TeleGeography 2000 Global Telecommunications Traffic Statistics and Commentary This work is based upon sources believed to be reliable, but the publisher does not warrant the accuracy or completeness of any information for any purpose and is not responsible for any errors or omissions. This work is for the confidential use of subscribers. Neither the whole nor any part of this publication may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopied, recorded or otherwise, with- out prior written consent from TeleGeography, Inc. All rights reserved. Copyright © TeleGeography, Inc. 1999 First Printing--November 1999 Second Printing--April 2000 This is copy TGO0-106141 Do not reproduce. ISBN 1-886142-21-1 Printed in the United States of Amedca TeleGeography, Inc. ¯ 1730 Rhode Island Avenue, NW ° Suite 400 Washington, DC 20036 USA Tel. +1 202 467 0017 Fax +1 202 467 0851 E-maih info@telegeography.com http://www.telegeography.com TeleGeography 2000 © TeleGeography, Inc. 1999 TeleGeography, Inc. This report was prepared by TeleGeography, Inc., an independent research and pub- lishing company based in Washington, D.C. The TeleGeography 2000 research and production team included: Gregory Staple .................. Editor in Chief Jason Kowal ................. Managing Editor Timothy Stronge .......... Senior Research Analyst Bram Dov Abramson ............ Research Analyst Aleetalynn Schenesky ........... Research Analyst Markus Krisetya ............... Research Analyst Jenifer Neidhart de Ortiz ......... Research Analyst Janice Gentile ................ Research Analyst Graham Finnie .............. Contributing Writer Kenneth Cukier .............. Contributing Writer Mickie Kennedy ............... Director of Sales Jessica Marantz ............ Director of Marketing Gregory Staple & lrena Slage ......... Cover Design The following also are available from TeleGeography, Inc.: Direction of Traffic 1999 (ISBN 92-61-08041-3) New International Carriers Americas Edition (ISBN 886142-19-X) New International Carriers Asia/Pacific Edition (ISBN 886142-17-3) New International Carriers Europe Edition (ISBN 886142-18-1 ) New International Carriers Online Database (www.internationalcarriers.com) TeleGeography 1999 International Traffic Statistics (ISBN 886142-15-7) Global Communications Cable & Satellite Map (ISBN 886142-16-5) Global Communications Traffic Map (ISBN 886142-20-3) The International Bandwidth Report (forthcoming ISBN 886142-22-2) For all inquiries, please contact: TeleGeography, Inc. 1730 Rhode Island Avenue, NW Suite 400 Washington, DC 20036 USA Tel. 1 202 467 0017 Fax +1 202 467 0851 E-maih info@telegeography.com For more information, please visit our web site: http://www.telegeography.com © TeleGeography, Inc. 1999 TeleGeography 2000 Contents List of Figures ............................................................................ 4 List of Country Traffic Tables ................................................................. 6 Preface ................................................................................. 7 Introduction: The Soft Network ................................................................ 9 Carriers The Growth of Internationa] Services Competition ............................................... 30 Market Shares of International Carriers ...................................................... 32 The Top 40 international Carriers .......................................................... 36 The Top U.S. International Carriers .......................................................... 37 Global Carrier Ownership Charts ........................................................... 38 Carrier Affiliates and Acquisitions .......................................................... 51 Pricing Elements of an International Call ........................................................... 59 International Call Costs from/to the U.S .................................................. 60 International Private Line Pricing ...................................................... 64 international Settlement Rates ....................................................... 65 FCC Settlement Benchmarks ......................................................... 67 National Interconnection Rates ....................................................... 69 Retail Pricing Retail Prices for a Three Minute Call .................................................... 70 Retail Pricing Trends, 1997-1999 ..................................................... 72 Origin of U.S. Carrier Revenues ............................................................ 73 Trading International Telecom Minutes ....................................................... 75 International Call Quality Metrics .......................................................... 76 Facilities The Bandwidth Wagon Rolls On ........................................................... 80 Submarine Cable Systems ............................................................... 86 International Circuit Usage by U.S. Carriers .................................................... 98 international Communications Satellites ..................................................... 101 The Internet international Internet Backbones .......................................................... 106 Internet Exchange Points ............................................................... 120 Global Internet Indicators.. ............................................................. 122 VolP Traffic and Settlements ............................................................. 125 Internet Governance: Then and Now ....................................................... 131 Traffic Analysis Overview of International Traffic Trends ..................................................... 139 Global Traffic Review International Traffic by Region ....................................................... 143 international Traffic by Country ...................................................... 147 international Traffic by Route ........................................................ 150 International Traffic on Mobile Networks ..................................................... 153 Alternative Traffic Arrangements .......................................................... 158 Country Traffic Statistics international Traffic Tables (see page 6 for a list of countries) ................................................... 162 Methodology ....................................................................... 271 Global Reference: Blue Pages National Telecommunications indicators ..................................................... 274 international Dialing Codes .............................................................. 279 A Primer on Bits ..................................................................... 288 End Pages About TeleGeography, Inc ............................................................... 290 One Hundred Years of International Telecommunications .................................. pull-out section TeleGeography 2000 © TeleGeography, Inc. 1999 List of Figures Introduction: The Soft Network Figure 1. The Broadband Future and Why It Matters .......... 11 Figure 2. Between Two Worlds: PSTN vs. Internet Architectures .. 12 Figure 3. Fiber Optic System Cost Trends .................. 13 Figure 4. Service Codes~atching the Next Wave ........... 14 Figure 5. Routing Codes--Trading Places .................. 16 Figure 6. Mobile’s Broadband Future ..................... 19 Figure 7. Open Codes--May the Source Be With You ......... 20 Figure 8. Changing international Service Models: From Half Circuit to No Circuit ................... 22 Figure 9. From Telco to Comco ......................... 25 Figure I0. Encoding Law--The Future of Regulation ........... 26 Carriers The Growth of International Services Competition Figure 1. The Intemational Carrier Boom .................. 30 Figure 2. Countries with International Telephone Services Competition ..................... 31 Market Shares of International Carriers ......................... 32 The Top 40 International Carriers ............................. 36 Global Carrier Ownership Charts Figure 1. AT&T Corp. and British Telecommunications plc ....... 39 Figure 2. AT&T--BT Global Venture ...................... 40 Figure 3. Bell Atlantic Corp. and GTE Corp ................. 41 Figure 4. Cable & Wireless plc ......................... 42 Figure 5. Deutsche Telekom AG, France T~l~com SA, and Sprint Corp ......................... 43 Figure 6. Global Cressing Ltd .......................... 44 Figure 7. Global TeleSystems Group ..................... 45 Figure 8. KPN, Qwest Communications, and US West Corp ...... 46 Figure 9. MCl WorldCom, Inc .......................... 47 Figure 10. SBC Communications, Ameritech Corp., and Bell Canada lnc .......................... 48 Figure 11. Telecom Italia SpA .......................... 49 Figure 12. Telef6nica SA ............................. 50 Carrier Affiliates and Acquisitions Figure 1. Selected Multinational Carrier Operations ........... 51 Figure 2. Major International Carrier Mergers and Acquisitions, Oct. 1998~ct. 1999 .................... 54 Figure 3. Major International Carrier/ISP Mergers and Acquisitions, Oct. 1998~ct. 1999 .................... 55 Figure 4. Major International Carrier/Cellular Mergers and Acquisitions, Oct. 1998-~9ct. 1999 ........... 56 Pricing Elements of an International Call Figure 1. International Call Components .................. 59 International Carrier Call Costs from the U.S ...................... 60 International Carrier Call Costs to the U.S ........................ 62 International Private Line Prices Figure 1. International Private Line Lease Prices from U.S., 1995-1998 ........................... 64 Figure 2. Band-X Bit Index, 1998-1999 .................. 64 International Settlement Rates ............................... 65 FCC Settlement Benchmarks ................................. 67 National Interconnection Rates ............................... 69 Retail Prices for a Three Minute Call ........................... 70 Retail Pricing Trends, 1997-1999 Figure 1. Retail Tadffs for Selected Countries ............... 72 Origin of U.S. Carrier Revenues Figure 1. Advertised Tariffs vs Billed Revenues .............. 73 Figure 2. U.S. Carder Revenues and Settlement Outpayments .... 73 Figure 3. U.S. Carder International Call Revenue by Destination, 1998 ....................... 74 Figure 4. U.S. Carder Revenues for International Voice Service, 1998 ...................... 74 Trading International Te]ecom Minutes Figure 1. Minutes Sold on Arbinet Exchange, 1999 ........... 75 Figure 2. Band-X Index of U.K. International Wholesale Prices, 1997-1999 ...................... 75 International Call Quality Metrics Figure 1. Quality Metrics on Calls from Germany and the U.K, June-August 1999 ....................... 77 Facilities The Bandwidth Wagon Rolls On Figure 1. Competition for Telecommunications Bandwidth Heats Up ..................... 80 Figure 2. Western Europe’s Fiber Frenzy .................. 81 Figure 3. The Bandwidth Market Matrix ................... 83 Figure 4. Simplified View of an International Circuit ........... 84 Submarine Cable Systems Figure 1. Behind the Bandwidth Explosion ................. 86 Figure 2. Components of a Submarine Cable System .......... 87 Figure 3. Map of Major Trans-oceanic Submarine Cables ....... 88 Figure 4. Map of Major Submarine Cables in the United States (Pacific Coast) ..................... 90 Figure 5. Map of Major Submarine Cables in the United States (Atlantic Coast) & Caribbean ................ 91 Figure 6. Map of Major Submarine Cables in Latin America and the Caribbean ................. 92 Figure 7. Map of Major Submarine Cables in Europe .......... 93 Figure 8. Map of Major Submarine Cables in East Asia ........ 94 Figure 9. Map of Major Submarine Cables in South Asia ....... 95 Figure 10. Major International Submarine Cables by Region ..... 96 International Circuit Usage by U.S. Carriers Figure 1. International Circuit Usage by Region, 1996-1998 ..... 98 Figure 2. Idle Circuits of U.S. Carders by Region, 1996-1998 .... 99 Figure 3. International Circuit Usage for Selected Routes, 1996-1998 ........................... 100 © TeleGeography, Inc. 1999 TeleGeography 2000 International Communications Satellites Figure 1. Orbital Map of Major Geostationary Satellites ....... 102 Figure 2. Major International Telecommunications Satellites in Geostationary Orbit ..................... 103 Figure 3. Geostationary Satellite Launch and Cost Trends ...... 104 Internet International Internet Backbones Figure 1. Inter-regional Internet Bandwidth, 1999 ........... 107 Figure 2. The Top 50 International lnternet Routes, 1999 ...... 10B Figure 3. The Top 50 International lnternet Hub Cities, 1999 .109 Figure 4. Map of Major Asian International Backbone Routes, 1999 ......................... 110 Figure 5. The Top 50 International Backbone Routes in Asia, 1999 ........................... 111 Figure 6. Map of Major European Internationa~ Backbone Routes, 1999 ......................... 112 Figure 7. The Top 50 international Backbone Routes in Europe, 1999 ......................... 113 Figure B. Map of Major South American International Backbone Routes, 1999 .................. 114 Figure 9. Top S0 International Backbone Routes in Latin America and the Caribbean, 1999 ........... 115 Figure 10. Map of Major International Backbone Routes in the U.S. and Canada, 1999 ................ 116 Figure 11. The Top 50 International Backbone Routes in the U.S. and Canada, 1999 .................. 117 Figure 12. Map of Major U.S. Domestic Backbone Routes, 1999 ......................... 118 Figure 13. The Top 50 U.S. Domestic Backbone Routes, 1999 ......................... 119 Internet Exchange Points Figure 1. Major Internet Exchange Points, A-K ............. 120 Figure 2. Major Internet Exchange Points, L-Z .............. 121 Global Internet Indicators Figure 1. Host Growth and Distribution by Region, 1997-1999..122 Figure 2. Host Count by Top Level Domain, 1997-1999 ...... 123 Figure 3. Host Count by Country, 1997-1999 ............. 124 VolP Traffic and Settlements Figure 1. Major Clearinghouse Networks and Standards ....... 125 Figure 2. IP/PSTN Equipment Makers ................... 126 Figure 3. How a VolP Clearinghouse Works ............... 127 Figure 4. A Guide to XolP Standards, Protocols, and Vendor Forums ........................ 128 Figure 5. IP Carders, Traffic, and IPOs ................... 129 Internet Governance: Then and Now Figure 1. Growth of Internet Users and RFCs .............. I31 Figure 2. The Internet Governance Stack ................. 132 Figure 3. ICANN: The Bottom Up ...................... 134 Traffic Analysis Overview of International Traffic Trends Figure 1. International Traffic in Asia, 1996-1998 ........... 139 Figure 2. Traffic Base of Incumbent and New Carriers ......... 140 Global Traffic Review Figure 1. International Traffic and Main Line Growth ......... 141 Figure 2. International Traffic, Revenue, and Subscriber Growth .141 Figure 3. Intercontinental Traffic Flows, 1995 & 1998 ........ 142 International Traffic by Region Figure 1. Interregional Traffic Flows, 1998 ................ 143 Figure 2. International Traffic by Origin, 1998 ............. 143 Figure 3. Traffic Growth by Region, 1997-I998 ............ 143 Figure 4. European Telecommunications Traffic Flows, 1998 .... 144 Figure 5. Latin Amedca Telecommunications Traffic Flows, 1998 .......................... 145 Figure 6. Asian Telecommunications Traffic Flows, 1998 ....... 146 International Traffic by Country Figure 1. Outgoing International Telephone Traffic Growth for Selected Countries, 1997-1998 ............. 147 Figure 2. Telephone Traffic Balances for Selected Countdes, 1998 ....................... 147 Figure 3. International Traffic Indicators, 1998 ............. 148 International Traffic by Route Figure 1. The Top 50 International Routes, 1998 ........... 150 Figure 2. Traffic Imbalances on Selected U.S. Routes, 1998 .... 151 Figure 3. Traffic Imbalances on Selected Non-U.S. Routes, 1998 ................... 151 Figure 4. International Outbound Routes with Rapidly Growing Traffic, 1997-1998 ................ 152 Figure 5. International Outbound Routes with Declining Traffic, 1997-1998 ............... 152 International Traffic on Mobile Networks: Current Trends and Pricing Figure 1. Mobile’s Share ............................ 153 Figure 2. Fixed vs Mobile Subscribers and International Traffic by Wodd Region, 1998 .............. 154 Figure 3. Retail Pdces for Intemational Calls ............... 155 Figure 4. Top Mobile Operators ....................... 155 Figure 5. The Tromboning Effect ....................... 156 Figure 6. The High Cost of Calling Mobiles, 1999 ........... 157 Alternative Traffic Arrangements Figure 1. Adternative Traffic Arrangements as a Share of International Voice & Fax Traffic, 1996-1998 ............. 158 Figure 2. Top ISR Routes with the U.S., 1997-1998 ......... 158 Figure 3. Call Delivery Methods ....................... 159 International Traffic Tables (see page 6 for a list of countries) Blue Pages National Telecommunications Indicators .................. 274 International Dialing Codes ........................... 279 Map of Worid Dialing Codes .......................... 282 North Amedcan Area Codes .......................... 284 Map of North Amedcan Area Codes ..................... 286 TeleGeography 2000 © TeleGeography, Inc. 1999 List of Country Traffic Tables Albania .................... 162 Algeria .................... 163 Andorra .................... 164 Angola .................... 165 Argentina ................... 166 Armenia ................... 167 Aruba ..................... 168 Australia ................... 169 Austria .................... 170 Azerbaijan .................. 171 Bahamas ................... 172 Bahrain .................... 173 Bangladesh ................. 174 Belarus .................... ! 75 Belgium .................... 176 Bolivia .................... 177 Bosnia-Herzegovina ........... 178 Brazil ..................... 179 Bulgaria ................... 180 Canada .................... 181 Chile ...................... 182 China ..................... 183 Costa Rica .................. 184 Cote d’lvoire ................. 185 Cyprus .................... 186 Czech Republic ............... 187 Denmark ................... 188 Dominican Republic ........... 189 Egypt ..................... 190 El Salvador ................. 191 Estonia .................... 192 Finland .................... 193 France ..................... 194 Germany ................... 195 Ghana ..................... 196 Greece .................... 197 Hong Kong .................. 198 Hungary ................... 199 India ...................... 200 Indonesia ................... 201 Iran ...................... 202 Ireland .................... 203 Israel ..................... 204 Italy ...................... 205 Jamaica .................... 206 Japan ..................... 207 Jordan .................... 208 Kazakhstan ................. 209 Kenya ..................... 210 Korea, Rep .................. 211 Kuwait .................... 212 Latvia ..................... 213 Luxembourg ................. 214 Macau .................... 215 Macedonia .................. 216 Malaysia ................... 217 Malta ..................... 218 Mauritius ................... 219 Mexico .................... 220 Moldova ................... 221 Monaco .................... 222 Morocco ................... 223 Namibia ................... 224 Netherlands ................. 225 Netherlands Antilles ........... 226 New Zealand ................ 227 Nicaragua .................. 228 Norway .................... 229 Oman ..................... 230 Pakistan ................... 231 Palestinian Authority ........... 232 Panama .................... 233 Paraguay ................... 234 Peru ...................... 235 Philippines .................. 236 Poland .................... 237 Portugal ................... 238 Qatar ..................... 239 Romania ................... 240 Russia ..................... 241 Saudi Arabia ................ 242 Singapore .................. 243 Slovak Republic .............. 244 Slovenia ................... 245 South Africa ................. 246 Spain ..................... 247 Sri Lanka ................... 248 Sudan ..................... 249 Sweden .................... 250 Switzerland ................. 251 Syda ...................... 252 Taiwan ..................... 253 Thailand ................... 254 Trinidad & Tobago ............. 255 Turkey ..................... 256 Ukraine .................... 257 United Arab Emirates .......... 258 United Kingdom-Outgoing ....... 259 United Kingdom-Incoming ....... 260 U.K. Top 100 Correspondents ..... 261 United States-Outgoing ......... 262 United States-Incoming ......... 263 U.S. Top 100 Correspondents ..... 264 Uruguay ................... 265 Uzbekistan .................. 266 Venezuela .................. 267 Vietnam ................... 268 Yugoslavia .................. 269 Zimbabwe .................. 270 © TeleGeography, Inc. 1999 TeleGeography 2000 Preface New Day As we enter the new century and the new millennium, no other industry is changing our world so quickly, impacting people around the globe so profoundly, or redefining our new day so definitively as global communications. Our industry is in a state of creative upheaval, gen- erating unbounded opportunities for both communications providers and those served by the communications industry. It is a moment when, as Albert Einstein said, "Imagination is more important than knowledge." What was a voice-dominated business is rapidly becoming a data-dominated business. Switched cimuits are morphing into routed packets, and former monopoly environments are turning competitive. Also, the mother of all disruptive technologies, the lnternet, is itself con- tinuing to undergo rapid change. A few years ago we thought of the lnternet as a path to the "global village," where technologically savvy intellectuals could go to communicate cheaply. But in an amazing instant, the lnternet has been transformed from a rather small community into a global metrepolis--creating opportunities for people in all professions and socio-economic groups around the world. It has become a global marketplace for goods as well as ideas, growing at rates larger than anything the world has seen previously. Fueling this rapid growth of advanced data services is bandwidth. Bandwidth availability is growing so quickly that it dwarfs the much-heralded metrics of Moore’s Law. Global band- width is doubling every 100 days with no slow-down in sight. Futurologist George Gilder calls bandwidth the "defining abundance" of our day, playing the same role as electricity in the industrial age and transistors in the computer era. And like the "defining abundances" before it, bandwidth shows extraordinary elasticity of demand. Falling prices increase demand so rapidly that revenue rises faster than prices fall. The bandwidth explosion on international telecommunication routes is among the fastest growing of all communications sectors. New competitors as well as existing providers are dri- ving an increasingly robust and competitive marketplace. The new routes nearing comple- tion across the Pacific--China-US, Japan-US, and Southern Cross--will increase by a factor of forty the total cross-Pacific capacity that existed previously. Also, across the Atlantic, bandwidth availability is increasing on a logarithmic scale. TAT- 14, scheduled for completion late in 2000, will handle an astounding seven million simultaneous voice conversations. And that is just phase one--there are plans to double this capacity by 2002. Bandwidth growth just continues to create the opportunity for more applications, reaching ever-wider audi- ences. Ultimately, the growth of bandwidth, along with that of the Internet is creating the Golden Age of Communications. As we go forth in these exciting and sometimes confusing times, it helps to have a reliable guide that shows a detailed snapshot of the flows of these global communications capabili- ties that drive this change. That is why MCI WorldCom and BCE Nexxia are proud once again to sponsor TeleGeography 2000. For customers and providers, for regulators and investors, and for futurologists and opportunists, this industry standard is essential reading. Walter Schonfeld, Senior Vice President International Development MCI WorldCom, Inc. Walter Blackwell, Vice President International & Carrier Services BCE Nexxia TeleGeography 2000 © TeleGeography, Inc. 1999 Acknowledgements W le wish to thank the numerous carriers, government departments, regulators, and international organizations from around the world who responded to our requests for information. This report would not exist without the help of the dedicated people at these organizations who took the time to ensure that the data reported here are as current and accurate as possible. We would also like to thank the people who helped review early and final drafts of this book, including: Stephan Beckert, Gordon Cook, Laura Cozijnsen, Hendi Crosby, Michael Froomkin, Mike Jensen, Tim Kelly, Robert Luczak, Sanford Marble, Mike Minges, Takeo Miura, Cindy Obenstine, Larry Press, Ellen Rony, John Ryan, Robert Shaw, Reijo Svensson, and Donald Weightman. TeleGeography 2000 was supported, in part, by a publication grant from MCI WorldCom, Inc. and BCE Nexxia. As in the past, however, this grant was made without any precondition; TeleGeography, Inc. is solely responsible for the report’s editorial contents. The Editors © TeleGeography, Inc. 1999 TeleGeography 2000 The Soft by Gregory Staple, Network TeleGeograph y, Broadband networks which are Internet-compatible and software-driven will change almost everything--services, routing, revenues--for today’s long distance telephone companies. Forward This essay is about the future of global telecommunications. It begins with a local story though--a story about one day last summer when a company named e.spire began laying fiber optic cables through my Washington, D.C. neighborhood. Why start here? The answer has less to do with e.spire or my hometown than the fact that every global network is also a local one. It has a beginning and an end. And, as the high- speed digital capacity at the center of the global network--say between Paris and Washington or between Los Angeles and Tokyo--is brought to the last mile, tens of millions of people will have access to an integrated set of trans-national voice, data, and video services. This will happen in a handful of countries at first, but over the next decade tomorrow’s network will be characterized by four main features. It will be: (I) broadband from end-to-end, meaning high-speed; (2) Internet-friendly, meaning that tele- phony and other services will use the Internet’s common pack- et-switched transmission protocols; (3) always on, like electric- ity or tap water; and (4) pervasive. On this network, like the Internet today, software applications created by thousands of different vendors will be the driving force--hence the title: "The Soft Network." The body of this essay explores this new terrain in more detail and describes in general terms how we may get from here to there. It is followed by a concluding section that looks at the impact high-speed Internet services will have on the revenues and products offered by major international telecom providers. Legacy carriers already face mounting economic pressure as the Intemet begins to price all bits alike, whether they are voice bits or data bits, and whether they originate next door or around the world. The closing section of this essay also briefly describes the new regulatory agenda which broadband packet networks are bring- ing forward. This agenda includes the terms on which compet- ing service providers will have access to "last mile" facilities as well as network performance, privacy and consumer protection issues. Now on with the story. Rites of Passage--Asphalt Tattoos It’s the end of the world as we know it. It’s the end of the world as we know it. And I feel fine .... REM The millennium came early to my street: on May 15, 1999. I remember the day well, having been jogged awake before 7:00 AM by a cacophony of late 20th century machines preparing the ground for their successors. Diesel engines, pneumatic drills, backhoes and concrete crushers. Who needs alarms? The work crew began at one end of the block and by the fol- lowing week, the street had been gifted with a clutch of poly- ethylene conduits which would soon carry more fiber optic strands than cross the Pacific Ocean. As I watched all the commotion, I was reminded of the "clouds" engineers love to draw for the Internet--you know, those wavy- edged balloons on a white board which seem to float some- where above the real world. But the truth was in front of me: cyberspace is less about clouds than ditches miles upon miles of ditches, each wormed with strands of laser-thin glass fiber. Down one street and along the next. Across one city and on to another. City by city, country by country. Today’s public works program is tomorrow’s telegeography. Yes, I thought, almost every click of a mouse begins with a shovel of dirt. Our local trench-and-fill caravan was led by a compact, pur- pose-built truck which straddled a tree trunk sized saw blade. It was a circular saw strong enough to cut through 14 inches of tire-worn macadam and concrete. The blade cut parallel inci- sions about two feet apart running the length of the block. The severed pavement was then unearthed by a small backhoe, after which a pick-and-shovel team dug out another two or three feet of soil and gravel. The bottom of the trench was lev- eled for the conduits. They came in 20-foot sections and were trussed together in sets of four: olive, blue, orange, and spring green. Joiners worked quickly; once the trench was open, it took only a few hours to put the pipes in place. Later, I was told, another crew would carefully pull strands of fiber optic cable through two of the pipes, leaving the remain- ing conduits empty for the future. Even so, the communications capacity which would thread its way down our block was enormous--enough to carry all the telephone calls in the world, the foreman assured me. TeleGeography 2000 © TeleGeography, Inc. 1999 It seemed astonishing, but there is nothing like a ditch across your doorstep to drive a point home. Our tree-lined street had suddenly become a sea-front property. Last week we had been landlocked. Now we were moored in a deep-water port on the busiest waterway in the world. The events on my block were being duplicated across Washington last summer. And, e.spire, the competitive local exchange carrier at work in our neighborhood, was only one of many new service providers active in the city. The Washington Post reported that nine different companies were busy laying fiber optic cable across the town. The Post called it the ’Great Dig of 99’: "From coast to coast, the digital age has arrived in America as dozens of companies dig up city streets to install fiber optic cables. Freed... from burdensome regulation, heartened by the explosion of Intemet use, these companies are part of a digital gold rush." The trench-and-fill operation that began on Capitol Hill in May had traversed much of the city by mid-summer. In July, 1 saw some of the same crew working their way down Pennsylvania Avenue to the White House. It led to a further thought: The lnternet has put America at the center of the telecosm. Yet, in most people’s minds, Washington remains far removed from the cybertribes of Silicon Valley. Now perhaps things will be dif- ferent. You may not see it on television. But in January 2001, when the next President of the U.S. takes the oath of office on Capitol Hill and begins the traditional parade down Pennsylvania Avenue to the White House, that storied boule- vard will have become a virtual thoroughfare to rival all com- ers. The summer’s long days of dry heat have now given over to a wet autumn and the construction crews seem to have melted away. The streets are still hazardous for shoe leather and car axles alike. Block-long welts of oily black asphalt ooze in the midday sun and subside with every rain. Yet, 1 have heard few complaints. The landscape’s ritual scarification by jackhammer and steamroller seems to be a source of quiet pride. Like its urban cousins around the world, Washington has come of age. The asphalt tattoos which snake across its streets bear witness to the city’s rite of passage into the new economy. 11. Code Is a Compass "If you don’t know where you’re going, you may end up somewhere else." Yogi Berra Alright. Let’s step back from this story to check our directions. Fiber optic communication cables are hardly novel. Most tel- cos have been replacing copper and coaxial cables with fiber on trunk routes for years and, in the U.S., new players such as MFS (now owned by MCI WorldCom) and Teleport (owned by AT&T) have been building fiber rings in big cities since the 1980s. So what is the point here? How does Washington’s rite of pas- sage relate to the business of providing international telecom- munications? Let me offer some short answers now and then flesh them out later on. I. Global services need local networks. The end-to-end capac- ity of any network is no greater than the bandwidth available over the "last mile." Thus, so long as the high-speed capacity which now exists in the middle of the network is poorly matched at the ends, the network’s potential will remain largely unful- filled. 2. The distinction between local and long distance networks, between domestic and international ones, are becoming mean- ingless. Because fiber optic cables provide essentially unlimit- ed bandwidth, the marginal cost of transmitting additional information on any portion of the route tends towards zero. 1 Long distance networks are becoming local area networks, and hence long distance and local transmission prices are converg- ing. To which there is an important corollary: All bits soon will be priced the same, whether they are "voice" bits or "data" bits. Witness the lnteroet. These new pricing rules have already begun to leave their mark in North America and will bite even deeper in the years just ahead (see Section lll.A.). 3. Networks are becoming service independent. Again, look to the Internet. That is perhaps the main reason why the business of providing network capacity, of laying down conduits, has become one job; the business of providing services quite anoth- er. Put another way, so long as the lnternet’s open-access model prevails (more on this in Part Ill), one can become a major communications company--like Yahoo! or ReaLAudio [I For example, Williams Communications, one of the U.S.’ new long distance carriers, reports that its longhaul fiber network will have 2 or 3 conduits for cables with the first conduit having at least 96 fibers. Using Dense Wave Division Multiplexing (DWDM), every fiber may carry data on 32 wavelengths, each operating at about I0 Gbps, thus giving the network a total capacity per fiber of over 3 Terabits per second. That is a very large throughput by any measure, and next generation WDM networks may operate at over 80 Gbps on 80 or 120 paths per fiber. By way of comparison, assuming average cir- cuit loading (5 hours per day), and without any compression, the total volume of 1998 traffic on the U.S. interstate public switched network--some 336 billion minutes--could be sent from one point to another using less than I/2 a Terabit of fiber optic capacity. © TeleGeography, Inc. 1999 TeleGeography 2000 Networks--without owning any transmission capacity. Which leads to a final point. 4. As bandwidth becomes ever cheaper and broadband net- works proliferate (see Figure I ), it is services which will largely differentiate one player from the next. In this digital age, ser- vices mean software. And software means code--code for con- versations; code for painting pictures; code for routing traffic; code for authoring new codes; and code for masking old ones. In this sense, code acts like a compass. It points toward the network’s future, showing us the underground springs of traffic which today are just a trickle but tomorrow will be a Mississippi. Look at the codes if you want to forecast tomor- row’s traffic booms (see Figure 4) and how they may be man- aged (see Figure 5). Look at the codes if you want to see where network access is open and where not (see Figure 7). Look at the codes if you want to know whether the route forward will be private or public and, correspondingly, where (and how) regu- lators may intervene (see Figure 10). But I am getting ahead of myself. It may be useful first to try and provide additional context for the "Great Dig of ’99" by drawing a more complete picture of the networked world that Washington is joining. For in so many ways, the picture is quite novel. What was once a formal garden painted in familiar hues has, almost overnight, become a jungle plot whose unchecked growth and riotous palette threaten to overwhelm us. Consider the time line which accompanies this book (see insert in the back of the book). In 1900, there were but I0 million phone lines worldwide. It took another 60 years for that number to reach 100 million (of which 40 percent were still in North America) and 30 years Figure 1. The Broadband Future and Why It Matters [] xDSL [] Cable modems [] Satellites [] LMDS Note: Does not include ISDN Million subscribers 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 By mid*2000, there should be about three million broadband subscribers in North America, composing nearly half of seven million worldwide subscriptions. That may rise to over I36 million by the end of 2007, according to Pioneer Consulting, who sees cable--now the dominant broadband access technology--settling in at 30 per- cent of the market. At the bottom end, Local Multipoint Distribution System (LMDS), using 28 GHz microwave signals, is expected to be the least-used technology. As a wireless broadband aiternat~e, it’s easy to install, bypassing the first-mover advantage of incumbent telecom and cable providers. But to maximize signal performance, LMDS’s high frequency channel requzres Source: Pioneer Consulting LLC, 33.6 Kbps modem 56 Kbps modem 128 Kbps ISDN B Channel 13 minutes 8 minutes 3 minutes 784 Kbps xDSL 32 seconds 5 Mbps Cable modem 5 seconds TeleGeography 2000 © TeleGeography, Inc. 1999 PSTN STP STP PST~/IP Gat,~way IP Networks Router Router RSP Dial up PABX Customer SS7 Components: SSP: Service Switching Point--enables central offices (CO) to communicate with SCP database SCP: Service Control Point--has databases with all routing and billing inforrna- tion STP: Signal Transfer Point--routes queries and responses beWveen COs and SCPs Dial up Customer ISP: Intemet Service Provider RSP: Regional ISP IBP: Internet Backbone Provider PABX: Private Automatic Branch Exchange PABX © TeleGeography, Inc. 1999 The challenge of making the circuit sw tched telephone network SS7 (Signaling System 7) is the protocol used on the separate chan- compatible with the packet-switched lnternet is at the top of the nel created by telcos to send call set up and control and billing agenda for telephone and Internet service providers alike. Not so data. It is a critical part of the tN architecture. Separating the long ago, however, creating a multimedia platform was viewed pfi- billing and control network from the c rcuit used to canty a call marily as a telephone company project. In response, telcos built makeSit far easier to ntroduce new biiting or routing services, such today’s intelligent network (lN) based upon di~l switches, smart as free phone, call forwarding, conference calling and the like. software and a new broadband transmission protocol. The main _ ........ But, the smarter the te ephone network has become for switched Du~tmng D~OCKS are: phone service, the more trouble it has had in adapting to the ISDN (Integrated Services Digital Network) is a digital, worldwide lnte_n~ revolution at its edges. The lnternet is comparatvely public standard for sending voice, video and data over the public "dumb ’; the intelligence needed to carry new services (e.g., musc, switched telephone network. There are two varieties: basic rate video graphics~ is provided by the applications it runs~ not by the (I 2_8 Kbps) and primary rate (t .54 Mbps). Both are dial-up set- network. Thus, although today engineering groups are now working vices that :use existing copper wires. As yet though, deployment is to mesh the next generation arch tectures of the telephone network limited. ISDN reaches less than one percent of U,S. telephone sub- and the lnternet, in the short run lnternet/PSTN geteways may be scdbers, but in France, Germany and Japan, the take-up is higher the ma n way for multimedia serv ces to move from one network to (in Germany alone there are about five million ISDN lines), the other (for more on the comparative merits of IP vs. IN architec- tures, see www.tndenton.com[netheads3.htm). a. ain for the number to approach 500 million. But in the next (It was not always so, see Figure 2). As noted earlier, this con- decade (1990-2000), the network doubled in size; 400 million sensus holds that tomorrow’s network will be: (I) high-speed; mobile connections were also added. And the Internet wove a (2) packet-switched; (3) always on; and (4) pervasive. Let’s further 200 million users into its global skein. The network we look more closely at each of these features. know today is really but a child--barely I 0 years old. This very rapid change in the telecom landscape has neverthe- less been accompanied by growing agreement about the future. © TeleGeography, Inc. 1999 TeleGeography 2000 Trends i Trans-Atlantic Submarine Systems Total Capacity (Gbps) 7,000 6,000 5,000 4,000 3,000 2,000 1,000 Cost per Circuit Cost per Circuit $1,800 $1,600 $1,400 $1,200 $1,000 $8OO $600 $400 $200 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 North American Terrestrial Systems Millions Fiber-kin 5 Average Cost Average Cost per Fiber-km $140 1997 1998 1999 2000 2001 2002 2003 2004 2005 $120 $1o0 $8o $80 $40 $20 Note: Data reflect TAT cabte systems construction costs only, amortized over the Note: Fher deploymentand cost estimates sre for North American competitive local expected lifetime of each system, exchange carriers only. Source: ToleGeography, Inc. Source: Pioneer Consulting, LLC, G/obaf Broadband Access Markets Successive generations of fiber optic transmission systems have Fiber optic transmission systems are a classic "disruptive technol- sharply decreased the cost of carrying telecommunication traffic, ogy," as defined by Clayton Christensen’s popular 1997 book, The As a recent Bell Labs review noted, "’the operating speed of inte- Innovator’s Dilemma. According to Christensen, a disruptive tech- grated circuits has increased by a factor of 60 in the past 10 nolog~ is one where the rate of improvements in the technology is years; the capacity of optical fiber systems grew by a factor of 200 faster than the increased product performance demanded in the in the same period." The tiny core of an optical fiber may have u p market. Disruptive technologies give customers more than they to 50 trillion hertz of raw bandwidth. Current commercial systems currently need or are typically willing to pay for. That is why such use dense wave division multiplexing (DWDM) to mute data over technologies are usually commercialized first by newcomers serv- discrete optical frequencies ["colors") on each fiber to obtain ing niche markets (e.g., data networks; competitive local access transmission speeds of 500 billion bits per second. Laboratory providers). By the time the incumbent begins to adopt the tech- experiments have doubled that speed "...enough to carry every nology for its own customers, however, those customers may have nation’s voice traffic on a wave guide that looks for all the world migrated elsewhere. like a monofilament fishing line." © TeleGeography, Inc, 1999 A. High Speed We’ll start with a definition: about 384,000 bits per second (bps) or 384 Kbps. That’s a practical definition of a high-speed communications network. With that capacity one can browse the Web and sample audio or video clips along the way as if one were flipping the pages of a book. Definitions of high speed vary, however. The throughput of a network, and hence how quickly it can deliver data-intensive services, depends upon the bandwidth of the transmission medium (the wire or radio channel; see Figure 1). Accordingly, the definition of "high speed" and "broadband" (these terms are used interchangeably here) tend to vary depending upon the desired application, and some engineers would not consid- er a network to be broadband unless it operated at a speed of 1 Megabit per second (1,000 Kbps) or more. The public telephone network is indisputably low-speed; it was designed to carry voice, not video. The transmission circuits and switches allocate just enough capacity to handle a conver- sation with reasonable fidelity (3 Khz for analog networks and 64 Kbps for digital ones). The long distance and local trunk facilities used by telcos operate at much higher speeds. But, over the last mile, available bandwidth is only a fraction of that although, as we shall see, the copper wires which connect most telephones to the network are able to carry over 1 Mbps using Digital Subscriber Line (DSL) technologies. To make local access to the telephone network high-speed, one must modify (or replace) the transmission circuits available on the last mile. And one must change the electronics at end- office switches or subscriber terminals or both. Network wide, that can be very costly. Moreover, in most countries, telcos must factor in the competitive response of cable %1 operators TeleGeography 2000 © TeleGeography, Inc. 1999 cols (TCP/IP) make it comparatively thing impossible with HTML XML likewise can support style sheets easy to separate the job of transmit- for text to speech translation on the Web. That is likely to foster a ring information from the business of new genre of "talking browsers," ideal for mobile applications and originating it. This means that, on the for users with visual disabifities. A large cross-industry forum (IBM, Internet, new services need not belong Sprint, Ericsson, Sun) also has formed to craft an XML-based spec- but can be ification for voice-enabled tnternet applications. The first specs user-driven rather than network- were released in August 1999 (see www.voicexml.org). based. Many of the wodd’s mobile carders also have agreed to use an By contrast, on the public telephone network, it is the network oper- XML-compliant code, known as WML [Wireless Markup Language] ators and their preferred equipment suppliers which generally are for the microbrowsers that will display the Internet on the screens of the gatekeepers of new services (e.g., call forwarding, conference cell phones and the next generation of personal digital assistants, calling; see Figure 2). The Internet model--driven by thousands WML provides the user interface for WAP (Wireless Application of independent software groups vying to run their latest code on the Protocol) an open standard that enables wireless devices to access Net--consequently has radical implications for the way tetcos do the Internet [see www.wapforum.com). business. It affects the introduction as well as the routing of new The vast range of new XML applicat ons now being services [see Figure 5), coded has ted some to herald XML as the "second coming of the Web." A list of Perhaps the best example of how much things have changed is XML development groups can be found at www.w3.org/xmt. HTML (Hyper Text Markup Language), the standard code used to MP3/MP4-- Just as XML may drive a new wave of e-commerce and display text and graphical objects on the Internet. In the summer .................. voice traffic on the Net standard codes for downtoadin~ audio or ~[, wnen rltML was first pos~:ed on ~ne Net it was a rather -Ji ................. {MP3) and video (MP4} nformaton on the Web already have rua men~ary son:ware ~ool (]eve opea t)y a [~nen/ unKnown t~ritsn forced carriers to rethink the demand for network capacity. mathematician, Tim Barriers-Lee, to link documents on different computers, at a rather obscure research facility (CERN} in Streaming music and video on the Net has been possible for years. S~tzertand. Seven years later, our research suggests that HTML Software distributed by Real Networks Inc. is probably the best services probably generated more international traffic than the 800 known (see www.real.com) {Microsoft also has a product known as million telephones connected to the public fixed network (see Net.Show). Today, hundreds of radio broadcast stationsaround the Direction of Traffic 1999 for more details}, world stream their services over the Net {see www:netradio.com) and new movie and video services are announced almost daily. And that is only the beginning. Each week, the lnternet becomes a original content. Uncompressed CD-quality music files are very XML (Extensible Markup Language)--Launched in t998, XML HTML (e.g., those for displaying text as title, paragraph, row or col- umn) with XML one can add content-oriented labels {say, for a put- by the Motion Picture Experts Group (MPEG) in 1992, MP3 offers tags for price or s ze}. to the Internet, and then download them anywhere in the world-- As ell XM good for network traffic but bad for copyright holders. w, L supports unicode, a standard method for encoding characters in all the wodd’s major languages, thereby allowing com~ © TeteGe0graphy, lnc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Figure 4. Service Codes---Catching the Next Wave (continued) Music is big business. Worldwide sales of recorded music totaled record companies can be paid per tune, for details visit $38.’/billion in 1998 with the U.S. accounting for about a third of ............ the sales. To manage the copyright issues posed by MP3 and its The potential for the telephone network to function as a global cousins {there are several competing codecs available}, over 120 jukebox has not been lost on some telcos. AT&T has developed its record companies have formed a secure digital music initiative {SDMI}. own proprietary digital music format for the Net, known as A2B {www.a2bmusic.com}. An AT&T a~tiate also hosts the site of In August 1999, SDMI announced that it favored a digital water- mp3.com, one of the largest sources of MP3 encoded music on the marking technology known as MusiCode, developed by Ads Net. Technologies of Cambridge, Massachusetts. With MusiCode, any In case you are wondering, MPEG-~ provides a digital, CD-Rom- SDMI-compliant music player {such as the pocket-size devices pop- quality video standard. MPEG-2 is the standard for digital video utarized by Nomad {www.nomadworld.com} would be able to iden- disks (DVD}, and MPEG-4 offers a next generation multimedia tify watermarked music tracks from unauthorized MP3 sources on which already own a second (potentially broadband) wire to many customers. Until recently, therefore, with the exception of Integrated Services Digital Network (ISDN) offerings in a few countries (typically operating at 128 Kbps; see Figure 2), telcos have been cautious about upgrading the "last mile." They have pre- ferred to install fiber first on ]onghaul routes, where the cost per mile has fallen dramatically (see Figure 3), and secondly, on local broadband facilities for business customers who can afford premium service. Residential and small business markets were largely ignored. Then came the lnternet. As the Web transformed the Net into a mass market, by the mid-1990s the economic prospects for broadband over the last mile began to change. At last, there was a service which might actually let carriers recoup the cost of bdnging high-speed access to millions of new customers. Slowly, at first, but then with increasing urgency, incumbent tel- cos began to respond, spurred on by tens of new market entrants. During the last year across Europe and the Americas, there has been an unprecedented rush to buy, build or lease the facilities needed to offer end-to-end broadband services ahead of the competition. This global scramble for bandwidth has been a boon to scores of new network players in dozens of cities, including Washington, D.C.’s e.spire. Internationally, the activ- ity has been equally frenetic. Level 3, GTS, COLT, Global Crossing and Viatel are but a few of the companies which plan to offer broadband capacity to and from the world’s major cities (An authoritative survey of this new market will be pro- vided by TeleGeography’s international bandwidth guide; for a preview, see the Facilities section of this report). Over the "last mile," however, telcos typically have two choic- es: (a) to supply DSL services over their existing copper wires or (b) to provide access over the broadband facilities of third parties, i.e., cable TV systems or fixed wireless networks e.g., Multichannel Multipoint Distribution Services (MMDS).[2] A DSL modem taps the unused frequencies available in a stan- dard copper wire to send and receive data at speeds of 384 Kbps to 1.54 Mbps or more. However, DSL services generally are only feasible if the local loop between the subscriber and the telco’s end office is less than about 2.5 miles. A cable FV modem allows two-way digital communications at 5 Mbps or more, provided the cable system is properly configured. As with DSL technologies, with cable modems the downstream speed (to the subscriber) typically exceeds the upstream rate (from the user). Most major international players have now begun to marry their local broadband strategies with their long distance plans. For example, MCI WorldCom, whose UUNet subsidiary runs one of the world’s fastest Intemet backbones, has teamed with sev- eral ISPs and competitive local exchange carriers, notably Rhythms Net, to offer U.S. customers xDSL service over the last mile. MCI WorldCom also has bought several MMDS oper- ators, as has Sprint, which recently agreed to merge with MCI Woridcom. By contrast, AT&T proposes to buy two of America’s largest cable TV companies (TCI and Media One), whose 17 million subscribers would be tied to the Net exclu- sively via Excite@Home, AT&T’s affiliated lntemet Service [2] As more governments allocate sufficient microwave spectrum for fixed wireless operators, this service may well become a preferred vehicle for deliver- ing broadband services over the last mile to multiple dwe]linl~ units and medium-sized professional buildings, particularly where cable TV systems do not exist. In the U.S., for example, companies like Telil~ent and Winstar are now significant competitors in the local access market. TeleGeography 2000 © TeleGeography, Inc. 1999 Figure ~or such as XML and MP3, drive ever instance, Akamai’s customers are given a simple utility to modify larger volumes of traffic over TCP/tP the tags for every Web resource covered by the FmeFtow service. networks, other codes are being craft- The software rewdtes a file’s URL so that a particular file can be ed to manage the flow of these new routed to any one of hundreds of servers depending upon current traffic streams. These new load-bal- network conditions. The location of a content file is thus a function ancing techniques may well supplant of the dynamic routing algorithm resident in Akamars network of some of the network-based traffic servers: it is a-~eographical. management functions used by incumbent te cos They a so may Akamai has co-located servers on over 30 backbones indud ng affect the demand for new fac ires. Eady adopters of oad-man- .. .. networks run by Teleglobe and S ngnet Each geographica region agement software say smart co(]es can boost the performance or a has one mapping server that monitors the local state of the net- popular lnternet site more efficiently and without the fixed costs work [their job is to figure out which server should carry which file associated with acquiring additional bandwidth and how to distribute file requests amongst servers) and one or Akamai Technology in Cambridge, MA (www.akamai.com), and more content servers. Web pages themselves break down into Sandpiper Networks in Westiake Village, CA (www.sandpiper.com] units: the HTML page plus imbedded file--images, animations, are the current market leaders in this new fie d of software-based sounds, video. Major customers of Akamai include CNN and toad balancing. Akamai’s FreeFtow service and Sandpiper’s Yahoo; Sandpiper serves MSNBC and AOL, among others. Footprint each allow Web-based content providers to improve user _ access speeds, Both services work by port ng a subscriber’s con- Bob Collett, vice president of Data Services at Teleglobe, told the Cook Report {www.cookreport.com], "Akamai and Sandpiper have tent to a private network of servers; modifying the HTML tags asso- ciated with each Web object (text, graphic, video clip); and then applying special algorithms [mathematical formulas) so that a user clicking on a given link is automatically directed to the hi~hest per- forming server in the network. The key to their success lies in the way both Akamai and Sandpiper break the historical link between a network address (the domain name or URL) and the physical location of the corresponding data {servers and backbone links}. the potential for making some fundamental changes in the Internet business." It may presage a shift from "mana~:~ecl co-location ser- vices which are generally placed at key network access points or routers to intelligent mirror sites at the edges of our network, thus making performance of the mirror more directly related to the whole network, rather than a few routes." The not-so-hidden mes- sage is: Akamai is good for global networks with route diversity. Pont-to-point p ayers serv ng niche routes may not fare so well. As Paul Sp nrad explained in Wired {"The New Cool Akamai ........... ...... Akamai and Sandpiper are not the first to use d str bated traffic uvercomes the mternet s Hot bpot ~’roblem," AUgust 1999), "the ........... .... management but their software-based systems are un que Older ~e~ was on~natly Qes gned las! roads connecting distinct sources Systems, still available and n use do not a ter the HTML code of of content... An address like nasa.gov woud always correspond the packets which they manage According to Wired examp es to oed catea servers ~oca~ed at tiqe N~ facility. When you visit- include GTE lnternetworking, which acquired its Hopscotch product ed www.ksc.nasa.gov to see a shuttle launch, you connected to NAS~s servers at Kennedy Space Center" But as millions of Web users began to jockey for access to the most popular sites and as audio or video clips on sites made files bigger and bigger, the Internet protocols which tied information to a ~iven server began to cause bottlenecks and lead to event-driven "hot spots." from BBN, and the Central Dispatch solution offered by Resonate. Cisco has also developed a "tag switching" technology which iden- tifies and tags flows of associated traffic, such as streaming video clips; instead of exam ning each packet, routers need only read the tag, enabling faster performance. Other router vendors have pub- lished similar technologies and the tnternet Engineering Task Force Most content providers have responded by asking their ISPs to add [IEFF) has proposed an industry standard called multiprotoco abel bandwidth or by mirroring and caching content at multiple sites, switching (MPLS)to speed the flow of multimedia IP packets (see Akamai and Sandpiper both run their own private "mirror servers," www.ietf.or~ids.by.w~/mpls,html). but with one key difference---the content on their servers has been © TeleGeography, Inc. 1999 Provider (ISP) (This strategy is strongly opposed by other ISPs and it may ultimately prompt AT&T to divest Excite @ Home to win regulatory favor). Across the Atlantic, British Telecommunications, which has agreed to pool all of its high- speed international facilities with AT&T via the Concert joint venture, will chiefly rely upon xDSL for the last mile (regulation bars it from using cable %1 facilities in the short term). In a similar vein, the 1998 annual report of Deutsche Telekom (DT) advises that DT "plan[s) to serve [its] international cus- tomers by seamlessly linking" the network for international traf- © TeleGeography, Inc. 1999 TeleGeography 2000 tic in Germany, its domestic high performance network (a 12,000 kilometer fiber optic net joining major economic cen- ters) and its 36 City-Netz fiber rings with a European backbone network. By 2000, DT states that this network will span 20,000 kilometers and link some 40 points in 16 countries. Over the last mile, DT will rely upon ISDN and xDSL services; regulatory pressure has led the company to divest its nation- wide cable TV system, the largest in Europe with 18 million sub- scribers (Microsoft and Rupert Murdoch’s News Corporation, among others, have reportedly made the short list of bidders). Whatever the strategy, bringing broadband to the last mile will require time. Even in rich countries, most forecasts put high- speed access to the Inter-net at no more than 10-15 percent of households until 2003 or 2004. And even then, it is not at all clear that network owners will be the big economic winners. To see why, let’s turn to the second main feature of tomorrow’s networks--packet-switched. B. Packet-Switched The Internet’s packet-switched protocols are likely to become the common denominator of tomorrow’s networks although, as we shall see, they may "ride" on other standards such as ATM (Asynchronous Transmission Mode), a method of transmitting Internet traffic along with other digital services. Remember first that TCP and IP are protocols--software, not hardware. As well, it bears emphasis that the TCP/IP protocols always have been open standards available without charge on the Internet--a precedent with far-reaching implications for other Internet-based applications. More on this below (see also Figure 7). TCP (transmission control protocol) is a standard for communi- cating digital information as a series of packets that may be separately transmitted, but reliably reassembled at the other end. TCP thus provides the glue that binds the Internet togeth- er. It ensures that communications can be sent from one ter- minal on the Net to another, even though there is no direct con- nection between the terminals and even though the networks transited are otherwise incompatible due, for example, to the use of different hardware or operating systems. IP (Internet Protocol) defines the structure of data packets transmitted over the Internet. It also provides the basis for assigning unique addresses (numbers) to network terminals so that data packets which are properly coded can be routed across multiple (and often incompatible) networks. An IP address consists of four numeric blocks and looks like this: 209.8.101.138 (for www.peacockmaps.com, which hosts TeleGeography’s new Internet posters). The general acceptance of TCP/IP as the network protocol for sending and receiving almost any type of information---e-mail, phone calls, .music, films--has two hugely important conse- quences. TCP/IP protocols make telecommunication services: (1) network-independent and (2) application-blind. I. Network-Independent The lntemet is, by definition, a network of networks and, as we have seen, TCP was designed to deliver bits from one commu- nication terminal to another regardless of the number (or type) of the networks in between. For service providers, therefore, TCP/IP is inherently pro-competitive. As well, because TCP works on a best-efforts basis--checking to be sure that all packets have been received and resending those that are miss- ing---end-to-end performance for any given service may be improved if alternative networks are available (e.g., because one network is congested or experiences an outage). The connectionless, network-independent routing of services which TCP/IP makes possible has cross-cutting implications for existing telephone network operators. On the one hand, by making their networks Internet compatible (e.g., by adding InternetlPSTN gateways), they can shift part of the capital costs of building out and upgrading network facilities to third parties. If the network is software-based, why own all of the hardware? Let someone else pick up the tab. On the other hand, the competitive network model implicit in TCP/IP may radically lower the entry barriers for new network players as well as expanding service competition. In addition, if a telco must rely upon unknown third parties for end-to-end services, it may lose the ability to control the quality of the ser- vices upon which its customers have come to rely. Thus, a telco’s inability to guarantee network security or to meet per- formance targets (Quality of Service (QoS)) may outweigh the benefits from off-loading construction costs. (Private TCP/IP networks run over a small number of dedicated facilities may offer an alternative though. See Section ll.C. below.) 2. Application-Blind TCP/IP networks are likewise application-blind. In popular terms, that means the Internet is a "multimedia platform," but this really misses the crux of the matter. The actual services provided to end users on the Internet are not defined by TCPIIP but by higher level application protocols such as HTTP (hyper text transport protocol, the basis of Web services) or SMTP (simple mail transfer protocol, for e-mail). Thus, in contrast to the public network, where new services are typically tied to net- work-specific software (see Figure 2), TCPIIP makes it possible to rapidly introduce a new application; the underlying software used to transmit the application across the network remains the same. That makes new code, not conduits, the driving force behind new services (see Figures 4 and 5). This is worth saying again in a different way. Because the Internet is application-blind, if you want to have any idea of what the mix of services will be or the demand for more capac- ity one, three or five years out, you must closely monitor the TeleGeography 2000 © TeleGeography, Inc. 1999 locales where new applications are being developed, i.e., code space, not network space. And, as with the lnternet’s basic TCP/IP protocols, the more open the code, the greater may be its long term impact (see Figure 7). C. Always On Let’s move on to a third feature of tomorrow’s network: "always on." The lnternet has become synonymous with "round the clock" networks. Someone is always online when you’re not--trad- ing, hacking, posting. 24/7 networks are the bane of many people’s existence but the salvation of others--for instance, making vital health and public safety services available on demand. And yet, for most people, the Intemet is not really "always on" at the flip of a switch, like electricity or water. Even when a switch has been turned, the lights may flicker and the tap run dry. The 99.999 pement ("five nines") reliability we expect on the public telephone network still seems to be a distant dream on the Internet. Worse still, the more widely available broad- band access becomes, the more heavily trafficked will be the network. Keeping the network "always on" under these condi- tions is likely to be a major challenge for years to come and will require attention from almost everyone who uses or provisions network services, from local hosts to upstream routers to back- bone facilities. Let’s start with the access issue. For the lntemet to appear "always on," a permanent window on your electronic desktop, a dedicated high-speed connection is required. Yet, as noted, only a small number of users (businesses and universities aside) now have such access. To access the Net today, most sub- scribers must first secure a telephone dial tone and then log on to an ISP, a process which may take 20 or 30 seconds on good days but very much longer when demand peaks. Then one must get to the desired Web site. In fact, this part of the trip seems to be getting easier, though it may be marked by delays along the last mile. Most indicators show the lnternet’s per- formance has improved over the last year despite adding 50 million or more users. For example, Key Note Systems has found that the average time to access the Web pages of 40 major business sites in the U.S. decreased by half to approxi- mately five seconds during the first nine months of 1999. Subscribers with dedicated access to the Internet have benefit- ted accordingly. But caveat emptor. Always on does not mean "always works." TCP/IP is a best-efforts protocol. By itself, depending upon the network paths available, TCP/IP may never provide the QoS available on the public switched network where a dedicated (and, often, carrier-owned) circuit is assigned to each conversation.[3] The OoS problem thus con- tinues to challenge many of the Internet’s talented engineering and software groups. Some believe that the problem will only be solved by routing traffic over private intranets which can guarantee performance and keep unwanted traffic off their facilities. Hence, the push by several.major carders to use their own facilities to provide virtual, private TCPIIP services to large corporate clients. Likewise, performance standards also have led major Intemet backbone operators (telco-based or not) to manage traffic flows by using other standard protocols to package and trans- mit IP flows with bit streams originating on other networks. A favored approach is ATM. With ATM, traffic can not only be sent at high speeds (up to 2.5 Gbps) but the network operator can accord priorities to different types of packets or cells, thus making QoS possible. In engineering terms, therefore, IP is commonly "run over ATM" and today ATM is the networking protocol of choice for convergent traffic systems. [4] Still other engineers believe that OoS can be addressed by rely- ing more heavily on various IP protocols such as RSVP (Resource Reservation Protocol) which essentially alerts IP net- work switches in advance that a certain path will be required, thus speeding the transit of the subsequent data packets. Another approach is offered by the multi-protocol label switch- ing (MPLS) standard proposed by the Internet Engineering Task Force (IETF) to tag priority traffic. Norte] Network and Cisco both offer "tag switching" products for managing IP networks and several major companies including Concert (the AT&T-BT [3] It is worth explaining briefly that today’s public telephone networks do not usually dedicate an end-to-end circuit to a call. Rather, to increase the network’s capac- ity, numerous calls may be multiplexed over a given channel Thus, apart from the truly dedicated connection along the copper wire between a subscriber’s terminal and the local switch, the so-called "dedicated circuit" assigned to a voice call is really a set of time slices or frequency assignments on a circuit used for multiple calls, especially on long distance networks. [4] Some engineers nevertheless contend that whole layers of network technology, such as ATM, can be eliminated. Today, as discussed, installing a multiservice net- work with high reliability and QoS involves IP (at Level 3), ArM (at Level 2) and a pdmary layer of protocols to mesh electronic transmission streams with optical transport facilities, each of which adds something to the end-to-end solution. Yet, a new protocol is expected to emerge that combines the attributes of IP and ArM to provide for interoperability and end-to-end quality of service. The transmission and optical layers will merge to form an IP over DWDM (Dense Wave Division Multiplexing) layer, which will have the service characteristics now associated with today’s optical layer, typically based on SONET (Synchronous Optical Network)--a family of fiber-optic transmission rates, frame formats and control standards that allows interworking of optical systems manufactured by different vendors. When that happens, IP will finally become, in the words of MIT’s David Clark, "the great spanning layer" between the glass fiber below and tee bitstream of ideas above. © TeleGeography, Inc. 1999 TeleGeography 2000 Figure 6. Mobile’s Broadband Future []Digital []Analog 1985 1990 Source: ITU Total Global Cellular Subscribers (millions) 250 200 150 I00 5O 0 1995 1998 Multimedia Capable Systems 384 kpbs 2 Mbps Present and sma# new spectrum chunks GSM++ EDGE ) I GSM Core Architecture WCDMA ) Large new spectrum chunks Source: Ericsson Compared to the traditional technology used for GSM cellular telephone service (9.6 Kbits/second), new 3rd generation terminals using WCDMA (Wide Code Division Multiple coverage Access) support data rates from 144 to 512 Kbits/second within a wide area. At fixed loca- tions, 2 Mbits/second rates are achievable, making possible a range of wireless multimedia services. © TeleGe0gra phy; Inc~ ~999 venture), UUNet (owned by MCI WorldCom) and Frontier (now a Global Crossing company) have announced plans to deploy some form of MPLS. Whatever the method, the business of providing OoS and trou- bleshooting network failures will become a bigger and bigger industry as broadband local access delivers ever-increasing vol- umes of network traffic. And the demand for assistance already is making itself felt all down the line. The larger and more diverse the network becomes, the harder it will be to predict the network’s weakest links or where "flash crowds" or "hot spots" may disrupt service. Much as power pools, back-up generators, and circuit breakers are a standard part of our elec- tric world, so too will e-commerce make their digital counter- parts commonplace in the online world. TeleGeography learned this firsthand one Thursday in September 1999. At 8:20 AM, a certain "cmdrtaco"--the erstwhile Rob Malda, founder of Slashdot (www.slashdot.org)-- relayed a "this is cool" post from a loyal foot soldier about pea- cockmaps.com. By 8:25 AM, thousands and thousands of Slashdot’s faithful tried to enter our portal, overwhelming our current network resources (they have since been upgraded). D. Pervasive It is time to turn to the last feature of tomorrow’s networks-- their scope. And here something paradoxical seems to be hap- pening. On the one hand, there is little doubt that the future networks will be more pervasive than ever, a constant helper and witness--and, yes, snoop--at work and at home. On the other hand, access to broadband services will remain limited well into the next decade for most users. Recall Section II.A. The network thus will grow faster vertically than horizontally for some time to come. A global service vision has long animated the telecom indus- try’s largest players. Any time, any place is also the battle cry of the mobile satellite industry, especially the proponents of next-generation systems such as Teledesic and Skybridge, which promise high-speed Internet backbones in the sky. Meanwhile, back on earth, the cellular telephone industry has its own ambi- tious plans to migrate its large customer base --now 400 mil- lion and growing--to a third generation (3G) service with data rates up to 2 Mbps for fixed sites (see Figure 6). Even so, most telephone users and, indeed, the majority of non-telephone users, are likely to be affected by new high- speed networks indirectly first due to the more rapid integra- tion of narrowband TCPIIP services into a larger economy. As Intel (and others) put the Internet inside almost every con- sumer device, and as software agents become more and more powerful, the Internet will be everywhere--from refrigerators to street lamps, from bar code scanners to office doors, and from © TeleGeography, Inc. 1999 TeleGeography 2000 Figure 7. Open Codes--May the Source Be With You engi Stallman set for himself was largely mitment to open standards (though unrealized. Enter Torvald, then a student at Helsinki University. In privately agreed--international stan- 1991 Torvald posted a new "kernel" for the OS on the Internet. A dards groups are usually dominated "kernel" is the heart of an operating system providing microcode for by carriers) so that one telephone net- shuttling instructions between the keyboard, the mouse, the moni- work can seamlessly hand-off calls to tot, the hard drive and microprocessors of a computer. But "a ker- another. But telcos and their switch nel does not run a computer by itself," as Chades Mann stressed in vendors have jealously guarded the an insightful review of Linux (The Atlantic Monthlg, "Living with custom software used to implement these standards. Linux August 1999). ’t works in concert with hundreds of sup- plementary programs inc uding drivers ut I ties pro~amming The Intemet grew up with a different engineering ethic: open stan- ’ ’ tools and window managers"--hundreds of megabytes of code. dards openly agreed to are the norm. And the line-by-line instruc- tions for encoding these standards--the source codes---are also In Mann’s words, Torvald’s invitation for help in transforming his posted on the lnternet. The lnternet therefore runs primarily on "kernel" into a viable operating system "was like a group of fdends open source software, decid[ing] to build their own space shuttle in a basement." Yet build it they did. Drawing on Torvald’s initial post, by 994 coders Sendmail is the most popular mail transfer agent with about 80 percent of the market. Apache, another open-source program, runs from around the wodd had cobbled together a complete operating system that soon became the choice of many computer enthusiasts at least 50 percent of the servers hosting Web content. The story ...... (see e.g. ’In The Beginning Was The Command Line by Nea] s similar for computer operating systems (OSesj Yainoo, one or ’ ’ Stephenson at www cryptonomicon com/beginn ng.html) the Web s most heavily trafficked s tes, uses FreeBSD. And fu y 30 percent of Internet hosts reportedly use Linux, another non-propri- Today, all 800 megabytes of Linux (some versions are much small- etary OS. er) can be downloaded from the Net for free or, if preferred, pur- These open source Internet products and the ethic which has made chased on CD-ROM from non-profits and loose collectives like them possible are likely to have a greater and greater impact on the Debian (price: $ t .99) or from commercial enterprises like Red Hat, telephone wodd for one simple reason: most popular services on which bundles value-added manuals with its distribution ($29.95~. the telephone network and the Internet will soon traverse tt-~ same computer switches and terminals. Some familiarity with the lnternet’s open source philosophy and its recent successes is thus which birthed Linux has puzzled many. How did it happen? One essential for navigating tomorrow’s net-centdc wodd. answer is provided by Eric Raymond in "The Cathedral and the Bazaar" {available at www.tuxedo.org/’esrtwdtings). Raymond, one of the hundreds of programmers who helped to code Linux, like a space shuttle, can be built only by a billion-dollar coml~any argues that open source has a psychologca edge "Play" is a under the watchful eyes of a master planner, not by a ga~gle of more "economically efficient mode of organizing work." There also ....... is a practica reason says Raymond Given enough eyes ’ geeKs swapping code omine. Hence, the success ot Linux seems to deL comn~n sense Le~ s take a t {scores of expert beta testers) a l bu~s are sha low Open ry c oser ooK. source culture will triumph not because cooperation is morally right Version 1.0 of Linux was released in March 1994 and is named for ... but simply because.., open-source communities.., can put orders the young Finnish pro~ammer, Linus Torvald, who contributed a of magnitude more skilled time into a problem." key batch of code. Linus did not invent Linux, which has its ori- gins in a non-proprietary OS proposed in t984 by Richard Staltman, known as GNU (GNU’s Not Unix]. To ensure that his Unix-like OS could never become a commercial product, Stallman invented the GNU ~eneral public license which copyrighted the source code and released it to the world on the sole condition that So what will the software world look like once the open source tran- sition is complete? According to Raymond, the basic infrastructure of computing for the Internet--the Web, operating systems and the lower level of communication software that has to cross boundaries between competing networks le.g.; IP messaging and signaling pro- tocols)--will almost all be open source cooperatively maintained any modification be covered by the same ucense--tnat is, could De ’ But higher-eve applications (e g., work group prosrams) wil prob- freely copied and changed. Stallman described this rule as "copy- ably remain closed. ]eft." © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 7. Open Codes--May the Source Be With You (continued) Ultimately, though, the line between open and closed source pro- say some observers, so goes its networked cousins from voice chat ~rams will be drawn in the marketplace by the give and take of to Intemet telephony and beyond. If this view is right, the open commercial practices. The recent battle between AOL and source movement may soon be followed as closely by the telephone Microsoft over Instant Messaging {IM) protocols--open source and industry as it is by software manufacturers. compatible, like e-mail or closed and company specific, like Windows 95--may be a precursor of what is to come. As goes IM, © TeleGeography, Inc. 1999 cars to soft drink machines. Taken together, in fact, many observers think that the traffic generated by these "silicon cockroaches," to use Robert Metcalfe’s term, will become the largest source of future network traffic. In theory, almost any device with a microprocessor can be linked to the Net, provided the device has an output port and the right software. Networked vending machines and Web cameras are already popular. However, as Kevin Werbach, Editor, Release 1.0, observed in an April 1999 survey of Intemet trends ("The Final Frontiers of Networking, Part 2: You are the Network"), virtually any company with a fleet of cars, trucks, field equipment or branch offices will benefit from tying their assets and people together over a network. That is one reason, suggests Werbach, why Sun Microsystems (among oth- ers) has devoted so much attention to its Java programming language and the related Jini technology which help to network different devices. According to Ken Arnold, a staff engineer at Sun, with the Jini software new devices (a pdnter, a micro- phone) can be quickly networked without having to pre-install the software at the device; the network does it. Says Arnold: "We want people to put little things on the network--tempera- ture sensors, motion detectors, clocks.., and then build more things out of them." The vertical reach of the network also will be extended by a new wave of business-to-business (B2B) commeme. Matthew Symonds highlighted this in a June 1999 survey for The Economist, "Business and the Internet". Once large firms move their purchasing online, writes Symonds, as have IBM, GE and Ford, for example, thousands of their business partners and suppliers must follow suit or risk being cut out of the value chain. The economic worth of these transactions will soon dwarf retail e-commerce, notes Symonds. And B2B e-com- merce is also likely to change the way people work across scores of fields which have almost nothing to do with such early Internet businesses as publishing and data processing. In Symonds’ words: "Because the main actors will be established rather than new businesses," the process "will [be] more like a highly telescoped evolution than a revolution. But revolutions come and go; evolution sticks." Nota bene. 111. Orienteering 101 "We are at an inflection point .... Steven A. Ballmer, CEO, Microsoft "Our model for conducting business and generating rev- enues is new and unproven .... MP3.com IPO Prospectus Now that we’ve looked at tomorrow’s network in more detail, it is again time to ask what it all means. What are the practical implications for the provision of international telecom services and regulation? Some further thoughts follow below, but 1 cau- tion that, as with the preceding narrative, the discussion remains at a fairly general level. Today’s networks are exceedingly complex and they vary signif- icantly country-by-country. Network evolution also is an uneven process, punctuated by extraordinary bursts of activity in a few locales, while whole regions go about their business as usual, at least temporarily. My perspective also has been influ- enced by my day-to-day experience with network providers in several industrialized countries. In much of the world though, making a phone call still requires a long walk and the lnternet is something you hear about on the radio. My observations thus must be applied with some care to different local circum- stances. A. Plotting a Course--Prices and Products When Microsoft’s CEO said that the software industry was at "an inflection point," he was referring to the challenge which Web-based software, available to any networked terminal, poses for Windows-based programs bought in shrink-wrapped boxes at shopping malls (see also note 6). But he might as well have been talking about the break point facing international telephone carriers. In the lnternet age, the old answers just don’t hold. Consider the fundamental question of what service a carrier will offer two years hence, between what points, and at what price. The answers used to be reasonably predictable. A telephone circuit was, with some exceptions (facsimile), used for voice telephony and the demand for services (and transmission capacity) on any route broadly tracked the size of the population or eco- nomic activity at each end. International transmission facilities TeleGeography 2000 © TeleGeography, Inc. 1999 Facilities Market Structure Technology Platform Service Model Traffic Routing Wholesale Pricing Retail Pricing Pre-1985 Half Circuits Monopoly/Oligopoly Circuit Switched Public Telephony Joint Venture/Correspondent Agreements Mutually Agreed by Carriers; Proportionate Return Accounting Rates Per Minute, Distance Sensitive 1985-2000 Whole Circuits Multiple Licenses Circuit Switched Public Telephony End-to-End Services, Global Backbones/Local Distributors Least-cost Routing by Originating Carrier Negotiated Rates; Domestic Interconnect Charges Per Minute, Less Distance Sensitive 2000 and beyond No Circuits Unlicensed Packet Switching (TCP/IP over ATM) Virtual Private NeWvorks (VPN)j Internet Best Efforts on Public InternetJ Quality of Service (QoS) Tiers Peering (SKA) and Packet- based Pricing Flat Rate, Distance Irrelevant were resource constrained and accordingly could be priced well above cost. The broadband, Internet-based networks described above have changed all of this. Services are user-defined not network- based, and telephony is but one of many networked traffic streams, and not the largest one. Places do not define traffic routing; servers and software do. In any case, many terminals have no fixed location--they are wireless. As well, when long- haul capacity is almost infinite, incremental transmission costs and prices approach zero. Though this Landscape still may be somewhere over the horizon even in richer countries, it can make the future seem unsettled everywhere. Let’s look more closely at some of the implications for interna- tional telecom prices and products. 1. Prices As broadband data networks carry the bulk of international traffic, the price per bit of delivering narrowband services, such as voice, must come into line with that charged per bit for car- tying data. After all, the same basic transmission facilities are used for both services. Likewise, when end-to-end bandwidth is plentiful, the cost for local and long distance transmission services also begins to converge. During the last year, as the proportion of data traffic on the major networks began to eclipse the voice stream[5], the pressure to align voice and data pricing has never been greater. Some examples suggest a magnitude of the economic challenge ahead. For instance, in 1998 the average gross international revenue per minute earned by AT&T for originating over I 0 billion min- utes of traffic was $0.68; net revenue per minute was $0.30. By comparison, rIXC, a major provider of Internet telephone service globally, generated gross revenues of $0.15 per minute during the first haft of 1999 and operated at a loss. Similarly, FCC data shows that in 1998 roughly 47 percent of the circuits used by U.S. carriers for overseas service carried voice traffic. These circuits, however, generated $9.4 billion in net revenue as compared to $908 million for the remaining (53 percent) of private line circuits which were primarily used for data services. There is more. Merrill Lynch reports that, in the U.S., the average monthly rev- enue per line generated by competitive local exchange carders (CLECs), many of which entered the market to offer broadband (xDSL) services to small business and residential customers, is approximately $55 per month. But only about 20 percent of this sum comes from data services, even though the average customer is known to spend more time (and bandwidth) each month surfing the Internet than making local and long distance telephone calls. Deutsche Telekom, Europe’s largest carder, generated only three percent of its revenues from data services in 1998, although its data portfolio included the country’s largest ATM network and a range of Internet access services. In [5] The datalvoice cross-over probably occurred in 1998 on the U.S. long distance network and on many international routes. By early 1999, for example, in Hong Kong, to cite but one case, the monthly total of dial-up Internet minutes was already three times that for international phone calls. Data traffic has eclipsed voice traffic first on longhaul networks because users generally are not directly connected to the Internet; they use the public telephone network for the last mile. With broadband local networks that will change, and the pdce pressure on vertically integrated telcos with local and long distance services will become that much greater. 22 © TeleGeography, Inc. 1999 TeleGeography 2000 the year to June 1999, Telstra Corporation, Australia’s largest carrier, reported that less than 14 percent of its communication revenues came from data services (broadly construed to include ISDN lines) and less than one percent from Internet ser- vices. Never mind that Telstra has more international trans- mission capacity dedicated to IP service than it does for tele- phony. You get the picture. The pressure to align the prices for voice and data traffic also exists at the wholesale level. During the last decade, many large carriers sought to reduce their settlement costs at the for- eign end by provisioning capacity on an end-to-end basis them- selves (that is, moving from a half circuit to a whole circuit regime; see Figure 8). With the rise of the lnternet, they must now cope with moving from a whole circuit to a "no circuit" regime under which traditional interconnection arrangements between long distance and local (foreign) carriers may be replaced by ad hoc "peering" arrangements between two or more tiers of Internet backbone operators. Only the top tier of operators (five or six carriers globally) exchange traffic without charge for destinations on the network. Other carriers general- ly must pay to have their traffic delivered on equivalent terms. To be sure, bandwidth costs for longhaul traffic are declining rapidly, but they are still far from trivial. Thus, as more and more subscribers see Internet-based telephony as an accept- able alternative, network service providers must somehow match the revenues they receive for connecting this traffic with their bandwidth and switching costs, plus, of course, the asso- ciated administrative marketing and billing expenses. It is far from certain how this will be done. The largest "pure" IP telephony carriers (ITXC, Net2Phone, Delta 3) are now operating at a considerable loss as facilities and marketing costs outstrip revenues (see the "VolP Traffic and Settlements" article in this book). Retail pricing trends in the U.S. and Western Europe for long distance services are not encouraging either. Low fiat rate charges seem to be the trend, and free (advertiser supported) service if you are an lnternet operator. In August 1998, for example, major U.S. carriers announced that for a $4 or $5 monthly fee they would offer customers .$0.05 per minute long distance telephone calls and overseas calls at .$0.12 per minute to many points. By con- trast, FCC data shows that in 1998 the gross domestic long dis- tance revenue of U.S. carriers was .$0.13 per minute; for inter- national services it was $0.59. In these circumstances, the path ahead may well be marked by a number of unexpected "profit warnings" and "restructuring charges," as providers scramble to align their business with this new economic reality. 2. Products Bits are bits. And as the premium for delivering voice bits between one country and another begins rapidly to decline, many international carriers will have to think hard about the business they want to be in years hence. This is particularly so in the most competitive markets, where the pressure on prices is likely to be the sharpest. One response is to offer a mix of different bit services to users, that is, to offer vertically integrated services by, for example, bundling international and local exchange services, where the entry barriers may be higher. Witness AT&T’s cable "IV strate- gy. Conversely, several major U.S. exchange carriers have sought to leverage their local assets with the aid of compatible international networks (the tie up of Frontier and Global Crossing is but one example). Carriers also may be tempted to bundle international telepho- ny with other digital services, such as mobile telephony where customers are still willing to pay a premium for convenience. Linking fixed and mobile serves also is plainly of interest to AT&T as well as to Bell Atlantic (which recently agreed to part- ner with Vodafone) and NTT (which has a large Internet and mobile base, but as yet, limited overseas reach). The mobile telephony business is now so large (mobile customers will exceed fixed line subscribers by 2005 in many countries) that it deserves a separate essay. Failing that, however, two points deserve emphasis. First, as revenue per customer from mobile telephony continues to fall, mobile data services and other Internet-based offerings will become as important to the future of the wireless industry to wireline telcos. Indeed, as more and more mobile operators deploy the Wireless Application Protocol (WAP), some estimate that data traffic will eclipse voice traffic on wireless networks by 2004. Second, whereas terrestrial mobile operators have his- torically been viewed as "local" carriers, multi-national mergers and alliances (e.g., Vodafone-Bell Atlantic; Orange- Mannesmann) coupled with the upsurge of roaming traffic (especially on Europe’s GSM networks) will give the largest mobile operators a significant international traffic base. (see the "International Mobile Telephone Traffic" article in this report for some preliminary data on international roaming traffic). The larger question for both wireless and wireline carriers, how- ever, is whether economies of scale (network lines) can actual- ly be translated into economies of scope (new services), and just how broad that scope might be. In the past, when services were network-specific (there was one network for voice, one for data and one for video), the ability of any network to leverage its position into the market space occupied by another network was much more limited. But once a network’s services have been digitized so that they can be networked with a common protocol (TCP/IP), that network’s economies of scale may more readily be translated into economies of scope elsewhere. This is most evident, of course, in the market for basic transmission services. Smaller carriers are already quite worried that the largest ATM backbone operators, for instance, can offer better end-to-end prices for high volumes of voice and Internet traffic and further concentration of backbone facilities (a la mergers of TeleGeography 2000 © TeleGeography, Inc. 1999 the AT&T and BT international networks or the proposed MCI WorldCom and Sprint merger) thus seems inevitable to many observers. Common digital transmission standards also may expand economies of scope at the service level too. Internet telepho- ny is a case in point. So too, notably, are streaming video and audio services. Telephone, lnternet and cable TV networks all hope to be able to bill their customers handsomely for access to a common suite of services and the software interface for these service suites has already led to a growing fight among competing software developers.J6] But bundling, like cross-market competition, has its risks. As lower value digital services are bundled with higher valued offerings, the greater may be the price pressure on the premi- um services. How long will users be willing to pay $0.05 per kilobyte of traffic for one service when they are offered an "all the bits you can eat" price for another service which is deliv- ered to the same set of terminals and over the same wires? The uncertain economics of tomorrow’s networks also is prompting a range of hybrid business models with as many dif- ferent products (see Figure 9). Some telephone companies see their future not so much in providing new network services as in hosting the services provided by others by, for instance, building data centers and Web server "farms." AT&T, UUNet, Frontier, Owest, Metromedia (which recently acquired AboveNet) and Exodus (a major IP backbone provider) all have plans in this area. Major "outsiders," such as SAI:~, Sibel, Hewlett-Packard and lntel, which plans to spend more than $1 billion on 12 worldwide data centers, have entered this market as well. A new umbrella organization, the application service providers industry consortium (www.aspconsortium.org), has been created by over 100 companies to agree upon the best practice for hosting high-end applications. The challenge of transforming a telco into a comco (communi- cations company) was also a prime topic at the Telecom ’99 Forum in Geneva. Drawing on a background paper by Ernst & Young LLP (see the executive summary of "The Connected Society," is available at www.ey.com/publicate/tce/pdf/connect- edsociety.pdf}, executives from several international telcos addressed the challenge of moving up the value chain as pure transmission becomes a commodity service. Press releases aside, the plans of most telcos are still embryonic, but there are some success stories. One was recounted by Peter Nicholson, Chief Strategy Officer, BCE, parent of Bell Canada. To jump start its systems integration business, BCE acquired a 45 per- cent stake in an independent venture, CGI, and then transferred key staff and contracts to the company, thus giving it the com- mercial scale and expertise to play in the larger North American market. Likewise, BCE increased the market capitalization (and visibility) of its e-commerce business, Emergis, by more than S 1 billion by separately listing the company’s stock while retaining a majority interest. Whatever the new strategy--e-business or Web hosting, sys- tems integration or application management--it is now clear that the core business of most international telcos has begun to enter a period of dramatic change. Whether one canvasses global consultants like Ernst & Young LLP and Pricewaterhouse Coopers (the source for Figure 9) or the telcos who hire them, there is now little doubt that the sustainable product set of most carriers 5 or 10 years hence will look quite different from what it does today. B. Rules of the Realm--Regulating Network Access and More Since the early 1980s, international telecom regulators have been preoccupied with market entry and prices, especially set- tlement rates. These are late 20th century concerns, however. The landmark 1997 WTO’s Agreement on Basic Telecommunications Services, which opened the telephone ser- vice market in over 70 countries to foreign investment, resolved the main market entry issues, at least in principle. Likewise, new fiber optic cables and the Internet will soon bring both wholesale and retail prices for most international calls more into line with costs. Competitive data networks will be the name of the game worldwide. Some regulators may try to prep up the common carrier para- digm for several more years or to manage competition, but it will be a stop-gap effort. In the main, regulators will find them- selves concentrating on a new agenda brought forward by the broadband, Internet-friendly networks described above. Topping this new agenda are likely to be the following: I. Promoting Network Access If the "last mile" is the gateway for new broadband services, then who may pass through this portal and on what terms becomes the issue of the day. Witness the widening legal and regulatory disputes in the U.S. triggered by AT&T’s proposed acquisition of cable TV operators, TCI and MediaOne. Witness too the disputes in Europe and North America regarding the terms on which incumbent telephone companies must resell to their competitors local loops capable of providing xDSL and other high-speed services. Encouraging broadband competition will require a new sensitiv- ity, however, to the changing relationship between network owners and service suppliers. In the telephone world, these businesses are one and the same; in the Internet world (and also the cable TV universe) they typically are not. So where should regulators intervene~at the network level, e.g., with rules mandating resale and unbundling of basic elements? Or at the services level, e.g., mandating open source standards for cable TV set-top boxes so that independent producers and TeleGeo{~raphy 2000 © TeleGeography, Inc. 1999 Figure 9. From Telco to Comco IT/e-business consulting Web hosting Network design and integration Tomorrow’s business focus Today’s business focus Rather than simply delivering network services, tomorrow’s comcos will be managing information and communications solutions Source: Adapted from Pricewaterhouse Coopers, InfoComm Review, voL 5, no. 3. (1999). Design © TeleGeography, Inc, 1999 unaffiliated Intemet service providers will have non-discrimina- tory access? Or both? Laws can codify access, but so can soft- ware standards and (seemingly neutral) access technologies (see Figure 10). As importantly, who should police the new regime--the regula- tor, the regulated or the market? When markets were closed, the pace at which regulators moved and the remedies for inac- tion typically affected but one company. Not so today when there are scores of competitors in key markets and business moves on Internet time. As Kathryn C. Brown, the FCC’s Chief of Staff, recently said, governments must learn how to enforce "real world rules in real time for real businesses.., something which is a real change for many regulators." That is why the FCC, among others, is encouraging warring par- ties (e.g., on interconnection and local access charge issues) to reach a negotiated agreement with self-executing enforcement mechanisms, including performance audits, liquidated damages and specific financial penalties for non-performance. These types of private contracts, publicly reviewed, are likely to be more and more popular in tomorrow’s deregulated markets. Just in case, though, the FCC has asked the U.S. Congress to create a new Enforcement Bureau equal in stature to other FCC divisions. 2. Allocating Common Resources The drive for broadband connectivity will place a new premium on networking resources that are comparatively scarce, and thus must be shared, notably the radio spectrum and network numbers. Competing demands for radio spectrum are everywhere. Mobile carriers are bidding for G-3 licenses; other entrepre- neurs want large blocks of microwave spectrum for fixed wire- less networks; and satellite companies want frequencies to sup- port their next-generation broadband systems. Spectrum auc- tions (including "after market" trading forums) as well as flexi- ble, non-service-specific spectrum licenses, may put spectrum to its best use domestically. Treaty negotiations will probably remain the rule internationally for some time, however, with the ITU’s radio conferences being perhaps the most important. So long as a common scheme for numbering telephone termi- nals and computer-based resources (the lnternet) are benefi- cial, governments will probably have a role here as well. Telephone number plans, local number portability (in markets with competitive carders) and access to "free phone" numbers have garnered the most attention thus far. But management of IP addresses and, as importantly, the rules for PSTN-IP number convertibility, are likely to become more pressing. ICANN’s recent history underscores this. (see the lnternet section for more details). 3. Monitoring Network Performance When the world’s telephone network was largely owned by a small number of carders, it was comparatively easy for regula- tors to set and verify network performance standards (call- seizure rates, set-up times, etc.). Because most engineers shared a common outlook, nationality aside, international stan- dard setting was also a clubby affair. Today, however, in many countries a long distance call may depend on the performance of several competing carriers; on the Internet, it is not uncom- [6] As the world’s largest software company, Microsoft, of course, is a key protagonist. It has recently spent over $8 billion buying minority stakes in mobile telecom, wireline and cable TV businesses (see the Carders section for more details) to advance its own soft network vision. As Microsoft Chairman, William Gates, recently explained to a Telecom ’99 audience in Geneva, the company’s growing interest in network service providers is "very similar to the collaboration that existed between the hardware industry (e.g., lntel and Compaq) and Microsoft in creating the PC business." By providing a symmetric suite of network interfaces (for mobile devices, cable TV, web TV and PCs) and by underwriting the cost of new broadband networks, Gates argues that Microsoft will stimulate new network-based services for every- one. However, competing software providers such as Sybian (backed by Motorola and Nokia), which is promoting its own interface for wireless data, contend that Micresoft’s networking push is just a reprise of its past business practices. Once again, they say, the software giant is trying to extend its proprietary codes (DOS and, later, the Windows NT Operating Systems) into new markets by making key network interfaces into technical choke points. Open standards, not private ones, are the best way to grow the market for new networked services, say Microsoft’s opponents (see also Figure 7). 25 © TeleGeography, Inc. 1999 TeleGeography 2000 the incumbents. But how could With laws? Court orders? Policy state- a technofogy-neutra] definition be enforced?, merits. In the d~ital age, these tools may "it does not take much imagination to see that if there is Ia new] fee for be less and less effective. Network switch- the use of IP telephony," says Scott Bradner~ a member of the Intemet es don’t understand words~ but they follow Engineering Task Force (IETF), "people would quickly start using non- software may do more assigned [switch] port(s)" on routers to mask the identity of the traffic ..... routed by their chosen tntemet Service Provider (ISP]. Given the flexibili- edicts.Sageprivate than any number or ~eg slat ve ty of ]nternet routing protocols, "there would be no way for an LSP to even know that it was handling IP telephony." As Lawrence Lessig puts it: "Code is law." "The coming of age of this sim- ple idea," says Lessig, a Harvard Law Schoot professor, "is the singIe most But here is the rub. If code is law, then the ad hoc application of code-- significant ch; for example, to tweak the switch ports used by Intemet telephony or the pages--can create pdvate laws. Hence as Lessig There are few texts, as telecom law. But a new book by says, East Coast and West Coast Code raise similar questions about proce- Lessig, Code ond Other Lows of C~lberspace (Basic Books, t999], offers a dures and values. Is the code wdtten in the open? ls it transparent about good starting point. (For a shorter version, try Lessig’s columns in The its means? How is it implemented?. One of the book’s main arguments is that the architecture of any electron. Coast Code--to the deliberations of the W3C [Worldwide Web Consortium) ic network--the hardware and software that makes it run--contains a set and |ETF. for example--as to the deIiberations that underlie East Coast of embedded rules which govern the behavior of users. Lessig calls these Code. "If the architecture of the Internet affects values as well as bits... electronic rules--typically minted in Silicon ~alley or Redmond--West Coast then as in any lawmaking context, we should be asking who are these law- Code, in distinction to the text-based-rules--East Coast Code--popular in makers and how do they make law?." Washington, D.C, tn Lessig’s view, West Coast Code simply works better: The ability of West Coast Code to answer these questions has been publicly "it’s faster, cheaper and more reliable tested during the last year by the efforts of the quasi-public lnternet Some examples may help to cladfy the choice. Consider the issue of limit- Corporation for Assigned Names and Numbers (ICANN) to change the rules ing access to pornography on the lnternet. East Coast Code typically relies for managing lntemet addresses (background on ICANN is provided in upon new criminal penalties and ~nancia] forfeitures for those who post or "lnternet Governance: Then and Now" later in this book). For example, distribute offensive bits. tn contrast, West Coast Code prefers to fight bits ICANN has recently proposed to make it harder for "cybersqoat-ters" to use with bits. using software (e.g., P3R a version of XMI. (Extensible Markup an lnternet address which matches a pre-existing corporate trademark or Language)--See Figure 4) to "tag" the content on Web pages in a standard brand. If tCANN’s mandatory arbitration approach prevaiIs, many cyber- format so that pages can be "rated" and Web browsers can fitter any con- squatters could lose their "h~me." For unless the Internet’s master direc- tent a user wishes to block. The first approach relies upon government to tory (the root servers) encode your address, no one can reach you. police the Net: the second seeks to harness the pov~r of the Net’s users. [CANN’s actvties have prompted considerable protest, not so much Privacy regulation is another example. Commercial Web sites often collect because the group’s approach to cybersquatters is misguided, but because significant persona] data from their visitors. In response, some govern- many observers fear that fCANN’s actions may lead others to use the merits want to require merchants to disclose their data collection practices intemet’s address codes to police undesirable behavior. "Domain names and to obtain customer consent. West Coast coders, Eke Lesslg, say this may provide the most efficient way of enforcing laws on the Net," corn- won’t work. "Rather than read a privacy policy each time I enter a new mented Mike France in Business Week ("What’s in a Name corn? P]enty"] site, 1 should be able to tel! my browser what my pdvacy preferences are," For example, "when people buy names for their Web sites, they could be that is, to set parameters for s~e access. Then my software can "negotiate" required to sign a detailed contract obligating them to comply with a car- with their softwaro. Codes, not words, wifi best protect privacy, says Lessig. rain set of rules," such as those governing obscene content or the protec- advantage because Federal Communications Commission (FCC) rules alas- The outcome of the ICANN debate may thus have much wider implications sify the service as "value added," and thus exempt from the payment of about the future of regulation in a society where the codes used online may local access charges to support universal service. The FCC’s rules should shape economic and social options even more than the bills passed by leg- isiatares. -~D TeleGeography, Inc~ ~999 © TeleGeography, Inc. 1999 TeleGeography 2000 mon to traverse 15 or 20 networks en reute. In this environ- ment, what QoS obligations should network service providers have to each other?. To their customers? Should QoS be gov- ernment-mandated or market-driven? Likewise, how do you assign the liability amongst various ser- vice providers (or their customers) for network failures whether due to negligence or not? Should the limited liability rule applicable to North American carriers be extended to the lnternet? Or should the contract law rules which now govern most data networks apply to all service providers? And what role, if any, should carrier insurance play--mandated or option- al? Recent network outages and fast-spreading software viruses have brought these questions forward, as has the Y2K issue. Tomorrow’s soft networks will make any network player even more vulnerable to actions of downstream and upstream ser- vice providers (as well as their customers). That has made net- work security of prime concern to every major service provider. Security is also a key reason why private intranets, with sophis- ticated fire wails, are attractive to companies with mission-crit- ical communication needs. Thus, with the economic costs of network outages rising daily and the uncertain remedies avail- able to injured parties, the pressure for government action is likely to mount. 4. Protecting Consumers and Their Privacy Open markets also have attracted their fair share of unscrupu- lous operators and almost all regulators have seen a rising level of consumer complaints. The lnternet likewise has been fertile new territory for deceptive business practices and worse, many of them multi-national in scope. Getting fleeced online is one thing; letting the whole world know about it is quite another and many lnternet users seem to be less troubled by the threat of "shady" Internet operators than by the inadvertent surren- der of their online privacy--privacy which they take for granted on the public telephone network. Numerous countries have laws protecting both the content of a telephone conversation from being intercepted as well as the identity of the originating and terminating numbers. Not so for Internet sessions. And the trans-national dimension of the Internet can make the matter even more complicated. There is also a flip side to this issue, of course--protecting children and non-consenting adults from unwanted invasions of their net- work space, whether from unsolicited banner adverts or porno- graphic images. The wider the network’s reach and the greater the range of services on offer, the more pressing these issues are likely to become. Which brings us to the last issue. 5. Reconciling Telecom and Broadcasting Rules When all networks can carry the same types of bits, why should some networks be licensed, or be subject to price or content rules, and others not? Tomorrow’s broadband networks will allow video, audio and telephone services to be delivered side by side. Internet telephony and streaming video services on the Net have already challenged traditional service definitions for regulators. However, in North America, at least for now, regulators have preferred to treat the Internet as sui generis and not saddle it with the market access or content regulation commonly applied to terrestrial broadcasters or telephone companies. This "hands off" policy has not always been even- ly applied; interconnection terms between Intemet service providers and incumbent telcos are subject to complex rules, at least at the local level. Convergence between wireless and wireline services is also forc- ing regulators to rethink their rule books. When a mobile oper- ator uses its spectrum to provide Internet services to a fixed site, how should it be treated? And who should pay? And what about competition policy: do wireless local loops or the poten- tial for 3G services make the current telcolcable TV duopoly a temporary feature of the telecom landscape in some countries or not? As always, new telecom technologie.s raise more questions than answers for those who would oversee them. In the short run, therefore, despite the clamor for action, the best approach may be "forbearance": do not apply old rules to new services and do not adopt any new rules either unless you have a clear sense of the networks’ current direction. Which brings us back to this essay’s title--The Soft Network. Though it is woven from silica, tomorrow’s network will be soft- er than ever. In the years ahead, it is lines of codes, not miles of conduits, which are likely to exert the strongest magnetic pull on the network’s compass. In this shifting terrain, "know the code," however written, is likely to be the first rule of naviga- tion for referees and players alike. ~i~.~ Gregory C. Staple (gstaple@velaw.com) is the founder of TeleGeography and a member of the telecommunications prac- tice group at the law firm of Vinson & Elkins L.L.P., Washington, TeleGeography 2000 © TeleGeography, Inc. 1999 For Further Reading Know the Code For background on the cross-cutting impact of open source pro- grams, such as Linux, see "The Open Source Revolution," edit- ed by Tim O’Reilly, Release 1.0, 20 November 1998, available from www.edventure.com and Open Sources, Voices from the Open Source Revolution, Chris di Bona, Sam Ockman and Mark Stone, eds. (O’Reilly & Associates, Inc., 1999). The XML revolution is profiled by Tim Bray, co-editor of the early specifications, in "XML and the Second Generation Web," Scientific American, May 1999, available at: www.sciam.com/1999/0599issue/0599bosak.html. See also the special report on XML in the October 1999, Network Magazine, available at www.networkmagazine.com. And if you wish to follow XML events regularly, visit vwvw.xml.com. Last, but far from least, is the new history by the man who invented the "http" in web addresses, Tim Berners-Lee: Weaving the Web, Harper, San Francisco, 1999. Broadband Local Access The roll-out of broadband capacity over the last mile in the U.S. is summarized in a recent (October 1999) report by the FCC Cable Services Bureau, titled "Broadband Today." See www.fcc.gov/Bureaus/Cable/News_Releases/1999/nrcbg017.ht ml. For a global survey, see Pioneer Consulting LLC’s Global Broadband Access Markets. See also the sources referenced at the Web page maintained by The ADSL Forum; www.adsl.com/adsl_forum.html. Prices and Products As international telcos begin to rethink their pricing and service strategies in earnest, they would do well to consult the work of two economists who have spent much of the last decade think- ing through the economics of network-based services. See Information Rules: A Strategic Niche to the Network Economy by Carl Shapiro and Ha] R. Varian (Harvard Business School Press, 1999). The Future of Regulation For the most complete global survey available, see Trends in Telecommunication Reform, Convergence in Regulation, 1999 (ITU, Geneva 1999). Adso worth reviewing is the U.S. Federal Communication Commission’s "Strategic Plan for the 21st Century," at: www.fcc.gov/Bureaus/Miscellaneous/News_ Releases/1999/nrmc9059.html. For Europe’s regulatory agen- da, see the European Commission’s "Fifth Report on the Implementation of the Telecommunications Regulatory Package," at: europa.eu.int/comm/dg13/Sthrep99.htm. However, the book of the year on regulation is Code and Other Laws of Cyberspace, by Lawrence Lessig (Basic Books, 1999). Runner up: Judge Thomas Penfield Jackson’s findings of fact in the Microsoft case available at: usvms.gpo.gov. Carriers TeleGeography 2000 © TeleGeography, Inc. 1999 The Growth of International Services Competition Forty Percent Growth in Competition--Again As of July 1999, over 1,700 companies worldwide were authorized to build facilities to offer international telephone service. Three years before, there were less than 500 (see Figures 1 and 2). In the fastest growing markets (the U.S. and Western Europe), the pace is not likely to slow down. Even by conservative estimates, the world should easily have more than 2,200 international carriers by mid-year 2000. Although these carriers are all authorized to operate inter- national transmission facilities, many of the new competitors do not own extensive submarine cable capacity and switch- ing assets. In the U.S. especially, the hundreds of small companies which are authorized to own networks may never build them. In fact, three quarters of the carriers authorized in the U.S. do not generate more than $10 million in inter- national service revenues each year. On a global scale, a pattern is emerging: the world’s top 20 carriers carry about 60 percent of the traffic whereas the thousands of new car- tiers which have started business since 1989 carry just 13 percent. The relationship between the network builders and the swarm of "virtual carriers"--which repackage the facilities and services of the handful of network builders--is one of symbiosis. New market entrants, while they represent a competitive threat, can also be the incumbent’s best cus- tomers. The New Breed of Virtual Carrier Although old-fashioned pure resale is still a staple (more than 30 percent of traffic originated in the U.S. was resold in 1998), virtual carriers are always looking for new ways to cut prices without shrinking profit margins. The next devel- opment in alternative traffic routing may create a new kind of packet-switched symbiosis. Once the network builders figure out how to send commercial grade traffic on IP net- works reliably and also to devise a way to settle accounts properly, the ranks of international carriers will swell even more rapidly (see the "VolP Traffic and Settlements" article in this book). Indeed, IP connectivity may lead to unregu- lated "international carriers" on virtually every street corner, in every corner of the world. ~i~ iiiFigureli: iThelilnternationalii! )om iiiiii!i Global Growth of International Carriers, July 1995 July 1999 2000 1800 1600 1400 1200 1000 8O0 600 4O0 200 0 367 470 1042 1760 1995 1996 1997 1998 1999 Regional Growth of Authorized international Carriers, 1996-1999 Americas Europe Asia-Pacific Africa ¯ July 1999 [] July 1998 July 1997 July 1996 July 1995 0 200 400 600 Number of Competitors 800 © TeleGeography, Inc. 1999 TeleGeography 2000 Number of Authorized International Carriers Rank Country July 1999 July 1998 July 1997 July 1996 July 1995 1. United States 679 393 175 115 65 2. United Kingdom 215 144 100 65 35 3. Hong Kong 80 4 1 1 1 4. France 50 29 1 1 1 5. Japan 50 13 3 3 3 6. Canada 49 21 21 19 18 7. Germany 40 32 1 1 1 8. Switzerland 40 21 1 1 1 9. Russia* 35 1 1 1 1 10. Netherlands 30 23 3 1 1 11. Australia 28 14 10 8 8 12. Ireland 25 5 3 3 1 13. Korea, South 24 3 2 2 2 14. New Zealand 19 11 9 9 2 15. Belarus 18 1 1 1 1 16. Belgium 18 11 1 1 1 17. Denmark 18 11 9 7 1 18. Peru 18 1 1 1 1 19. Austria 17 13 1 1 1 20. Mexico 16 15 9 1 1 21. Spain 16 9 1 1 1 22. Sweden 16 13 11 9 7 23. Italy 15 9 1 1 1 24. Norway 14 7 1 1 1 25. Philippines 12 12 9 9 9 26. Chile 10 9 9 9 9 27. El Salvador 10 11) 1 1 1 28. Finland 8 8 8 8 5 29. Malaysia 5 5 5 5 4 31). Luxembourg 4 1 1 1 1 31. Colombia 3 3 1 1 1 32. Dominican Rep. 3 3 3 3 3 33. Iceland 3 1 1 1 1 34. Israel 3 3 3 1 1 35. Kazakhstan 3 3 1 1 1 36. Argentina 2 1 1 1 1 37. Bermuda 2 2 2 2 1 38. Brazil 2 1 1 1 1 39. Brunei 2 2 2 2 2 40. China 2 2 2 2 2 41. Dominica 2 1 1 1 1 42. Georgia 2 1 1 1 1 43. Indonesia 2 2 2 2 2 44. Ukraine 2 2 2 2 2 Estimates include ca ~riers authorized TeleGeography 2000 © TeleGeography, Inc. 1999 Market Shares of International Carriers Percentage of Outgoing Minutes Country/Carrier 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Australia Telstra 100.0 98.0 87.0 76.3 73.4 62.0 55.0 49.0 C&W 0ptus 2.0 13.0 21.9 23.4 27.0 26.0 22.0 AAPT 11.0 12.0 Primus Telecom. 3.0 4.0 WorldxChange 4.0 Global One 3.0 Others 1,8 3.2 11,0 5.0 6.0 Austria Telekom Austria 100.0 95.0 UTA Telekom 1.5 Others 3.5 Belgium Belgacom 100.0 96.0 Others 4.0 Canada (Canada-U.S. route only) Stentor (alliance) 100.0 93,0 80.0 63.0 57.0 56.0 55.0 Sprint Canada 22.0 27.0 23.0 23.0 AT&T Canada Long Distance 2.0 8.0 8.0 8.0 9.0 13.0 ACC 3.0 4.0 5.0 5.0 Others 5.0 12.0 4.0 4.0 7.0 4.0 Chile CTC Mundo/Globus 1.0 17.5 31.2 31.0 31.5 33.0 35.0 Entel Chile 100.0 80.0 57.5 40.0 40.6 37.3 33.0 31.0 Chilesat 20.0 25.0 19.7 19.4 15.2 17.0 13.0 BellSouth Chile 6.6 6.8 10,0 10.0 10.0 FirstCom (lusatel) 1.2 1.0 2.8 3.0 5.0 Transam 1.0 1.0 2.8 3.0 3.0 Others 1.0 1.0 1.0 1.0 3.0 Denmark Tele Danmark 100.0 92.5 84.4 67.5 Tele2 4.0 6.6 12.4 Telia A/S 3.5 6.3 9.9 Others 2.7 10.3 Dominican Republic Codetel 100.0 >90.0 85.8 83.0 77.0 73.8 72.2 Tricom n.a. 6.7 7.5 12.8 12.9 15.5 AACR n.a. 7.5 9.5 10.2 13.3 12.3 El Salvador ANTEL 100.0 91.5 Others 8.5 Notes D ata based on outgoing international traffic for the public switched neWvork and International Simple Resale (ISR) covering the full calendar or fiscal year. Some data aggre- gated in "0thers" rows include market shares for carriers shown individually in later years. Market shares may not total to 100 percent due to rounding. Canada: Competition on non-U.S, routes commenced October 1998. Stentor alliance members include Bell Canada, BCT.Telus, SaskTel, MTS, and Aliant. By year end 1998, incumbent carrier Teleglobe held an approximate 80 percent market share on overseas routes. Sprint Canada market shares include Fonorola prior to 1998 merger. ACC merged with AT&T Canada in 1999. Source for some carrier market shares is NBI/Michael Sone Associates Inc., Toronto, Canada. Chile: CTC Mundo/Globus market shares prior to 1998 merger aggregate CTC Mundo and Globus traffic. Source; TeleGeography, Inc. © TeleGeography, Inc. 1999 © TeleGeographyo Inc. 1999 TeleGeography 2000 Country/Carrier Finland Sonera Finnet International Telia Others France France T~l~com "releglobe Cegetel Siris Others Germany Deutsche Telekom Mannesmann Teleglobe MCI WorldCom Viag Interkom Others Hong Kong C&W Hong Kong Telecom New World Others Indonesia PT Indosat PT Satelindo Ireland Eircom Esat Telecommunications MCI WorldCom Others Israel Bezeq Barak Golden Lines Italy Telecom Italia Infostrada Teleglobe Others 1989 1990 Percentage of Outgoing Minutes 1991 1992 1993 1994 1995 1~6 1~7 19~ 100.0 90.0 72.8 66.0 58.9 54.7 5.0 19.1 ~.2 28.2 28.0 3.0 7.7 8.8 9.3 12.0 2.0 0.4 0.9 3.5 5.2 100.0 99.5 0.5 95.4 4.6 88.5 11.5 100.0 100.0 100.0 100.0 100.0 84.8 15.2 91.0 5.0 3.0 1.0 72.5 15.0 12.5 100.0 93.0 3.6 1.9 1.0 1.0 85.7 7.3 1.8 1.8 1.5 1.8 90.0 2.0 8.0 88.3 11.7 78.0 8.0 3.0 11.0 55.0 22.5 22.5 88.6 4.5 3.8 3.0 Notes Data based on outgoing international traffic for the public switched network and International Simple Resale (ISR) covering the full calendar or fiscal year. Some data aggre- gated in "Others" rows include market shares for carriers shown individually in later years. Market shares may not total to 100 percent due to rounding. Germany: Mannesmann market shares include international traffic carried by mobile operator Mannesmann Mobilfunk and the long distance carriers Mannesmann ARCOR and o.tel.o. Source: TeleGeography, Inc. © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 Market Shares of International Carriers (continued) Percentage of Outgoing Minutes Country/Carrier 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Japan KDD 93.3 88,0 73.3 69.7 66.9 66.3 66.2 63.9 62.7 58.0 Japan Telecom 3.0 5.5 13.4 15.0 16.2 16.4 16.5 17.5 19.0 18.3 C&W IDC 3.7 6.5 13.3 15.3 16,9 17.3 17.3 18.7 18.4 18.2 Others 4.2 Malaysia Telekom Malaysia 100.0 90.0 80.0 77.0 Celcom (TRI) 8.0 11.0 10.0 Binariang 8.0 TIME Telekom 5.0 Others 2.0 9.0 1,0 Mexico Telmex 100.0 83.0 78.0 Alestra 8.5 10.5 Avantel 7.5 8.5 Protel 1.0 Miditel 1.0 Others 1.0 2.0 Netherlands P~I Telecom (KPN) 100.0 95.0 89.0 Ene~Lel 2.0 2.0 Others 3.0 9.0 New Zealand TNZ 100.0 92.0 82.0 80.0 78.4 74.8 78.0 78.2 74.6 76.0 CLEAR 8.0 18.0 20.0 21.6 25.2 22.0 19.8 20.2 13.0 Teleglobe 7.0 Others 2.0 5.2 4.0 Norway Telenor 100.0 93.5 Telia 5.0 Others 1.5 Philippines PLDT 100.0 91.6 84.2 69.0 68.0 78,0 71.0 70.0 Globe Telecom 2.0 7.0 7.1 Eastern Telecom 7.0 6.0 5.0 7.0 6.8 Bayan Tel 1.0 4.0 5.0 5.1 Digitel 2.0 3.0 3.7 Capitol Wireless 1.0 1.0 1.0 1.0 3.1 Sma~ 1.0 1.0 2.0 2.5 Philippine Global Com. 8.4 15.8 23.0 23.0 6.0 3.0 1.0 Islacom 1.0 1.0 1.0 Notes Data based on outgoing international traffic for the public switched neWvork and International Simple Resale (ISR) covering the full calendar or fiscal year. Some data aggre- gated in =Others" rows include market shares for carriers shown individually in later years. Market shares may not total to 100 percent due to rounding. Japan: Japan Telecom market shares include ITJ prior to 1997 merger. Source: TeleGeography, Inc. © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Percentage of Outgoing Minutes Country/Carrier 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Republic of Korea Korea Telecom 100.0 79.9 74.5 68.7 72.6 73.5 69.0 66,6 Dacom 20.1 25.5 31.3 27.4 26.5 27.0 21.9 Onse 4.0 11.5 Spain Telef6nica 100.0 95.0 Retevision 4.5 Others 1.0 Sweden Telia AB 100.0 92.0 87.0 78.0 69.0 66,0 62.0 Tele-2 8.0 13.0 21.0 22.0 22.0 24.0 Others 3.0 9.0 12.0 14.0 Switzerland Swisscom 100.0 93.5 Sunrise 3.0 Diax 0.7 Others 2,8 United Kingdom BT 91.0 86.0 81.0 76.8 74.2 68.6 67.7 60.0 54.9 51.6 C&W Comm. 9.0 14.0 19.0 23.2 24.0 28.1 25.8 26.8 30.3 32,2 MCI WorldCom 6.6 5.1 5,1 Teleglobe 4.2 ACC 3.0 3.6 3.0 GlobalOne 3.1 1.5 1.5 Others 1.8 3.3 6.5 1.0 4.6 2.2 United States AT&T 83.3 78.4 74.8 70.3 62.2 60.1 54.3 50.2 44.7 39.6 MCI WorldCom 10.2 14.6 17.8 21.2 25.4 28.6 32.0 32.9 31.2 28.8 Sprint 5.8 6.4 6.3 7.3 10.3 11.1 11.3 13.2 12.0 11.7 Teleglobe USA 1.3 3.3 Pacific Gateway Exch. 0.9 3.2 2.6 WoddxChange 1.9 2.4 STAR Telecomm. 0.5 1.8 Others 0.7 0.7 1.1 1.2 2.1 0.2 2.4 2.8 5.2 9.8 Notes Data based on outgoing international traffic for the public switched network and International Simple Resale (ISR) covering the full calendar or fiscal year. Some data aggre- gated in =Others" rows include market shares for carriers shown individually in later years. Market shares may not total to 100 percent due to rounding. United Kingdom: The figures for Cable & Wireless Communications reflect data for Mercury prior to its April 1997 merger with Bell Cablemedia, Videotron, and NYNEX CableComms. United States: Market shares for U.S. carriers prior to 1993 exclude traffic to Canada and Mexico. MCI WorldCom market shares prior to 1999 merger aggregate MCI and WorldCom traffic. Source: TeleGeography, Inc. © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 The Top 40 International Carriers Outgoing Traffic 1998 Revenue (millions of minutes) (US$ billions) Rank Company Origin Country 1998 1997 Change 97-98 Total Int’l Service 1 AT&T (a) U.S. 9,878.7 10,290.0 -4% $53.2 $4.7 2 MCI WorldCom (a) U.S. 7,188.5 7,307.7 -2% $30.4 $2.7 3 Deutsche Telekom (a) Germany 4,711.0 4,813.0 -2% $37.7 $2.2 4 British Telecom (a, b) U.K. 4,249.3 3,735.0 14% $26.3 $2.6 5 France Telecom France 3,825.0 3,545.0 8% $25.9 $2.0 6 Sprint (a) U.S. 2,926.0 2,758.9 6% $17.1 $0.7 7 C&W Communications (a, b) U.K. 2,646.2 2,065.0 28% $4.3 $1.1 (est.) 8 Telecom Italia Italy 2,339.4 2,352.0 -1% $25.7 $1.9 (est.) 9 Swisscom Switzerland 2,258.0 2o164.0 4% $6.9 $1.2 (est.) 10 China Telecom China 1,711.5 1,631.8 5% $23.9 $2.2 11 C&W Hong Kong (a, b) Hong Kong 1,681.6 1,718.0 -2% $4.5 $2.2 12 KPN Telecom (a, c) Netherlands 1,600.0 1,535.0 4% $8.5 $0.9 13 Telef6nica (a) Spain 1,518.0 1,319.0 15% $19.6 $0.8 14 Belgacom (a, c) Belgium 1,400.0 1,340.0 4% $4.4 $0.5 (est.) 15 Bell Canada (c) Canada 1,350.0 1,250.0 8% $6.9 $0.4 16 Singapore Telecom (a, b) Singapore 1,235.0 1,161.0 6% $2.9 $1.1 17 Teleglobe Canada (a) Canada 1,145.0 1,113.5 3% $3.4 $0.8 18 Telekom Austria (a) Austria 1,100.0 995.5 10% $3.6 $0.3 (est.) 19 KDD (a) Japan 1,100.0 1,105.0 0% $3.3 $1.6 20 Rostelecom (a) __ Russia 1,038.3 939.0 11% $0.9 $0.2 21 Tel~fonos de M~xico (a) M~xico 1,022.8 1,008.9 1% $7.9 $0.9 22 Saudi Telecom Saudi Arabia 932.6 801.3 16% $3.3 $0.2 (est.) 23 Etisalat U.A.E. 874.8 738.0 19% $1.4 $0.2 (est.) 24 Chunghwa Telecom Taiwan 862.0 789.0 9% $6.0 (est.) $0.6 (est.) 25 Telstra (b) Australia 836.0 833.0 0% $11.9 $0.9 26 Teleglobe USA (a) U.S. 830.3 301.6 175% $3.4 $0.1 27 Telia Sweden 750.0 747.0 0% $6.2 $0.6 (est.) 28 OTE Greece 681.3 593.7 15% $3.0 $0.6 29 FaciliCom (a) U.S. 678.0 223.2 204% $184.2 $0.1 30 Turk Telekom Turkey 644.1 557.5 16% $1.9 (est.) $0.2 (est.) 31 Pacific Gateway Exch. (a) U.S. 641.4 742.9 -14% $0.5 $0.0 32 Eircom (a, b) Ireland 613.0 534.0 15% $2.4 $0.2 33 Korea Telecom Korea, Rep. 604.3 610.0 -1% $9.1 $0.7 34 TPSA Poland 602.4 529.0 14% $2.5 $0.3 (est.) 35 WorldxChange (a) U.S. 595.8 430.1 39% $0.4 $0.04 36 Embratel (a) Brazil 545.8 459.1 19% $3.4 $0.7 37 Telekom Malaysia (a, b) Malaysia 525.5 588.5 -11% $2.0 $0.3 (est.) 38 Telenor (c) Norway 500.0 481.0 4% $3.6 $0.3 39 Tele Danmark (a) Denmark 479.0 513.0 -7% $4.8 $0.3 40 STAR Telecom. (a) U.S. 457.4 116.2 294% $0.6 $0.04 Note: Traffic figures are for public switched telephone neWvork (PSTN) circuits only (service resale is excluded). Data for U.S. and U.K. carriers include International Simple Resale (ISR). Carrier rankings based on originating country minutes only; when based on the aggregated traffic of all subsidiaries, the top multinational carriers include: AT&T/BT, MCI WorldCom, Cable & Wireless, and Teleglobe. International service revenues generally reflect net of PSTN ser- vice revenues after adding or subtracting for settlement payments, but may also include some private line revenue. All revenue figures converted from origi- nal currency at conversion rate current to year end reported. Some revenues figures have been estimated (est.). a. Data based on billing point of call, not originating point. b. Data are for the fiscal year ending 31 March. Telstra FY ends 30 June. c. Traffic totals based on TeleGeography estimates. Source: TeleGeography, Inc., FCC, and company reports © TeleGeography, inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 The Top U.S. International Carriers Facilities-Based Traffic Outgoing Facilities-Based,Traffic (millions of minutes) Rank Company 1998 1997 Growth 1 AT&T 9,878.7 10,290.0 -4% 2 MCI WorldCom 7,188.5 7,307.7 -2% 3 Sprint 2,926.0 2,758.9 6% 4 Teleglobe USA 830.3 301.6 175% 5 FaciliCom 678.0 223.2 204% 6 Pacific 6ateway Exchange 641.4 742.9 -14% 7 WorldxChange 595.8 430.1 39% 8 STAR Telecommunications 457.4 116.2 294% 9 RSL COM USA 214.3 94.4 127% 10 Viatel 202.0 58.5 245% 11 Primus Telecom 199.5 75.2 185% 12 New Global Telecom 168.0 n.a. n.a. 13 Cable & Wireless USA 69.0 44.0 57% 14 Tricom USA 66.1 29.2 127% 15 GTE Communications 47.7 10.8 341% Switched Resale Traffic Outgoing Switched Resale.Traffic (millions of minutes) Rank Company 1998 1997 Growth 1 STAR Telecommunications/PT-1 2,435.5 1,350.8 80% 2 MCI WorldCom 1,887.8 600.1 215% 3 IDT Corp. 1,563.7 300.5 420% 4 Telegroup 899.3 276.9 225% 5 Cable & Wireless USA 742.9 1,228.1 -40% 8 RSL COM USA 737.2 321.1 130% 7 Primus Telecom. 658.9 158.2 316% 8 Qwest Communications 489.7 343.4 43% 9 Intermedia Communications 409.3 27.0 1418% 10 Startec 388.1 120.9 221% 11 Corn Tech International 291.3 n.a. n.a. 12 AT&T 256.7 29.0 785% 13 Tel-Save 255.8 24.4 950% 14 AmeriVision Communications 238.5 n.a. n.a. 15 WorldxChange 188.5 n.a. n.a. 16 DirectNet 171.9 32.4 431% 17 Excel (Teleglobe USA) 186.6 119.6 39% 18 Sprint 150.0 124.2 21% 19 6TE Communications 144.8 17.4 730% Note: Traffic figures are for public switched network circuits based on billing point of call, not originating point. International Simple Resale (ISR) is included in facilities-based totals. Source: TeleGeography, Inc. and FCC © TeleGeography, Inc. 1999 TeleGeo~raphy 2000 © TeleGeography, Inc. 1999 Global Carrier Ownership Charts Sample Ownership Chart Primary Carrier Listing International Carrier 100 ~__~ocal Service Provider~ 100 Mobile Operator 100 ISP -~ 100 ~_~Non-telecom investor~ I Region or Division> Cable TV Operator -- -- -~ International Carrier 9O 10 Submarine Cable Key 5O solid lines represent ownership relationships numbers indicate percent interest held dashed lines show proposed acquisitions arrows indicate strategic alliance or other cooperative agreements Note: Charts include all carrier affiliates authorized to provide international facilities-based service, as well as major affiliates providing local, mobile, cable TV, or Internet service. Some communications investments have been highlighted with bold lines to denote importance; oth- ers have been omitted due to space limitations. All charts and market capitalizations are cur- rent to mid-October 1999. © Te[eGeography, tnc11999 © TeleGeography, Inc. 1999 TeleGeography 2000 TeleGeo~raphy 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Carrier Affiliates and Acquisitions Parent Carrier Headquarters Majority-Owned Affiliate Foreign Affiliate Market AT&T Corp. United States British Telecommunications plc United Kingdom Cable &Wireless plc United Kingdom COLT Telecom Group plc Global TeleSystems Group, Inc. United Kingdom AT&T CIS Ltd. Russia AT&T Communication Svcs Nederland BV Netherlands AT&T Communications (UK) Ltd. United Kingdom AT&T EasyLink Services Asia/Pacific Ltd. Hong Kong BT Australasia Pty Ltd. Australia BT (Hong Kong) Ltd, Hong Kong BT North America United States BT (Worldwide) Ltd. Belgium Clear Communications New Zealand Concert Global Networks (Canada) Ltd. Canada Cable & Wireless (Bermuda) plc Bermuda Cable & Wireless BET Ltd. (85%) Barbados Cable & Wireless Comms. plc (52,6%) United Kingdom Cable & Wireless Dominica Ltd. (80%) Dominican Republic Cable & Wireless France France Cable & Wireless Global Markets Inc. Canada Cable &Wireless Hong Kong Ltd. (54%) Hong Kong Cable &Wireless IDC Inc. (98%) Japan Cable & Wireless Ireland Ltd. Ireland Cable & Wireless Jamaica Ltd. (79%) Jamaica Cable & Wireless Optus Ltd. (52.3%) Australia Cable & Wireless Sakhalin Telecom (60%) Russia Cable & Wireless St. Kitts & Nevis (65%) St. Kitts & Nevis Cable & Wireless (Switzerland) AG Switzerland Cable & Wireless Tonga Tonga Cable & Wireless USA, Inc. United States Cable &Wireless (West Indies) Ltd. Antigua & Barbuda Cable & Wireless (West Indies) Ltd. Cayman Islands Cable & Wireless (West Indies) Ltd. St. Lucia Cable & Wireless (West Indies) Ltd. St. Vincent & Grenadines Companhia de Telecom. de Macau (51%) Macau Nakhodka Telecom (60%) Russia TeleYemen (51%) Yemen COLT Telecom AG Switzerland COLT Telecom Austria GmbH Austria COLT Telecom BV Netherlands COLT Telecom Espafia SA Spain COLT T~l~communications France SAS France United States COLT Telecom GmbH Germany COLT Telecom SA Belgium COLT Telecom SpA Italy COLT Telecom U.S. Corp. United States Esprit Telecom Benelux BV Belgium Esprit Telecom Benelux BV Netherlands Esprit Telecom de Espafia SA Spain Esprit Telecom Deutschland GmbH Germany Esprit Telecom France SA France Esprit Telecom UK Ltd. United Kingdom GTS Carrier Services Switzerland GTS Carrier Services BV Belgium GTS Carrier Services BV Netherlands GTS Carrier Services (France) SARL France GTS Carrier Services (Germany) GmbH Germany GTS Carrier Services Italy Srl Italy GTS Carrier Services Spain SA Spain : : Note: Affiliates include all ma r unless Otherwise noted; A!! p~rent Cgmp~nies are also ~odzed to p~-Ovide Service in t~ei~ home rnark~; TeleGeography 2000 © TeleGeography, Inc. 1999 Parent Carrier Headquarters Majority-Owned Affiliate Foreign Affiliate Market Global TeleSystems (continueo) United States Interoute Communications Group Long Distance International, Inc. MCI WorldCom, Inc. Primus Telecom. Group, Inc. RSL Communications, Ltd. United Kingdom United States United States United States United States GTS Carrier Services Sweden AB Sweden GTS Carrier Services (UK) Ltd. United Kingdom GTS Russia Russia InTouch Telecom NV Belgium ITL Ltd. Ireland NetSource Europe ASA Norway 0mnicom SA (98.6%) France WESTCom GmbH Germany ATI Telecom, Inc. (54.4%) United States Interoute Belgium BV Belgium Interoute Communications France France Interoute Comunica(;Oes Digitals SA (80%) Portugal Interoute Denmark A/S (88%) Denmark Interoute Ireland Ltd Ireland Interoute Netherlands BV Netherlands Interoute Networks Ltd. Canada Interoute Telecom Deutschland GmbH Germany Interoute Telecommunications Inc. United States Interoute Telecomunicaciones SA Spain Interoute Telecomunicazioni Italia SpA Italy Interoute SA, Switzerland Switzerland Dynamic de EspaSa Spain LDI (Sweden) Sweden LDI Telecom SA France Long Distance International Denmark ApS Denmark Long Distance International Ltd. Switzerland Long Distance International Ltd. United Kingdom NETnet AB Sweden NETnet A/S NonNay NETnet Italy SpA Italy NETnet Telekommunikation GmbH Austria Televersa NETnet GmbH Germany Embratel SA (51.79%) Brazil MCI WorldCom AB Denmark MCI WorldCom AB Sweden MCI WorldCom AS Norway MCI WoddCom Deutschland GmbH Germany MCI WorldCom Ireland Ltd. Ireland MCI WorldCom Japan Ltd. Japan MCI WorldCom Ltd. United Kingdom MCI WorldCom Netherlands Netherlands MCI WorldCom SA France MCI WorldCom SpA Italy MFS Communications of Canada, Inc. Canada WorldCom AG Switzerland WorldCom Asia Pacific Ltd. Hnng Kong WorldCom Australia Pry Ltd. Australia Corporate NeWvorks Ltd. United Kingdom Primus Canada Canada Primus Netherlands Netherlands Primus Telecommunications AG Switzerland Primus Telecommunications Deutschland GmbH Germany Primus Telecommunications KK Japan Primus Telecommunications Ltd. Ireland Primus Telecommunications Ltd. United Kingdom Primus Telecommunications Pty Ltd. Australia Primus Telecommunications SA France Telegroup Italia Srl Italy Telegroup Network Services Danmark Denmark European Telecom SA (90%) Belgium European Telecom SARL Luxembourg RSL C0M Asia International Services Corp. Hong Kong RSL C0M Australia Pty Ltd. Australia © TeleGeography, Inc. 1999 TeleGeography 2000 Figu Majority-Owned Affiliate Foreign Affiliate Market Parent Carrier Headquarters RSL Communications, Ltd. (continued)United States Teleglobe, Inc. Canada Viatel, Inc. United States World Access, Inc. United States RSL COM Austria AG (90%) Austria RSL COM Canada Inc. Canada RSL COM Danmark NS Denmark RSL COM Deutschland GmbH Germany RSL COM Finland 0y Finland RSL COM France SA France RSL COM Ireland, Ltd. Ireland RSL COM Italia Srl (99.3%) Italy RSL COM Japan KK Japan RSL COM Mexico (80%) Mexico RSL COM Nederland BV Netherlands RSL COM Schweiz AG (78.5%) Switzerland RSL COM Sweden AB Sweden RSL COM UK Ltd. United Kingdom RSL Commmunications Spain SA (90%) Spain Excel Communications, Inc. United States Teleglobe Australia Pty Ltd. Australia Teleglobe BV Netherlands Teleglobe Communications Corp. United States Teleglobe Danmark AS Denmark Teleglobe El Salvador, SA de CV El Salvador Teleglobe France SAS France Teleglobe GmbH Germany Teleglobe Hong Kong Ltd. Hong Kong Teleglobe International Communications SA Spain Teleglobe International (UK) Ltd. United Kingdom Teleglebe Ireland Ltd. Ireland Teleglobe Italia SpA Italy Teleglobe Japan KK Japan Teleglobe Norge AS Norway America First Ltd. United Kingdom American Telemedia Ltd. United Kingdom Destia BV Belgium Destia Canada, Inc. Canada Destia France SARL France Destia GmbH Germany Destia Service AG Switzerland Viatel Belgium SA Belgium Viaphone GmbH Germany Viatel Global Communications BV Netherlands Viatel Global Communications Espa~a SA Spain Viatel OpOrations SA France Viatel SRL Italy Viatel UK Ltd. Ireland Viatel UK Ltd. United Kingdom FaciliCom International BV Belgium FaciliCom International BV Netherlands FaciliCom International Denmark NS Denmark FaciliCom International, LLC New Zealand FaciliCom International SARL France FaciliCom International Ltd. United Kingdom FaciliCom Telekommunikation GmbH Germany FCI Austria Austria FCI Finland Finland FCI Italy Italy FCI Norway Norway FCI Spain Spain FCI Switzerland Switzerland TeleS/FaciliCom International Sweden AB Sweden World Access Telecommunications Group Ltd. United Kingdom FCI NonNay Norway FCI Switzerland Switzerland FaciliCom International Ltd. United Kingdom TeleGeography 2000 © TeleGeography, Inc. 1999 Buyer Country Target Country Date Value US$ AT&T Corp. U.S. BT’s international operations U.K. Oct. 99 Merger of Equals Japan Telecorn Co. Ltd. (15%) Japan Pending $950m Bell Atlantic Corp. U.S. GTE Corp. U.S. Pending $82,800rn British Telecommunications plc U.K. AT&T Corp.’s international operations U.S. Oct. 99 Merger of Equals LG Telecorn Co. Ltd. (23.49%) South Korea 0ct-98 $364m Japan Telecom Co. Ltd. (15%) Japan Pending $950m Binariang Bhd (33.3%) Malaysia 0ct-98 $396rn Cable & Wireless plc Japan International Digital Communications Inc. Japan Sep-99 $758m Communications Systems International U.S. Primecall Inc. U.S. Mar-99 Undisclosed Compafiia Telecoms. de Chile Chile VrR Larga Distancia S.A. (now Globus) Chile 0ct-98 $49.9m Deutsche Telekom AG Germany Hrvatska Telekom dd (35%) Croatia Pending $850m Energis plc U.K. Unisource Carrier Services Netherlands Pending $95m Esat Telecom Group plc Ireland Global One’s Irish unit Ireland Apr-99 $5m France T~l~com S.A. France Telecom Argentina (19.5%; 30%) Argentina Aug-99 $265m Global Crossing Ltd. Bermuda Frontier Corp. U.S. Pending $11,200m Global TeleSysterns Group Inc. U.S. Esprit Telecom Group plc U.K. Mar-99 $849.9m Hermes Europe Railtel B.V. Belgium Jul-99 Undisclosed InTouch Telecom Belgium N.V. Belgium Aug-99 Undisclosed NetSource Europe S.A. Norway Dec-98 $46.1rn Omnicom (98.6%) France May-99 Undisclosed ICG Communications Inc. U.S. ChoiceCom LP U.S. Dec-98 $55m Interoute Communications Group Ltd. U.K. Skipper Telecomunicazioni (majority %) Italy Nov-98 Undisclosed Network Telecom Services (75%) Netherlands Nov-98 Undisclosed ATI Telecom Inc. U.S. Mar-99 Undisclosed American International Telephone Inc. U.S. Mar-99 Undisclosed IXnet U.S. Saturn Global Network Services Holdings Ltd. U.K. Dec-98 $47.5m Japan Telecom Co. Ltd. Japan AT&T Jens Corp. Japan Aug-99 $453m BT Communication Services Ltd. Japan Aug-99 $274.4m KDD Corp. Japan Teleway Corp. Japan Dec-98 $424m Kingston Communications Group U.K. Teleport London International U.K. Feb-99 Undisclosed Long Distance International Inc. U.S. NETnet International AB Sweden Dec-98 Undisclosed Luca S.A. Guatemala Telgua (95%) Guatemala 0ct-98 $700m Mannesmann AG Germany Infostrada SpA (50%; 100%) Italy Jun-99 Undisclosed ISIS Multimedia Net GmbH (64.9%) Gerrnany Pending Undisclosed Mannesrnann Arcor (15%; 70%) Germany May-99 Undisclosed o.tel.o GmbH Germany May-99 $542m Metromedia International Group, Inc. U.S. PLD Telekom Inc. U.S. Pending $132m MCI Worldcom, Inc. U.S. Sprint Corp. U.S. Pending $115m NTL Inc. U.S. CWC ConsumerCo. U.K. Pending $10,000m Eastern Group Telecoms U.K. Pending $144.4m Nippon Telegraph and Telephone Corp. Japan Philippine Long Distance Telephone Co. (15%) Philippines Pending $372m Olivetti SpA Italy Telecom Italia SpA Italy Jun-99 $33,000m Pointe Cornmunications Corp. U.S. Pensat International Communications Inc. U.S. Jul-99 Undisclosed PrimusTelecornmunications Group,Inc. U.S. London Telecorn Group Inc. Canada Pending $50.6m Telegroup, Inc. U.S. Jun-99 $72m Telephone Savings Network Ltd. Canada Jun-99 $5m Qwest Cornms. International, Inc. U.S. US West Inc. U.S. Pending $50,900m SBC Communications, Inc. U.S. Ameritech Corp. U.S. Pending $60,000m Southern New England Telecommunications Corp. U.S. 0ct-98 $5,700m Startec Global Communications Corp. U.S. PCI Communications, Inc. U.S. Dec-98 Undisclosed Phone Systems and Network S.A. (64.6%) France Feb-99 $3.8rn Star Telecommunications, Inc. U.S. PT-1 Communications Inc. U.S. Feb-99 $195m Telecom Italia Italy Telecom Argentina (19.5%/30%) Argentina Aug-99 $265m Entel Chile S.A. (55%; 22.5%) Chile Jun-99 Undisclosed Telefbnica S.A. Spain European Telecom International GrnbH Austria Feb-99 Undisclosed Teleglobe, Inc. Canada Excel Communications, Inc. U.S. Nov-98 Undisclosed Telia AB Sweden Telenor AS Norway Pending Merger of Equals Telenor AS Norway Telia AB Sweden Pending Merger of Equals tele.ring Telekom Service GmbH Austria Citykom Austria Telekommunikation GmbH Austria Jan-99 Undisclosed Televista, Inc. U.S. Allied Global Telecommunications, Inc. U.S. Pending Undisclosed Cross Communications, Inc. U.S. Pending Undisclosed Transaction Network Services Inc. U.S. Transline Communications Inc. U.S. May-99 $4m Viatel, Inc. U.S. Destia Communications, Inc. U.S. Pending $600m World Access, Inc. U.S. Comm/Net Holding Corp. U.S. Jul-99 $27m FaciliCom International, Inc. U.S. Pending $436m Resurgens Communications Group U.S. Dec-98 Undisclosed Shares un!ess~e~i~ in ~se~ ~e~etwo pe~’~geS are listed, th~ i ~ompleti~ #f ~a~sa~o~ lather than armou~eme~0~)~tentio~ t# ~ge © TeleGeography, Inc. 1999 TeleGeography 2000 Buyer Country Country Date Allegiance Telecom Inc. U.S. U.S. Jun-99 AT&T Corp. U.S. U.S. Pending Brazil Pending British Telecommunications plc U.K. Spain Feb-99 Italy Pending Cable & Wireless plc U.K. Germany Jan-99 U.K. Jul-99 Cable & Wireless Hong Kong Ltd. Hong Kong Hong Kong Mar-99 CoreComm Ltd. U.S. U.K. Jul-99 U.S. Jun-99 U.S. Dec-98 France T~i~com S.A. France U.S. Dec-98 Belgium Nov-98 France Dec-98 Denmark 0ct-98 Global TeleSystems Group, Inc. U.S. Denmark Jul-99 ICG Communications, Inc. U.S. U.S. Jan-99 MCI WorldCom, Inc. (UUNet) U.S. Primus Telecommunications Group, Inc. U.S. PSINet Inc. U.S. Telia Sweden Uni2 Spain Target Value US$ KIVEX.com $34.5m IBM Global NeWvork $5,000m Netstream $300m Arrakis $15.7m I.Net (32.5%) $21.5m ECRC NeWvork Services GmbH $43.6m Internet Network Services Group Undisclosed FIC NeWvork Service Inc. (85%) $5m Internet NeWvork Services Group Undisclosed MegslNet Inc. $16.8m Stratos Internet Group Inc. Undisclosed Stratos Intemet Group Inc. Undisclosed EuroNet Internet BV Undisclosed NCTech (67%) Undisclosed WEB NS (80%) Undisclosed Ebone A!S (25%; 100%) Undisclosed NETCOM 0n-Line Communication Services, Inc. Undisclosed 0zEmail Ltd. Australia Mar-99 $322.8m GlobalServe Communications Canada Feb-99 $4.3m TCP/IP GmbH Germany May-99 Undisclosed TouchNet GmbH Germany Sep-99 Undisclosed Argentina 0n-Line Argentina Jul-99 Undisclosed AsiaNet Ltd. Hong Kong Dec-98 Undisclosed Caribbean Intemet Puerto Rico Jun-99 Undisclosed Ciberia Intemet Spain Sep-99 Undisclosed DataNet Mexico May-99 Undisclosed Domain Acesso e Servicos Internet Ltd. Brazil Aug-99 Undisclosed Elender Informatikai es Szamitastechnikai Rt. Hungary Sep-99 Undisclosed Global Link Hong Kong Aug-99 Undisclosed Horizontes Internet Brazil Apr-99 Undisclosed Huge Net Ltd. Hong Kong Dec-98 Undisclosed Inet Inc. South Korea 0ct-98 Undisclosed Infase Comunicaciones Spain Sep-99 Undisclosed Internet de M~xico Mexico May-99 Undisclosed Internet Exchange Europe b.v. Netherlands 0ct-98 Undisclosed Internet Network Technologies U.S. Sep-99 Undisclosed Netline Comunicaciones S.A. Chile Sep-99 Undisclosed Netsystem Argentina Sep-99 Undisclosed 0penlink Brazil Apr-99 Undisclosed Planete.net France Mar-99 Undisclosed Rimnet Japan Aug-99 Undisclosed Satelnet France Mar-99 Undisclosed ServNet Serv. de Informatica e Comm. Brazil Sep-99 Undisclosed Sinfonet Panama Aug-99 Undisclosed Spider Net Ltd. Hong Kong Dec-98 Undisclosed TBA Internet Brazil Sep-99 Undisclosed The Internet Access Co. U.S. Jun-99 Undisclosed The Internet Co. Switzerland Jun-99 Undisclosed Tokyo Internet Corp. Japan 0ct-98 Undisclosed TotalNet Canada Sep-99 Undisclosed Transaction NeWvork Services, Inc. U.G. Pending $720m Unix Support Nederland BV Netherlands 0ct-98 Undisclosed Vision NeWvork Ltd. Hong Kong Sep-99 Undisclosed Zircon Systems PW Ltd. Australia Oct-99 Undisclosed Unisource Iberia Spain Sep-99 Undisclosed CTV-Jet Spain Mar-99 Undisclosed TeleGeography 2000 © TeleGeography, Inc. 1999 Buyer Country Target Country AT&T Corp. U.S. American Cellular Corp. U.S. Rogers Cantel Mobile Communications (16.5%) Canada Bouygues SA France Bouygyes Telecom (20%) France Deutsche Telekom AG Germany max.mobil (46%; 71%) Austria 0ne20ne U.K. Polska Telefonia Cyfrowa SA (22.5%; 45%) Poland Russian Telecommunications D evelopment Corp. Russia Westel 450 (49%) Hungary Westel 900 (49%) Hungary France T~l~com SA France E-Plus (77.24%) Germany MATAV Rt. Hungary Westel 450 (49%; 100%) Hungary Westel 900 (49%; 100%) Hungary Interoute Communications Group Ltd. U.K. WorldWide Mobilcom Corp. U.S. Lietuvos Telekomas Lithuania Comliet UAB (24.5%; 49%) Lithuania Mannesmann AG Germany Omnitel (23.4%; 55%) Italy Orange plc U.K. SBC Communications Inc. U.S, Tele Danmark NS Denmark TelefOnica SA Spain Telef0nos de M~xico SA de CV Mexico Vodafone 6roup pic U.K. Cellular Communications of Puerto Rico Inc. (50%) Puerto Rico Comcast Cellular Corp. U.S. UAB Bit~ GSM (71.6%) Lithuania Telebahta Celular (50%) Brazil Telergipe Celular (50%) Brazil Telerj Celular Brazil Telest Celular Brazil Cellular Communications of Puerto Rico Inc. (50%) Puerto Rico AirTouch Communications Inc. U.S. Digital Phone 6roup (8%) Japan CommNet Cellular Inc. U.S. New Zealand GSM Australia Date Pending Pending Pending Pending 0ct-99 0ct-99 0ct-99 0ct-99 0ct-99 0ct-99 Pending Pending Mar-99 Sep-99 Jun-99 Pending Pending Pending Sep-99 Jul-99 Jul-99 Jul-99 Jul-99 Pending Jun-99 0ct-99 Pending Feb-99 Dates indicatetirne! Value US$ $370m $478m $1,130m Undisclosed $11,120m Undisclosed Undisclosed Undisclosed Undisclosed $1,091m $442.5m $442.5m $3.5m Undisclosed Undisclosed $34,700m $232m $400m $50m $30m $30m $60m $60m $232m $74,400m $411m $764m $385m Pricing TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Elements of an International Call An international service provider has a number of options to send its customers’ calls abroad. This section compares the cost ele- ments of those options. Referring to the table on pages 60 to 61, let’s use a call from New York to Berlin as an example. Not including call-back, refile, and other forms of non-traditional traf- fic switching, a U.S. carrier has four basic methods of transport- ing its customer’s call to the destination in Germany: I. Carrier Settlement. To switch the call from the customer’s telephone to its own long distance network, the international carrier first pays the Local Exchange Carrier (LEC) in New York an origination fee of 2.1 ¢ per minute. Next, the carrier moves the call along a "backhauI" route; that is, from its national network to the undersea cable landing station on the Atlantic Ocean shoreline. The carrier shifts the call onto the interna- tional "half circuit" it owns, then pays the German carrier a settlement fee to transfer the call onto its matching half circuit and to the final destination. The U.S. carrier’s marginal cost of using its own backhaul and international circuit is insignifi- cant: 0.1¢ per minute. (Note: because half circuit purchas- es on new cables are increasingly rare, the following tables show only whole circuit pricing to Germany.) The settlement rate, at 10¢ per minute, is far less inexpensive. Total cost: 12¢ per minute. 2. Carrier Interconnect. Competition rules in Germany per- mit foreign carriers to interconnect directly with the domestic telephone network. Rather than financing a half circuit and paying a settlement fee, a U.S. carrier can purchase a whole circuit all the way to an international gateway in Germany, then pay the German carrier a 2.7¢ per minute fee to switch the call to Berlin. Some international carriers may own a LEC themselves, in which case this charge is simply an intra-cor- porate transfer. Total cost, including origination and backhauh 5¢ per minute. 3. International Simple Resale (ISR). A carrier is not required to own its own circuits. Instead, it can switch traffic onto U.S.- Germany private lines leased from other carriers. Total cost, including origination, backhaul, private line lease, and inter- connection in Germany: 6¢ per minute. 4. Service Resale. A telephone service provider may wish to avoid carrying its own traffic to Germany altogether by pur- chasing the minutes transported over another carrier’s net- work in bulk and marketing those minutes as its own. The charge required for end-to-end service resale is a "wholesale rate" covering origination, U.S. domestic long distance, and the underlying carrier’s international transpor~ and termina- tion charges. Total cost: 7¢ per minute. The following pages examine the component costs of provisioning an international call in more detail. Statistics on settlement rates and interconnection charges plus a review of retail prices show how carriers are passing along their costs to end users. In addi- tion, tables and charts on telecom service trading and interna- tional call quality metrics examine the new models for purchasing minutes and bandwidth. ionents Country A Half Circuit I il II II Local National Backhaul International Exchange Transport Transport Network Elements Covered by Settlement Rate H~if Circuit Country B National Network Network Elements Covered by Sottiement Rate TeleGeography 2000 © TeleGeography, Inc. 1999 International Carrier Call Costs from the U.S. Origination Cost Per Minute (;nst (U.S. Cents) Int~l Circuit Int’l Circuit Settlement Interconnect Wholesale Ownership Lease Rate Rate Rate Total Americas U.S.-Canada Retail Price ...... 10.0 Carrier Settlement 2.1 -- -- 10.0 -- -- 12.1 Carrier Interconnect 2.1 0.1 -- -- 0.7 -- 2.9 ISR 2.1 -- 0.3 -- 0.7 -- 3.1 Wholesale for resellers .... 4.6 4.0 -- -- -- 42.0 19.0 -- -- 21.2 -- -- -- rl.a. -- -- -- n.a. -- -- 22.9 22.9 U.S.-Mexico Retail Price -- Carrier Settlement 2.2 Carrier Interconnect -- ISR -- Wholesale for resellers -- Europe U.S.-Germany Retail Price ...... 20.0 Carrier Settlement 2.1 -- -- 10.0 -- -- 12.1 Carrier Interconnect 2.1 0.1 -- -- 2.7 -- 4.9 ISR 2.1 -- 1.1 -- 2.7 -- 5.9 Wholesale for resellers .... 7.1 7.1 U.S.-U.K. Retail Price ...... 10.0 Carrier Settlement 2.1 -- -- 11.0 -- -- 13.1 Carrier Interconnect 2.1 0.1 -- -- 1.7 -- 3.9 ISR 2.1 -- 0.4 -- 1.7 -- 4.2 Wholesale for resellers .... 6.9 6.9 Notes: Prices are indicative of carriers’ cost per call, but may not reflect actual costs. All cost components expressed in U.S. cents and are exclusive of taxes. Cost totals expressed in U.S. cents. Rates are based on international calls originating from Washington, D.C. only. All rates are current as of August 1999. Origination cost includes access charges paid to Local Exchange Carrier (Bell Atlantic) and U.S. domestic network costs for transmitting calls to international gate- way. Circuit ownership costs reflect half circuit ownership for Australia, China, and Japan. All other circuit ownership costs are for whole circuits. Circuit ownership costs include price of backhaui. CaJculations converting circuit ownership prices to per minute costs assume that each 64 kbps circuit is compressed at a 5:1 ratio and is used for ten years and that each voice path is used 4 hours (240 minutes) per day. Interconnection rates show price for national termination. Direct interconnection by foreign carriers to the domestic public switched telephone neWvork is not permitted in China, India, and Mexico. Settlement rates are for peak rate traffic terminated by the largest foreign carrier. U,S.-Mexico settlement rates vat7 according to originating and terminating locations in both countries. The simple average for all U.S.-Mexico rates is presented here. Retail rates are based on the MCI One international calling plan. © TeleGeography, Inc. 1999 TeleGeography 2000 Origination Cost Per Minute Cost (U.S. Cents) Int’l Circuit Int’l Circuit Settlement Interconnect Wholesale Ownership Lease Rate Rate Rate Total Asia U.S.-Australia Retail Price ...... 25.0 Carrier Settlement 2.2 -- -- 15.0 -- -- 17.2 Carrier Interconnect 2.2 2.2 -- -- 4.0 -- 8.4 ISR 2.9 -- 4.9 -- 4.0 -- 11.8 Wholesale for resellers .... 6.8 6.8 U.S.-China Retail Price -- Carrier Settlement 2.1 Carrier Interconnect -- ISR -- Wholesale for resellers -- 58.0 -- -- -- -- 48.2 131.0 60.1 n.a. n.a. 48.2 U.S.-India Retail Price .... Carrier Settlement 2.1 1.0 -- 64.0 Carrier Interconnect ... ISR Wholesale for resellers ... 122.0 67.1 -- -- 69.4 69,4 U.S.-Japan Retail Price ...... 35.0 Carrier Settlement 2.1 -- -- 14.0 -- -- 16.1 Carrier Interconnect 2.1 0.1 -- -- 5.5 -- 7.7 ISR 2.3 -- 5.5 -- 5.5 -- 13.3 Wholesale for resellers .... 12.4 12,4 Sources: Federal Communications Commission, European Union Commission, national regulaton/agencies, and industry interviews Concept: TeleGeography, Inc. © TeleGeography, Inc, 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 International Carrier Call Costs to the U.S. Origination Cost Per Minute Cost (U.S. Cents) Int~l Circuit Int’l Circuit Settlement Interconnect Wholesale Ownership Lease Rate Rate Rate Total Americas Canada-U.S. Retail Price -- 14.0 Carrier Settlement 1.3 -- -- 10.0 -- -- 11.3 Carrier Interconnect 1.3 0.1 -- -- 1.5 -- 2.9 ISR 1.3 -- 0.3 -- 1.5 -- 3.1 Wholesale for resellers -- 4.6 4.6 -- -- -- 86.0 19.0 -- -- 21.6 -- -- -- -- -- -- n.a. -- -- 26.0 26.0 Mexico-U.S. Retail Price -- Carrier Settlement 2.6 Carrier Interconnect -- ISR -- Wholesale for resellers -- Europe Germany-U.S. Retail Price ...... 23.0 Carrier Settlement 1.8 -- -- 10.0 -- -- 11.8 Carrier Interconnect 1.8 0.1 -- -- 1.5 -- 3.4 ISR 1.8 -- 1.1 -- 1.5 -- 4.4 Wholesale for resellers .... 5.5 5.5 U.K.-U.S. Retail Price ...... 32.0 Carrier Settlement 1.6 -- -- 10.0 -- -- 11.6 Carrier Interconnect 1.6 0.1 -- -- 1.5 -- 3.2 ISR 1.6 -- 0.4 -- 1.5 -- 3.5 Wholesale for resellers -- 3.8 3.8 Notes: Prices are indicative of carriers’ cost per call, but may not reflect actual costs. All cost components expressed in U.S. cents and are exclusive of taxes. Cost totals expressed in U.S. cents. Rates are based on international calls originating from Washington, D.C. only. All rates are current as of August 1999. Origination costs reflect regional access charges. U.K. origination costs are for BT local exchange segment and do not include trunking charges. Origination costs for Japan are estimates. Non-U.S. carriers may own significant portions of home country local network, in which case origination costs are counted as intra-corporate transfers. Calculations converting circuit ownership prices to per minute costs assume that each 64 kbps circuit is compressed at a 5:1 ratio and is used for ten years and that each voice path is used 4 hours (240 minutes) per day. Circuit ownership costs reflect half circuit ownership for Australia, China, and Japan. All other circuit ownership costs are for whole circuits. Circuit ownership costs include price of backhaul. Direct interconnection to the U.S. domestic public switched telephone neWvork is not permitted for carriers from China, India, and Mexico. Settlement rates are for peak rate traffic terminated by the largest foreign carrier. U.S.-Mexico settlement rates vary according to originating and terminating locations in both countries. The simple average for all U.S.-Mexico rates is presented here. © TeleGeography, Inc. 1999 TeleGeography 2000 Per Minute Cost (U.S. Cents) Origination Int~l Circuit Int’l Circuit Settlement Interconnect Wholesale Cost Ownership Lease Rate Rate Rate Total Asia Australia-U.S. Retail Price ...... 53.0 Carrier Settlement 2.2 -- -- 15.0 -- -- 17.2 Carrier Interconnect 2.2 1.1 -- -- 1.5 -- 4.8 ISR 2.2 -- 4.9 -- 1.5 -- 8.6 Wholesale for resellers .... 9.8 9.6 China-U.S. Retail Price ...... 222.0 Carrier Settlement 2.1 -- -- 58.0 -- -- 60.1 Carrier Interconnect ... n.a. ISR n.a. Wholesale for resellers ... n.a. India-U.S. Retail Price ...... 138.0 Carrier Settlement 2.1 -- -- 64.0 -- -- 66.1 Carrier Interconnect ... n.a. ISR n.a. Wholesale for resellers ... n.a. Japan-U.S. Retail Price ...... 69.0 Carrier Settlement 1.6 -- -- 14.0 -- -- 15.6 Carrier Interconnect 1.6 0.1 -- -- 1.5 -- 3.2 ISR 1.6 -- 5.5 -- 1.5 -- 8.6 Wholesale for resellers .... 18.4 18.4 Sources: Federal Communications Commission, European Union Commission, Philips Tarifica, national regulatow agencies, and industw interviews Concept: TeleGeography, Inc. © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 International Private Line Prices S30,O00 $25,000 $20,000 $15,000 $10,000 $5,000 $0 O E Europe O E Europe Africa & M East Asia S America W Europe ¯ World Avg © N & C America Africa & M East S America Asia W Europe World Avg N & C America 1995 1996 E Europe Africa & M East S America Asia W Europe World Avg © N & C America O E Europe Africa & M East S America Asia ~W Europe World Avg N & C America 1997 1998 Note: Data reflect averages of annual revenue collected by U.S. international carriers for 64 kbps circuit ]eases to countries within each region. Source: FCC carrier filings, TeleGeography, Inc. © TeleGeegraphy, |nc, 1999 Figure 2. Band-X Bit Index, 1998-1999 120.0 80.0 ~~ Trans.Pacific Intra-Europe 20.0 © TeleGeography, Inc. 1999 TeleGeography 2000 International Settlement Rates United States Destination 1997 1998 1999 Andorra 0.29 0.29 0.29 Argentina 0.43 0.35 0.33 Australia 0.21 0.15 0.15 Austria 0.21 0.14 0.14 Bahamas 0.30/0.15 0.30/0.15 0.30/0.15 Bahrain 0.80 0.65 0.55 Bangladesh 0.80 0.80 0.80 Belarus 0.50 0.43 0.43 Belgium 0.19 0.14 0.14 Bolivia 0.55 0.40 0.43 Brazil 0.43 0.33 0.30 Canada 0.10/0.06 0.10/0.06 0.10/0.06 Chile 0.50 0.35 0.35 China 0.85 0.70 0.58 Colombia 0.50 0.40 0.38 Costa Rica 0.40 0.35 0.29 Croatia 0.34 0.28 0.25 Cyprus 0.48 0.38 0.37 Czech Republic 0.31 0.28 0.18 Denmark 0.14 0.11 0.11 Dominican Republic 0.35 0.30 0.26 El Salvador 0.44 0.39 0.31 Finland 0.21 0.16 0.14 France 0.13 0.11 0.10 French Polynesia 1.25 0.70 0.70 Germany 0.10 0.11 0.10 Ghana 0.50 0.50 0.50 Greece 0.43 0.28 0.17 Guyana 0.85 0.85 0.85 Hong Kong 0.40 0.30 0.07 Hungary 0.31 0.28 0.19 Iceland 0.38 0.24 0.14 India 0.71 0.64 0.64 Indonesia 0.65 0.53 0.48 Iran 1.25 1.05 0.90 Ireland 0.17 0.11 0.10 Israel 0.35 0.30 0.15 Italy 0.17 0.11 0.11 Japan 0.43 0.15 0.14 Jordan 0.75 0.68 0.50 United Kingdom 1998 (US$) 1999 (US$) 0.22 0.13 0.86 0.56 0.28/0.17 0.24/0.08 0.20 0.19 0.38 0.36 0.82 0.64 0.99 0.97 0.35 0.34 0.10 0.10 0.90 0.89 0.49 0.36 O. 10/0.08 O. 10/0.04 0.90 0.89 1.08 0.89 0.90 0.56 0.69 0.47 0.33 0.33 0.25 0.20 0.21 0.20 0.10 0.07 0.67 0.56 1.54 1.18 0.15 0.13 0.11 0.10 1.64 1.27 0.10/0.08 0.10/0.04 0.66 0.52 0.29 0.24 0.90 0.89 0.45 0.42 0.18 0.18 0.23 0.23 O.95 O.87 1.21 0.64 1.21 1.18 0.16 0.16 0.25 0.24 0.16 0.13 0,59 0.48 1.21 0.97 Notes: 1. All rates expressed in US$. Equivalent dollar values are presented for accounting rates that are established in Special Drawing Rights (SDRs), gold francs, or pounds sterling. 2. The average U.S. accounting rate for 1997 and 1998 is weighted by the total minutes between the U.S. and each location in that year. U.S. 1999 rates are for Augus~ 1999. U.K. rates are for July 1998 and April 1999. 3. Where Wvo rates are shown, there are peak/off-peak rates or growth-based rates (traffic above a benchmark level is eligible for a lower rate). 4. Rates are for largest carrier serving the route. Different accounting rates may apply to competing carriers. © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 Destination Kazakhstan Korea, Rep. Kuwait Luxembourg Macau Malaysia Mexico Moldova Netherlands New Zealand Norway 0man Pakistan Panama Paraguay Peru Philippines Poland Portugal Russia Saudi Arabia Singapore SIovak Republic SIovenia South Africa Spain Sri Lanka Sweden Switzerland Taiwan Thailand Turkey Ukraine United Arab Emirates United Kingdom Uruguay United States Uzbekistan Venezuela Vietnam Yugoslavia 1997 0.80 0.49 0.79 0.14 0.60 0.40 0.35 1.04 0.14 0.14 0.11 1.13 1.00/0.60 0.60 0.60 0.50 0.50 0.35 0.30 0.80 1.02/0.81 0.43 0.34 0.34 0.50 0.24 1.00 0.06 0.17 0.57 0.60 0.41 0.60 1.00/0.65 0.10/0.07 0.55 0.80 0.49 1.65/1.00/0.93/0.85 0.41 United States 1998 0.69 0.43 0.80 0.14 0.50 0.40 0.37 1.04 0.10 0.14 0.09 0.75 0.60 0.48 0.50 0.43 0.36 0.28 0.22 0.40 0.87 0.26 0.29 0.35 0.40 0.13 0.80 0.06 0.14 0.23 0.45 0.38 O.50 1.00/0.65 0.11/0.07 0.43 II.a. 0.70 0.4O 1.65/1.00/0.93/0.85 0.38 1999 0.51 0.36 0.78 0.14 0.50 0.4O 0.19 1.04 0.07 0.14 0.08 0.68 0.60 0.45 0.40 0.33 0.29 0.21 0.15 0.35 0.68 0.15 0.29 0.34 0.35 0.14 0.60 0.06 0.14 0.23 0.35 0.33 0.50 1.00/0.65 0.11/0.07 0.33 0.60 0.32 0.78 0.34 United Kingdom 1998 (US$) 1999 (US$) 0.82 0.64 0.64 0.50 0.82 0.80 0.23 0.24 O.53 0.51 0.51 0.5O 0.45 0.44 0.30 0.24 0.07 0.06 0.20 0.19 0.07 0.07 0.82 0.80 0.66 0.64 0.77 0.64 0.90 0.80 0.74 0.72 0.49 0.48 0.28 0.25 0.23 0.18/0.14 0.37 0.27 1.27 0.89 0.59 0.32 0.19/0.10 0.19/0.10 0.18/0.16 0.16/0.09 0.66 0.48 0.16 0.16 0.90 0.89 0.12 0.12 0.08/0.07 0.08/0.04 0.57 0.44 0.82 0.80 0.32 0.30 0.31 0.29 0.49 0.32 n.a. n.a. 0.97 0.95 0.12/0.08 0.12/0.04 0.99 0.80 0.82 0.80 1.31 1.29 0.28 0.27 Source: FCC; Office of Telecommunications (OF-TEL) © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 FCC Settlement Benchmarks Motivated by the annual multi-billion dollar settlements outflow by U.S. carriers, the Federal Communications Commission (FCC) pro- posed in 1996 a set of "benchmark" or model settlement rates (see Report and Order IB Docket No. 96-261, FCC 96-484, (www.fcc.gov/Bureaus/Intema~Jonal/No~es! 1996ifcc96484.txt). Beginning in 1999, these benchmarks capped the amount U.S. carriers could pay their foreign correspondents for traffic exchange at rates ranging from $0.15 to $0.23 per minute. The FCC calculated benchmarks based on the price for the three net- work elements used to provide international phone services, as iden- tified by recommendation D. 140 of the ITU’s Standardization Sector (ITU-~ see www.itu.int/intset/itu-t/d140/dl 40_e_rev.htm). Known as Tariffed Component Prices (TCPs), these include international trans- mission facilities (cable and satellite half circuits); international switching facilities; and national extension facilities (domestic trans- port and termination). To obtain prices for the first and last network elements--international transmission and national extension--the FCC used 1996 rates for the largest carrier in each country, based on standard equipment. For international transmission, this meant an international private line at 1.5 Mbps or above, a 4:1 compression ratio on each 64 Kbps circuit, and a usage level of 8,000 minutes per 64 Kbps circuit per month. For national extension pricing, the FCC relied upon national long dis- tance tariffs, smoothing according to average expected time-of-day and distance for inbound traffic. The per minute costs for interna- tional switching were derived from the accounting rate share figures identified in ITU-T Recommendation D.300R’s international exchange section (www.itu.intlintsetlitu-tld3001d300r.htm). The FCC adopted the Benchmarks Order in August 1997, with imple- mentation staggered over several years, based on national incomes. Settlement rates to high income countries have already been affect- ed, following the 1999 deadline; other countries are being covered by the Benchmarks Order in turn, ending with the low income countries by 2003. As the table below demonstrates, most countries in the upper income bracket have fallen into line, with settlement rates at or below benchmarks. Settlement rates for countries that have met benchmarks are shown in bold. Tariffed Component Prices for FCC Benchmarks (U.S. cents) International International Netional Tariffed Component FCC Settlement August 1999 Country Transmission 4- Switching 4- Extension -- Price Benchmarks Settlement Rate Upper Income Bracket: Effective I January 1999 Australia 4.8 1.9 12.0 18.7 15.0 15o0 Austria 8.1 1.9 21.4 31.4 15.0 14.0 Bahamas 5.2 1.9 12.8 19.9 15.0 30.0/15.0 Belgium 3.0 1.9 9.2 14.1 15.0 14.0 Denmark 5.9 1.9 6.6 14.4 15.0 11.0 France 2.9 1.9 12.7 17.5 15.0 10.0 Germany 4.3 1.9 13.6 19.8 15.0 10.0 Hong Kong 5.1 1.9 0.0 7.0 15.0 7.0 Ireland 2.7 1.9 13.4 18.0 15.0 10.0 Israel 4.2 1.9 2.4 8.5 15.0 15.0 Italy 4.8 1.9 11.5 18.2 15.0 11.0 Japan 6.5 1.9 11.3 19.7 15.0 14.0 Kuwait 7.1 1.9 0.0 9.0 15.0 78.0 Netherlands 2.6 1.9 5.3 9.8 15.0 7.0 New Zealand 5.7 1.9 16.2 23.8 15.0 14.0 Norway 3.2 1.9 6.5 11.6 15.0 0.0 Portugal 4.6 1.9 17.4 23.9 15.0 15.0 Singapore 5.0 1,9 0.7 7.6 15.0 15.0 Spain 4.8 1.9 11.4 18.1 15.0 14.0 Sweden 3.6 1.9 4.5 10.0 15.0 6.0 Switzerland 4.4 1.9 14.3 20.6 15.0 14.0 Taiwan 5.7 1.9 6.3 13.9 15.0 23.0 United Arab Emirates 3.3 1.9 2.5 7.7 15.0 1.00/65.0 United Kingdom 2.4 1.9 8.7 13.0 15.0 11.0/7.0 TeleGeography 2000 © TeleGeography, Inc. 1999 Tariffed Component Prices for FCC Benchmarks (continued) International International National Tariffed Component FCC Settlement Country Transmission 4- Switching 4- _- Extension Price Benchmarks Upper Middle Income Bracket: Effective I January 2000 August 1999 Settlement Rate Argentina 6.7 3.4 22.0 32.1 19.0 33.0 Barbados 8.6 3.4 0.0 12.0 19.0 53.0 Brazil 6.6 3.4 17.8 27.8 19.0 30.0 Chile 2.9 3.4 12.3 18,6 19.0 35.0 Czech Republic 8.1 3.4 7.5 19.0 19.0 18.0 Greece 5.2 3.4 14.4 23.0 19.0 17.0 Hungary 6.1 3.4 4.9 14.4 19.0 19,0 Korea, Rep, 5.1 3.4 4.3 12,8 19.0 36.0 Malaysia 6.6 3.4 12.4 22.4 19.0 40.0 Mexico 0.9 3.4 12.5 16,8 19.0 19.0 South Africa 5.2 3.4 8.3 16.9 19.0 35.0 Trinidad 3.6 3.4 7.6 14.6 19.0 45.0 Uruguay 12.7 3.4 6.2 22.3 19.0 33.0 Lower Middle Income Bracket: Effective I January 2001 Colombia 5.1 4.8 8.6 18.5 19.0 38.0 Costa Rica 3.3 4.8 2.2 10.3 19.0 29.0 Dominican Republic 3.6 4.8 6.1 14.5 19.0 26.0 Ecuador 2.9 4.8 2.6 10.3 19.0 41.0 El Salvador 5.9 4.8 1.1 11.8 19.0 31,0 Guatemala 3.1 4.8 2.4 10.3 19.0 34.0 Indonesia 6.8 4.8 23.9 35.5 19.0 48.0 Jamaica 2.9 4.8 1.0 8.7 19.0 58.0 Jordan 15.9 4.8 2.3 23.0 19.0 50.0 Panama 4.7 4.8 9.9 19.4 19,0 45.0 Peru 5.8 4.8 5.5 16.1 19.0 33.0 Philippines 6.5 4.8 12.6 23.9 19.0 29.0 Poland 4.7 4.8 15.1 24.6 19.0 21.0 Russia 5.4 4.8 25.2 35.4 19.0 35.0 Thailand 4.0 4.8 8.3 17.1 19.0 35.0 Turkey 5,4 4.8 7.7 17.9 19.0 33.0 Venezuela 3.7 4.8 15,3 23.8 19.0 32.0 Lower Income Bracket: Effective I January 2002 China 8.7 4.8 4.2 17.7 23.0 58.0 Egypt 10.4 4.8 2.0 17.2 23.0 55.0 Guyana 6.6 4.8 0.6 12.0 23.0 85.0 Haiti 8.6 4.8 17,0 30.4 23.0 55.0 Honduras 3.1 4.8 8.7 16.6 23.0 97.0 India 8,1 4.8 18.3 31.2 23.0 64.0 Kenya 25.5 4.8 12.3 42.6 23.0 55.0 Nicaragua 3.8 4.8 3.7 12.3 23.0 43.0 Pakistan 14.7 4.8 7.2 26.7 23.0 1.11 Vietnam 9.3 4.8 10.6 24.7 23.0 78.0 Source: FCC 68 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 National Interconnection Rates Local Termination Regional Termination (US cents) (US cents) 1997 1998 1997 1998 Australia 1.62 2.15 5.30 2.18 Austria 2.00 1.90 2.00 1.90 Belgium 1.23 1.11 2.33 1.87 Canada n.a. n.a. 5.33 0.78 China n.a. 0.65 n.a. 0.81 Denmark 1.09 1.03 2,02 1.92 Finland 1.56 1.67 1.58 1.67 France 0.78 0.63 1.90 1.56 Germany 1.10 1.05 1.88 2.26 Greece 2.01 2.06 2.01 2.06 Ireland 2.44 1.08 4.61. 1.67 Italy 1.68 1.03 2.74 1.86 Japan 0.71 0.85 1.73 1.60 Luxembourg 2.23 2.34 2.23 2.34 Mexico n,a. n.a. n.a. 2.61 Netherlands 1.30 1.16 1.78 1.74 New Zealand n.a. 1.43 n.a. n.a. Norway n.a. 1.00 n.a. 1.38 Portugal 1.33 2.87 2,63 5.74 Spain 1.65 1.03 1.65 1.66 Sweden 1.27 1.12 1.96 1.75 Switzerland n.a. n.a. 2.72 1.50 U.K. 0.68 0.55 0.96 0.82 U.S. (Bell Atlantic) n.a. n.a. 0.79 1.46 National Termination (US cents) 1997 1998 9.90 4.00 2.63 2.50 3.26 2.67 n.a. n.a. n.a. 0.89 2.46 2.33 3.12 4.12 2.80 2.32 2.86 2.74 2.87 2.95 8.75 2.36 n.a. 2.69 5.98 5.55 2.23 2.34 n.a, B.a. 2.29 2.11 n.a. 6.66 n.a. 1.63 19.98 11.48 4.63 3,20 2.88 2.36 3.73 2.08 1.88 1.69 n,a, n.a. Notes: All interconnection charges are for peak period. Rates date to July and August 1999. Local termination is the lowest level of interconnection, typically giving a carrier access to a single town or part of a city. Regional termination generally gives a carrier access to all subscribers within a metropolitan area or a North American area code. China’s interconnection rates are for domestic operating companies only. Japan 1998 termination rates are estimated. U.S. termination fees vary according to Local Exchange Carrier (LEC). U.S. average for regional termination was 1.46€ as of August 1999. Source: European Union Commission, FCC, CRTC (Canada), MPT (Japan), 0FCOM (Switzerland), Telmex, China Telecom © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 Retail Prices for a Three Minute Call From]To Australia Belgium Canada Denmark Finland France Germany Greece Ireland Italy Australia n.a. 2.23 1.60 2.07 2.07 2.07 1.90 1.76 1.58 1.55 Austria peak 1.91 0.96 1.08 0.96 0.96 0.96 0.76 1.08 0.96 0.76 Austria off-peak 1.75 0.80 0.96 0.80 0.80 0.80 0.64 1.08 0.80 0.64 Belgium 3.12 n.a. 1.56 1.17 1.56 1.17 1,17 1.56 1.56 1.56 Czech Republic peak 2.01 1.27 1.93 1.27 1.31 1.27 1.21 1.31 1.31 1.27 Czech Republic off-peak 1.25 0.79 1.20 0.79 0.82 0.79 0.71 0.82 0.82 0.79 Denmark peak 2.15 1.24 1.21 n.a. 0.70 1.23 0.87 1.55 1.24 1.21 Denmark off-peak 1.65 0.95 0.83 n.a. 0.56 0.80 0.56 1.12 1.00 0.76 Finland 2.18 1.44 1.44 0.43 n.a. 1.44 1.25 1.44 1.44 1.44 France peak 1.62 0.66 0.70 0.70 0.80 n.a. 0.66 0.80 0.80 0.66 France off-peak 1.44 0.58 0.64 0.64 0.72 n.a. 0.58 0.72 0.72 0.58 Germany peak 3.00 0.66 0.66 0.66 1.35 0.66 n.a. 1.17 1.17 0.66 Germany off-peak 3.00 0,66 0.66 0.66 1.01 0.66 n.a. 1.01 1.01 0.66 Greece 1.36 1.36 1.36 1.36 1.36 1.36 1,36 n.a. 1.36 1.36 Ireland peak 2.25 1.00 0.76 1.26 1.26 1.00 1.00 1.26 n.a. 1.26 Ireland off-peak 1.80 0.89 0.67 1.10 1.10 0.89 0.89 1,10 n.a. 1.10 Italy 2.56 1.10 1.10 1.10 1.10 1.10 1.10 1.52 1.10 n.a. Japan peak 5.87 6.64 3.80 6.64 6.64 5.52 5.52 6.64 6.64 6.64 Japan off-peak 4.66 5.35 3.11 5.35 5.35 4.92 4.92 5.35 5.35 5.35 Korea, Rep. 2.75 2.28 3.59 2.28 2.28 2.28 2.28 2.28 2.28 2.28 Mexico peak 5.38 4.75 2.99 4.75 4.75 4.75 4.75 4.75 4.75 4.75 Mexico off-peak 3.59 3.17 2.99 3.17 3.17 3.17 3.17 3.17 3.17 3.17 Netherlands 1.15 0.36 0,36 0.41 0.72 0.40 0.33 1.02 0.77 0.52 Norway 0.79 2.21 0.57 0.40 0.40 0.57 0.48 0.87 0.67 0.63 Portugal peak 2.66 1.16 1.32 1.16 1.16 1.16 1.16 1.16 1.16 1.16 Portugal off-peak 2.66 1.16 1.32 1.16 1.16 1.16 1.16 1.16 1.16 1.16 Spain 3.40 1.16 1.94 1.16 1.16 1.16 1.16 1.16 1.16 1.16 Sweden 2.35 1.19 1.34 0.47 0.47 1.19 0.91 1.48 1.19 1.48 Turkey 3.53 1,37 2.06 1.37 1.37 1.37 1.37 0.88 1.37 1.37 U.K. peak 2.01 1.16 0,97 1.16 1.58 1.16 1.16 1.46 0,94 1.46 U.K. off-peak 1.72 1.09 0.92 1.09 1.42 1.09 1.09 1.21 0.81 1.2i U.S. (AT&T basic) 5.40 5.52 1.56 5.22 5.18 4.77 4.44 6.96 4.83 5.37 U.S. (AT&T One Rate) 0.75 0.87 0.30 0.87 0.87 0.66 0.60 0.87 0.66 0.66 U.S. (MCI basic) 5.51 5.66 2.20 5.30 5.39 4.85 4.55 7.10 4.91 5.45 U.S. (MCI One) 0.75 0.87 0.30 0.87 0.87 0.66 0.60 0.75 0.66 0.66 U.S. (Delta Three I P telephony) 0.42 0.48 0.20 0.51 0.51 0.42 0.42 1.20 0.51 0.42 Notes: 1. All rates are in US$ and exclusive of taxes and were current on August 31, 1999. 2. Rates have been calculated in real time using meter step (rounded up to next meter step for a 3 minute call). 3. Fees are $1 with domestic long-distance per month for AT&T One Rate and $3 per month for MCI One. 4. Rams for calls from the U.S. to Canada and Mexico are from Washington, D.C. to Toronto and Mexico City. © TeleGeography, Inc. 1999 TeleGeography 2000 Japan Korea, Rep. Mexico Neth’lands Portugal Spain Sweden Turkey U.K. U.S. 2.19 2.00 3.23 2.01 2.91 2.54 2.25 2.45 1.58 1.60 2.39 6.22 3.67 0.96 1.08 0.96 0.96 1.08 0.96 1.08 2.23 5.74 3.19 0.80 0.96 0.80 0.80 1.08 0.80 0.96 3.12 4.29 5.46 1.17 1.56 1.56 1.56 2.34 1.17 1.56 2.01 4.85 4.01 1.27 1.31 1.31 1.27 1.81 1.26 1.91 1.25 3.75 3.11 0.79 0.82 0.82 0.79 1.41 0.74 1.18 3.00 4.96 4.96 1.14 1.55 1.55 0.58 1.89 0.87 2.40 3.94 3.94 0.90 1.12 1.12 0.46 1.31 0.65 3.63 4.80 4.80 1.44 1.44 1.44 0.43 1.53 1.44 1.62 1.79 2.08 0.70 0.80 0.66 0.70 1.06 0.66 1.44 1.60 1.88 0.64 0.72 0.58 0.64 0.96 0.58 3.00 3.00 4.17 0.66 1.35 0.66 1.35 1.17 0.66 3.00 3.00 4.17 0.66 1.01 0.66 1.01 1.01 0.66 1.55 1.55 2.43 1.36 1.36 1.36 1.36 1.21 1,36 2.25 3,22 2.01 1.00 1.26 1.26 1.26 2.42 0.53 1.80 3.22 1.81 0.89 1.10 1.10 1.10 2.09 0.50 2.54 3.52 3.52 1.10 1.10 1.10 1.10 2.23 1.10 n.a. 4.06 6.30 6.64 6.64 6.64 6.64 6.64 4.75 n.a. 3.36 5.09 5.35 5.35 5.35 5.35 5.35 4.14 2.04 n.a. 3.73 2.28 2.28 2.28 2.28 2.28 2.28 5.38 5.38 n.a. 4.75 4.75 4.75 4.75 4.75 4.75 3.59 3.59 n.a. 3.17 3.17 3.17 3.17 3.17 3.17 1.75 2.18 2.92 n.a. 1.02 0.52 0.41 1.39 0.30 1.18 1.78 2.21 0.45 0.87 0.67 0.39 1.41 0.45 3.46 3.46 3.62 1.16 n.a. 1.00 1.16 2.03 1.16 3.46 3.46 3.62 1.16 n.a. 1.00 1.16 2.03 1.16 3.40 3.40 3.18 1.16 1.16 n.a. 1.16 1.80 1.16 2.93 5.24 5.82 0.91 1.48 1.48 n.a. 2.06 0.91 3.53 3.53 2.06 1.37 1.37 1.37 1.37 n.a. 1.37 2.76 4.43 4.43 1.16 1.46 1.46 1.16 2.76 n.a. 2.35 4.21 4.21 1.09 1.21 1.21 1.09 2.35 n.a. 5.19 6.48 5.07 4.71 5.88 5.79 4.71 6.57 3.87 0.78 1.05 1.26 0.75 0.75 0.75 0.60 1.41 0.30 5.27 6.65 1.73 4.79 6.02 5.87 4.79 6.71 3.95 1.05 1.95 1.26 0.75 0.75 0.75 0.60 1.86 0.30 0.54 0.42 1.11 0.45 0.99 0.42 0.39 1.35 0.42 To/From Australia Austria peak Austria off-peak Belgium Czech Republic peak Czech Republic off-peak 1.21 Denmark peak 0.92 Denmark off-peak 1.44 Finland 0.70 France peak 0.64 France off-peak 0.66 Germany peak 0.66 Germany off-peak 1.36 Greece 0.76 Ireland peak 0.67 Ireland off-peak 1.10 Italy 2.07 Japan peak 1.81 Japan off-peak 1.74 Korea, Rep. 2.65 Mexico peak 2.65 Mexico off-peak 0.27 Netherlands 0.57 Norway 1.24 Portugal peak 1.24 Portugal off-peak 1.33 Spain 0.91 Sweden 2.06 Turkey 0.97 U.K. peak 0.92 U.K. off-peak n.a. U.S. (AT&T basic) n.a. U.S. (AT&T One Rate) n.a. U.S. (MCI basic) n.a U.S. (MCI One) n.a. U.S. (Delta Three IPtelephony) Source: The Phillips Group, 3rd Floor, 19 Thomas More St., London E1 9YVV, U.K. Tel +44 20 7423 4500 Fax +44 20 7423 4501 ° Email: consult@tarifica.com ¯ wvwv.tarifica.com Source for U.S. rates: TeleGeography, Inc. © The Phillips Group and TeleGeography, Inc. 1999 71 TeleGeography 2000 © TeleGeography, Inc. 1999 Retail Pricing Trends, 1997-1999 From Asia $8.00 $6.00 from Australia from Japan $8.00 $8.00 $6.00 $6.00 from Rep of Korea $4.00 $2.00 $0.00 n.a. Australia Czech Germany Japan U.K. U.S. Rep. From Americaa from Canada $8.00 $4,00 $2.00 $0.00 Australia Czech Germany Japan U.K. U.S. Rep. $4.00 $2.00 $0.00 Australia Czech Germany Japan U.K. U.S. Rep. $8.00 from Mexico $8.00 from U.S. $6.00 $4,00 Au~-elia Cza~h Germany Japan U.K. U.$. $6.00 $4.00 $2.00 $0.00 Australia Czech Germany Japan U.K. U.S. Rep. From Europe $8.00 from France $8.00 from italy $6.00 $4.00 $2.00 $0.00 n.a. Australia Czech Germany Japan U.K. U.S. Rep. $8.00 from UK. $6.00 $4.00 ~.00 Australia Czech Germany Japan U.K. U.S. Australia Czech Germany Japan U.K. U.S. Rep. Rep. Key ¯ [] [] 1997 1998 1999 Rate for a 3 minute call to destination $6.00 $4.00 $2.00 $o.oo $6.00 $4,00 $2.00 $0,00 Australia Czech Germany Japan U.K. U.S, Rep. © Te!eGe~g~a p~r !nc~ ! 998 © TeleGeography, Inc. 1999 TeleGeography 2000 Origin of U.S. Carrier Revenues Although useful for tracking pricing trends, advertised retail tar- iffs only rarely match the per minute revenues that carriers actually collect. Billed revenues often fall far short of basic rates--a result of special calling plans offered to end users and of bulk discounts awarded to service resellers (see Figure 1). Nearly half of international voice service revenue collected from the customers of U.S. carriers flow to foreign telephone compa- nies in the form of settlement payments (see Figure 2). In 1998, the average price of an international call billed by the three largest U.S. carriers was 63¢ per minute--of which 31 ¢ were passed on to foreign carriers (see Figures 3 and 4). Because they send out far more international calls than they receive, U.S. carriers continue to incur a multi-billion dollar net settlements deficit. Yet thanks to falling settlement rates, gross settlement outpayments from the U.S. decreased in 1997 for the first time in twenty years. Rate per Minute $1.75 $1.50 $125 $1.00 $0.75 $0.25 [] AT&T Basic Peak Rate [] Actual AT&T Revenue per Minute [] AT&T One Rate Discount Calling Plan Australia Germany Japan U.K. $2.00 $12 Total Billed Revenues $6.0o $6 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 31 U!Si Carrier International Call Revenue by Destination, 1998 ...... Average U.S. Outgoing Call Revenue by Region $1.50 ~Settleme~ Payments ~:;:;~ Retained Revenue $1.25 $1.oo $0.75 $0.50 $0.25 $0.00 Average U.S. Outgoing Call Revenue by Country $1,50 ........................... ¯ Settlement Payments Retained Revenue $1.25 $1.00 $0.75 $0.25 $0.00 international Outgoing ca!IS for the four |argest U~Si international carriers Tota CO umn height sh0ws the average price foe ;;~al|s on a g!ven;~0~ ;Retai~d;re~e~Ue equ~iS average p~ice nuS ~ettlement p~ment, ~nd nC!udes s~h €omponents a~ ~ces~ ~s for 0rigination network €~sts AT&T 6,723.4 2,975.1 3,748.3 980.3 4,728.6 MCI WorldCom 4,292.2 2,307.4 1,984.8 748.8 2,733.6 Sprint 1,194.0 716.1 477.8 236.3 714.2 Top 3 Tota/ 12,209.6 5,998.7 6,211.0 1,965.4 8,176.4 Total Receipts (US$ millions) Net Revenue Average Revenue per Minute (US$/minute) Billed Settlement Retained Settlement Revenue Outpayment Revenue Inpayment 0.68 0.30 0.38 0.21 0.60 0.32 0.28 0.25 0.54 0.33 0.22 0.29 0.63 0.31 0.32 0.23 © TeleGeography, Inc. 1999 TeleGeography 2000 Trading International Telecom Minutes Lowest Matched Sell Rates (U.S. cents per minute) Top Routes Ranked by Minute Volumes Traded Q1 1999 Q2 1999 Q3 1999 Volumes Traded on Arbinet 1998 U.S. Outgoing PSTN Traffic U.S.-Americas Rank Destination Rank Minutes Brazil 22.0 18.9 14.7 1. China 14. 420.3 Mexico 16.4 14.3 13.0 2. Vietnam 28. 190.4 U.S.-Asia 3. Brazil 8. 581.3 China 30.1 21.5 18.3 4. Philippines 10. 552.6 Hong Kong 5.5 3.3 3.4 5. Japan 5. 796.1 Japan 7.1 6.2 6.1 6. Lebanon 67. 56.6 U.S.-Europe 7. Germany 4. 1,267.7 France 5.7 3.8 3.4 8. Italy 11. 510.4 Germany 4.8 4.1 3.7 9. Australia 12. 444.1 Greece 17.1 14.5 13.9 10. United Kingdom 3. 1,454.5 (m) Note: R~tes show the iow~t p~iC~ fo~ S~cC~ssfui trades am~n~ members of Arbinet’s ~x~hange during th~ glvsn quarter; Ro~ ~ankings reflect UiS. oUtbbund minutes ~999i .............................. TeleGeography 2000 © TeleGeography, Inc. 1999 International Call Quality Metrics Once ruled by monopoly PTOs, global telephony has become a competitive playing field. That makes phone calls cheaper. But price is iust one determinant of purchase behavior. Another is call quality. Traditionally, telephony Quality of Service (QoS) has been used to plan and manage the transmission system. Today, QoS is being built into Service Level Agreements (SLAs) that help formalize transactions in a telephony marketplace where, confronted with an increasing array of options, many wholesale and retailer buyers would rather focus on products than ingredients. Agreeing on pricing is easy; quality, on the other hand, is subjective. So how can QoS be quantified? A Metric Review Monnet UK Ltd., an independent QoS arbiter, is implementing one approach. Monnet monitors call quality on its clients’ net- works. It also constructs industry benchmarks, pooled from data provided by participating carriers. Figure 1 shows survey results for 45 destination countries, based on a sample of one million international calls from German and U.K. carders between June 1 and August 31, 1999. Monnet uses three indicators to measure call quality: ¯ Answer Seizure Ratio. ASR measures the percentage of successful call attempts made between a switch and a given destination. A 50 percent ASR means that only one half of all call attempts were answered by a person or device; an unanswered call or busy signal counts as an unsuccessful call. That means ASR is affected not only by performance factors--availability of dial tone and the network’s ability to establish a transmission path or switch a call--but also by phenomena ranging from a changed dialing code to holidays and sporting events, all of which might lead to more unan- swered calls. For calls to developed economies, 60 to 75 percent ASR is viewed as an acceptable range. ¯ Post Dial Delay. PDD measures the time it takes a net- work to establish a connection once the caller has finishing dialing. Hence, a PDD of 8.2 means that an average of 8.2 seconds elapse between the dial and the ring at the other end. ¯ Call Quality Index. CQI consists of a basket of five qua!- ity factors, and is expressed on a scale of 0 to 100. Based on a technical model provided in ITU-T Recommendation G. 107 (www.itu.int/’~udoc~-tlrec/g/g100-699/s_g107.htm), these factors include signal level, noise, echo path loss, echo path delay, and speech activity. A CQI score of 90-100 translates to a "best" ranking; a CQI of 50-60 is graded "poor." Beside reflecting basic infrastructure problems, CQI scores can decrease as a result of packet loss in IP networks, use of compression, and including many mobile calls in the traffic mix. Toward a Better Benchmark Monnet’s benchmark provides carriers with an objective point of comparison for their OoS. But industry-wide comparisons across different originating countries can be deceiving. In the survey summarized in Figure 1, for example, refiled traffic--the reorigination of calls in a hub country--was a key factor in determining call quality. Many of the U.K. carriers participat- ing in the survey acted as refile hubs for several other countries. Those countries’ network quality proved inferior to the carriers’ own networks, and as a result, a number of destinations appear to have lower call quality metrics from the U.K. than from Germany. Mobile traffic can dilute call quality: when a call transits a mobile network, a number of the characteristics which drive down Call (3uality Index scores--noise, echo, delay--tend to show up. The economics of sending calls to mobiles can also affect quality metrics. Interconnect fees to mobile networks are unusually high in many countries (see "International Traffic on Mobile Networks," pages 153 to 157), leading carriers termi- nating traffic to those countries to answer incoming calls des- tined for mobile phones with a busy signal. Such intervention artificially lowers ASR and raises ¢(31. Used inappropriately, call quality metrics can prove to be an inexact science. The figures shown here represent a snapshot summarizing quality from all types of international calls. A more sophisticated methodology would control for external fac- tors by segmenting call samples based on type of international service--fixed or mobile, originating or terminating leg of a refile arrangement, and so on. Given the bewildering range of new options for sending and terminating calls, we can expect companies such as Monnet to introduce such detailed analyti- cal tools soon. Just as ever-greater numbers of carriers have spurred demand for independent bandwidth exchanges, QoS metrics are taking on a larger role, too, acting as intermediaries that keep intercarrier relationships manageable. By quantifying service levels, benchmarks allow carriers to incorporate quality metrics into SLAs that are both verifiable and legally binding. Alongside price, quality has become part of the equation. © TeleGeography, Inc. 1999 TeleGeography 2000 Answer Seizure Ratio Post Dial Delay (seconds) Call Quality Index Destination from Germany from U.K. from Germany from U.K. from Germany from U.K. Australia 70% 47% 4.2 14.9 89.2 69.0 Austria 76% 39% 2.0 8.3 83.0 49.8 Belgium 68% 41% 2.0 17.7 77.1 59.4 Brazil 68% 47% 6,3 3.6 87.6 44.2 Canada 71% 75% n.a. 1.5 n.a. 63.2 Chile 57% 17% 2.8 4.2 73.7 69.7 China 60% 45% 3.6 5.4 85.7 52.3 Colombia 52% 12% 9.2 4.5 80.5 49.1 Denmark 76% 68% n.a. 2.1 n.a. 64.7 Ecuador 32% 15% 1.4 16.7 74.8 48.1 Finland 69% 66% 8.2 3.4 76.1 67.5 France 59% 55% 1.7 3.6 80.6 65.0 Germany n.a. 32% n.a. 4.6 n.a. 64.9 Ghana 33% 36% 6.6 6.7 69.9 55.3 Greece 56% 51% 3.9 3.1 73.9 43.7 Hong Kong 56% 67% 7.9 6.2 85.0 68.1 India 67% 33% 2.4 10.5 85.5 39.4 Ireland 61% 65% 6.4 1.1 62.5 67.6 Israel 71% 50% n.a. 6.6 60.8 40.8 Italy 63% 46% 1.9 4.1 67.6 40.8 Japan 63% 67% 5.9 4.9 81.7 67.9 Korea, Rep. 64% 52% 8.6 4.1 89.3 80.1 Kuwait n.a. 22% n.a. 17.9 80.3 42,2 Macedonia 42% 40% 7.0 6.8 81.1 43.9 Malaysia 32% 58% 5.2 3.6 83.0 63.2 Mexico 37% 43% 2.9 7.5 82.6 80.2 Netherlands 76% 40% 1.7 8.2 82.3 33.4 Norway 69% 66% 6.7 3.2 83.7 64.3 Pakistan 55% 20% n.a. 5.4 n.a. 40.7 Peru 79% 38% 2.2 13.0 90.0 49.6 Philippines 48% 42% 3.0 2.9 82.2 41.7 Poland 51% 51% 3.1 4.7 77.1 57.5 Romania 69% 28% 2.0 11.1 79.2 37.7 Russia 45% 35% 3.0 5.6 86.1 41.8 Saudi Arabia 72% 50% 7.6 7.6 n.a. 81.1 Singapore 70% 49% 2.8 4.0 89.3 42.0 South Africa 66% 55% 4.8 4.7 55.8 69.3 Spain 70% 48% 2.3 3.5 68.4 52.5 Sweden n.a. 63% n.a. 3.9 n.a. 62.3 Switzerland 61% 51% 2.0 3.7 83.9 62.2 Taiwan n.a. 52% n.a. 3.0 n.a. 53.2 Turkey 57% 20% 2.1 5.1 91.8 38.0 UAE 54% 42% 7.8 2.9 64.4 60.9 United Kingdom 62% n.a. 1.7 n.a. 75.6 n.a. United States 71% 69% 6.1 2.9 86.8 73.4 World Average 58.0% 42.9% 3.4 4.7 75.6 60.0 TeleGeography 2000 © TeleGeography, Inc. 1999 Facilities TeleGeography 2000 © TeleGeography, Inc. 1999 The Bandwidth Wagon Rolls On Four years ago, a true market for international bandwidth-- "raw" telecommunications capacity--barely existed. For the most part, specific bandwidth products (for example, a 2 Mbps circuit between Milan and Frankfurt) were supplied as two halves by national incumbent monopoly telcos (in this case Telecom Italia and France Telecom). Prices were high, and ser- vice was often indifferent. Most of the buyers were multina- tional companies such as American Express and Shell, with a sprinkling of value added network service companies like SITA--the airlines’ network--and a few newly emerging ISPs. These bandwidth buyers needed skills more usually to be found in minor foreign embassy officials--infinite patience, subtle negotiating powers, and good personal contacts inside supply companies. Negotiations focused on ensuring that a product was delivered on time, that it worked, and that when it went down (as it all too frequently did) it came back up as quickly as possible. Negotiations rarely focused on price, since the price-- generally far above underlying cost--was fixed by the monop- oly owners. That was then, and this is now. An extraordinary and radical makeover of the sector is in progress, and the bandwidth mar- ket is rapidly becoming one of the most dynamic segments of the industry, at least in those regions where competition has been permitted. The number of suppliers is increasing rapidly, new products are emerging monthly, and prices are falling faster than in almost any other sector of the market. As Figure 1 shows, competition is now established in western Europe and the Atlantic Ocean region. In other key regions, it’s coming soon, and by 2002, few areas will still be without alter- native suppliers of bandwidth. Why the sudden upsurge in interest? Four primary factors are at work, each reinforcing the others to create an explosive upward momentum. Competitive positioning: Competition is the key to the cur- rent changes. The U.S. and the European Union effectively freed up the trans-Atlantic and western European bandwidth markets in 1996, and many new providers see this as a once-in-a-generation opportunity to dominate a key emerg- ing market. Others want to lay infrastructure or buy fibers in order to lower unit costs for other higher value or retail products. These perceptions have led to a boom in plans for new bandwidth, especially in western Europe and between Europe and the U.S. Figure 2 lists key projects in western Europe. New technology: Dense Wavelength Division Multiplexing (DWDM) is the most disruptive technology to hit telecom- munications transmission since the first communications Highly competitive Highly or fairly competitive Fairly competitive Limited competition Very limited competition Almost no competition No competition ¯ 1999 State of Competition [] 2001 State of Competition Source: TeleGeography, In €: © TeleGeography, Inc. 1999 TeleGeography 2000 Organization Coverage Description of Commercial Infrastructure Launch Capacity Strategy BT 47 cities said to be Largely based on the Initial network covered;exceptionally domestic fiber networks completed Q2 dense coverage in partner of partners, e.g. C~g~tel 1999 countries including m France Germany, France and Italy Carrier1 14 cities; 12 more in Building its own network 1998 Germany and 12 in France in Germanywith MFN. ~lanned Elsewhere leasing dark fiber 160 Gbps lit I nta y No plans to sell bandwidth to third parties Carriers’ carrier which focuses on switched minutes and packets COLT 18 city MANs will be All-new network with End 2000 connected by long multiple ducts to ease distance network being upgrade to new fiber jointly created with Level types 3 B.a. Full service telco focused on corporate business. Unique MAN play Global Crossing 12 cities by December Pan European Crossing End 1999 1999; a further 6 in the first will be comprised of quarter of 2000, and 27 in 11,000 km multiple self- total by the end of 2000 healing rings connecting 24 cities 10 Tbps theoretical capacity Wilt sell full range of bandwidth products, positioning itself as a global provider GTS Carrier Services 30 cities covered, 70 percent leased dark 1996 (formerly Hermes) including many in fiber; own cables and southern Europe ducts especially on border crossings Upto 160 Gbps, depending on routes; 2-18 wavelengths lit First mover in many locations, e.g. southern Europe; full range of bandwidth products iaxis (formerly 18 cities so far; Based primarily on dark May 1999 Telemonde) aggressively looking to fiber leased from utilities new locations, especially in southern Europe Upto 10 Gbps per wavelength; up to 192 wavelengths per link Wholesale bandwidth only, plus value-added products such as telehousing Interoute (i-21) Aiming to cover 48 cities, All-new network based May 2000 Up to 1,000 Tbps with dense national i on Coming G.655 LEAF in theory; 144 coverage in major I fiber and 144 fiber pairs fiber pairs countries ~er duct Seeking dark fiber sales initially; full range of bandwidth products and corporate services likely Level 3 Initial network will cover 8 All-new network with Sep 2000 n.a. cities with long distance multiple ducts to ease network, plus 5 MANs upgrade to new fiber types Will seekto sell ducts and dark fiber initially; bandwidth focus will be followed by shift to corporate service portfolio KPNQwest Six Euro-rings being built, Largely new cabled 0.2 1999 n.a. with initial coverage of networks with high fiber major cities in France, count and multiple ducts Germany, UK and Benelux; 40 cities will be covered at completion in 2001; 14 cities live in 1999 Selling dark fiber and a full range of bandwidth products; strategic objectives include shift towards corporate value added services MCI WorldCom 13 cities; denser national Mix of mainly new cable Q3 1998 20 Gbps coverage planned in and some dark fiber major countries. Owns MANs in some cities Built network primarily to meet own needs, but does sell all bandwidth products MFN Primarily focused on All-new network based 1999 n.a. Mainly selling dark fiber MANs, but will build on very high fiber count regional networks in (up to 840 fiber pairs) Germany and possibly other territories Netsgroup Nine cities in France, Early entrant building Q1 1999 n.a. n.a. Benelux, Germany and UK low-cost dark fiber already covered; a further network; pure carriers’ four in Switzerland, Italy carrier and Spain byyear end TeleGeography 2000 © TeleGeography, Inc. 1999 F gure 2, Western Europes Fiber Frenzy (cont=nued) Organization Coverage Description of Commercial Pangea Teleglobe Telia Viatel Ten cities in northern Europe Initial ring connecting three cities in October 1999; further ring will connect Germany and France in December 1999; third ring to other cities in these countries plus Austria and Switzerland in 2000 Scandinavian ring and European expansion now under construction. Network will initially cover nine cities, mainly ~n northern Europe 13 cities in first phase; 12 more second-tier German and French cities in second phase complete Infrastructure Mainly submarine cable, some terrestrial infrastructure Primarily dark fiber and fiber swaps; transoceanic cables planned Network will comprise primarily of own laid cable, both submarine and terrestrial. Some dark fiber Mainly new cable; German network a joint build with MFN Launch Q4 2000 Q1 2000 2000 Q2 1999 Capacity Initial capacity 160 Gbps 160Gbps Strategy Selling primarily IRUs and long leases, city to city, from STM1 upwards Full range of bandwidth products; Teleglobe reorganizing as a global carriers’ carrier !company Focusing on wholesale IP and IP services Selling full range of bandwidth products; first to sell terrestrial IRUs munications infrastructure in 1996, creating a momentum that almost entirely on dark fiber leased largely from utilities, and continues to build three years later, and making this region the then invested heavily in DWDM equipment. Others like Global most dynamic bandwidth market in the world. products, or both. ...... both Only a handful stit say they are only interested in being ’Golden Tdang e’, which ]inks London Amsterdam, Brussels ’ wholesalers; retai ing provides a second source or revenue, as Frankfurt and Paris. However, a widespread fear that this re~ion country networks, reasoning that domestic telecommunications .... telehousing Web hosting and applications service provision. traffic will always exceed cross-oorder tramc © TeleGeography, Inc, 1999 satellites were launched thirty or more years ago. Its most obvious impact is on capacity: transatlantic cables to be installed in 2001 will have a design capacity of 2.5 Tbps, a 30-fold increase over Atlantic Crossing-l, which went into service in 1998. Temporarily, at least, the year on year improvement in the capacity of optical fibers is easily outpacing year on year improvements in computer memory, storage space or pro- cessing power. But this new abundance is also affecting many other aspects of the business, for example by making satellite bandwidth on some routes uneconomical Most satellite system owners are doing well in bandwidth because of the overwhelming demand for capacity from ISPs. But that role will likely become restricted to regions that do not benefit from the bandwidth boom. DWDM is also creating new product sets and business strategies. New demands: the liberalization of telecommunications markets has led to a sur~ng market for wholesale or card- ers’ carrier products; in the European region, the number of licensed international carders has grown to almost 600 this year. Many do not own facilities, and most need bandwidth products of some kind, whether it be dark fiber, STM 16 IRUs or annual STM1 leases. On top of that, demand generated by the continuing global boom in mobile communications and by the lnternet and corporate data traffic are contribut- ing to a much more rapid growth in the utilization of inter- national bandwidth, and especially an increasing volatility in demand. © TeleGeography, Inc. 1999 TeleGeo{~raphy 2000 Element Seller’s Market Buyer’s Market Length of contract One, three, or five years One to 25 years; shorter leases including bandwidth on demand are also in prospect Discounting Generally no more than 5 percent, Steep discounts on volume and contract length; no published mainly for length of contract prices, so everything is negotiable Regular Price Did not exist Becoming widely negotiated by major buyers, often linked to an Reviews in Longer industry index or some other independent measure Contracts Forward Pricing Did not exist Beginning to be used by some suppliers (e.g. in a ten year contract, the average yearly price is based on the market price four years’ hence) Options on Future Did not exist Under consideration by some suppliers Bandwidth at an Agreed Price Network-based Did not exist (only point to point Becoming widely available. For example, some suppliers offer Pricing, or Flexibility pricing was available) to price on the basis of the whole network, with no penalties for on End Points shifting bandwidth to different locations Flexibility on None (most contracts called for Increasing; on longer ownership or IRU contracts, cash-starved Payment Terms payment in advance yearly or buyers can often delay part of the payment quarterly) Easy money: on the supply side, meanwhile, the huge inter- est in telecommunications generated by the lntemet has seen investment houses chasing after opportunities to take a position in the international market. That’s made it easy for new entrants to raise money, resulting in a major con- struction boom (nearly all the entrants are funded through venture capital, high yield bonds and IPOs). Boom or Bust? For buyers of bandwidth, all of this activity has created a bewil- dering landscape in which the wrong decision can fatally under- mine carefully constructed business cases. In the past three years, for example, the new availability of bandwidth on long so-called ’capitalized’ leases or on an Indefeasible Right of Use (IRU) basis has tempted some smaller players to take long-term positions in bandwidth that depend upon assumptions which have not stood the test of time. On the face of it, long leases are very attractive; often such leases have paid for themselves (when compared to conventional ]eases) in 2-3 years. And long ]eases appear as an asset on balance sheets, improving EBIT- DA--a crucial consideration for indebted new carriers. The new Gemini and Atlantic Crossing-1 cables across the Atlantic made it possible for anyone to buy IRUs for the first time, and a handful of companies therefore took aggressive positions. Rather than buying to meet immediate demand, as nearly all bandwidth pumhasers had until then, they bought speculatively against an unproven end user demand that was still years away. Consequently when prices fell far faster than almost anyone anticipated, their cost base looked unsustain- able. An STM 1 IRU selling for $7-8 million in 1998 sells a year later for around $2-3 million, driving down the price of con- ventional leased circuits---a calamitous outcome for those who bought a lot of IRU capacity at the higher price, because these companies end up saddled with nominal costs higher than those who are leasing conventionally--often a direct competi- tor. That’s not the only dilemma for pumhasers: creative pricing and contracting for bandwidth products is becoming widespread wherever competition has flourished. Figure 3 lists some of the options now becoming available; buyers must learn to analyze not only the deals they are actually offered, but the deals that might be on offer if they ask the right questions. Ideally (from a buyer’s viewpoint), this new volatility would have created a genuine commodity market in bandwidth, since bandwidth is after all one of the most undifferentiated products in telecommunications. In reality, though, market commoditiza- tion is taking longer than people expected, and may only impact a small part of the market in the next few years. Many of the key requirements--regulated, large-scale exchanges; for- TeleGeol~raphy 2000 © TeleGeography, Inc. 1999 Backhaul Local Tail Submarine Cable Cable Landing Station Telehouse/ NAP Customer mal, standardized contracts; international quality measures; and sufficient liquidity--are still missing. For the time being, buyers of bandwidth will be on their own, and must deal with a new set of circumstances in which the risk of making a serious error is as tangible as the advantage offered by the new oppor- tunities of a buyers’ market. The Downside of the Upside The new bandwidth markets have thrown up a set of new chal- lenges, some of which are only just being resolved by suppliers and regulators. We’re on the beach--now what? In the monopoly era, cable stations at which submarine cables came ashore were operated by the national incumbent telco in the country, and backhaul-- the line from the station to the nearest metropolis--wasn’t real- ly an issue. Now it is, but regulators in some cases are strug- gling to catch up. The so-called "club" cables in some cases have almost 100 owners, all of whom need backhaul and space at the cable station. That has provided a new opportunity for incumbents, and few are resisting it; in some cases, annual backhaul costs are higher than the cost of the transoceanic cir- cuit. Where shall we put that switch? In a competitive telecommu- nications market, a semi-formalized system is required to exchange traffic and connect networks at an agreed-upon loca- tion, and telehouses, a.k.a, carrier hotels or co-location facili- ties, have come into existence to meet the need. However, the extraordinary growth in new carriers has created enormous pressure on existing facilities and led to a big increase in space in cities like London and (prospectively) Amsterdam, Frankfurt and Paris. In more peripheral and second-tier cities, new carri- ers may be forced to create their own space and lease very expensive local access lines to connect to other carriers in the city. Or they may have to turn to their bandwidth provider, which is likely to have tied up some telehousing space for its own use in the city. Outside the biggest business metropolises, telehousing will continue to present a problem for bandwidth purchasers for a while yet. Forecasting Demand: Outlook Foggy When telephony dominated telecommunications, traffic fore- casting was relatively straightforward. Historical data provided a fairly reliable guide to future demand, and the sources and sinks of traffic were predictably linked to populations, wealth and commercial activity. Not so in an IP era. There are at least three new variables, which, taken together, make forecasting next year’s international telecommunications demand almost as error-prone as forecasting next week’s weather. What will we be carrying? Before the Inter-net and the Web, networks were designed for specific applications. But the Internet is a huge, unpredictable, global machine for generating new applications, and few of those applications have a deter- minate bandwidth. Where will we be carrying it? Telephone traffic is closely corre- lated to populations and geography, since it usually connects two people using permanently located telephones on a point to point link. Twenty-first century traffic is different: often people have no idea where they are connected to, and it’s relatively easy to shift those connections--Web sites, say--to other places. That’s just one uncertainty among many: for example, new carriers are often not clear about which cities they want to connect next; mobile users are not in a fixed place; Internet traffic may soon include a significant proportion of multicasting; and caching and other bandwidth-saving technologies may unexpectedly reduce demand. What about the access network? Third, intemational demand depends on what end users pump into the network. That in turn depends on whether their networks are upgraded to ’higher’ speeds, and that’s moot today. While xDSL and cable modems are now relatively mature technologies, predicting when they’ll be used depends on a close understanding of local regulation, © TeleGeography, Inc. 1999 TeleGeography 2000 the attitudes of incumbents and cable TV companies, and com- mercial decisions by providers. That’s only one local variable: forecasters must also factor in the potential of new appliances to generate demand (wireless devices, info-appliances), grow- ing device to device (non-human) conversations, and unex- pected commercial decisions (e.g., the sudden emergence of "free" lnternet services in Europe). Why does this forecasting uncertainty matter? Because demand, in part, determines price, and buyers must have an opinion about likely price trends in order to make an informed deal with suppliers. How real is that asset? Many new carriers have opted to buy bandwidth on long leases or IRUs, partly to reduce nominal annualized cost of ownership, but also because, under most accounting conventions, leases of ten years or longer generally qualify as assets on a balance sheet instead of liabilities. Where bandwidth is a maior input, as it often is for new carriers, that can substantially improve EBITDA--almost irresistible for high- ly indebted new carriers, especially if an IPO is in the offing. But the fashion for long leases raises several difficulties. One noted by several cynical observers is that the exact same bandwidth may appear as an asset on the balance sheets of two or even three companies as it is sold down the line--an accounting sleight of hand that may not survive a detailed audit. Moreover, leases and IRUs may be written off over periods which look more and more unrealistic. While IRUs are nominally sold over a 25 year period, few in the industry believe that a specific asset has a commercial life longer than 5-7 years. What hap- pens when convention meets reality? Will long leases drop out of favor?. Conclusion For bandwidth buyers, opportunities and options are multiply- ing. What to do? In a forthcoming report, TeleGeography will examine all the new options, region by region, and attempt to guide the unwary through the increasingly tangled web of cables. The International Bandwidth Report will be published early in 2000. Please check our Web site for details. ~i~ Report Scheduled for release in early 2000, TeleGeography’s International Bandwidth Report will provide expert insight into the increasingly complex bandwidth mar- ket: clarifies supply and demand bandwidth issues: explains the bandwidth market and what drives it: describes how to be a wise bandwidth buyer; includes a directory of suppliers operating in the top business cities; and provides sample contracts. Bookmark the International Bandwidth Report web site at http:t/www.internatiom~tbandwidth.com to find out how to get a copy. TeleGeography 2000 © TeleGeography, Inc. 1999 Submarine Cable Systems Trans-Oceanic Capacity Trends 7,000 ¯Trans-Paciflc ¯ Europe-Africa-Asia 6,000 5,000 4,000 3,000 2,000 1,000 0 1995 1996 1997 1998 1999 2000 2001 Trans-Atlantic 23 23 Trans-Pacific 4 4 Europe-Africa-Asia 1 1 Total 28 28 Aggregate Capacity(Gbps) 1995 1996 1997 1998 1999 2000 2001 23 153 168 2,088 5,928 14 24 189 349 349 11 11 91 111 111 48 188 448 2,548 8,388 Note: Capacity figures denote lit capacity (Gbps) at the end of a par- ticular year. Projected capacity is based on reported initial capacity of systems during the year they are scheduled to be ready for ser- vice. Per-Fiber Capacity Trends More capacity 128 ,~ per-fiber ’9~8 /" Rapid developments in photonics have given a new dimension to 64 trans-oceanic capacity trends: upgradeability. Through bit rate 32 ’998 modulation and wavelength addition, systems can be upgraded by adding add~tmnal equipment (see Rgure 2), The first system to 16 -- ’96 ¯ ¯ make use of th=s capabdzty was TAT12/13, which upgraded its capacity from 10 Gbps to 20 Gbps in 1998. 4 In 1998 and 1999, a total of 105 Gbps of new capacity was added on 2 trans-Atlant~c routes, Of that, 60 Gbps came from upgrades alone. ’80 ’83 ’e5 ’86 ’87 ’91 ’95 Today’s upgradeable submarine cable systems can double their ¯ ¯ ¯ ¯ ¯ ¯ ¯ capacity in half the time it takes to lay a new system. 10 Mbps 100 Mbps I Gbps 10 Gbps 100 Gbps Channel Bit Rate ¯ Year of Commercial Deployment TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 2. Components Of a Submarine Cable System Wet Plant Repeaters Landing Station ~e wet plant;, ;as the name ir~plies,"is thepart ~;rthe cable ~ystem that is~under Wa~er. The .most bas~c..efements of-.we~: p|ant-are cables and.repea"ters. "T~e cable encases and pro- tects~ptica[ffiber that is.laid on, theocean floor .fi~;e miles from shore. Beca~e.!ightcan on.fy ~ave~ a certain distance before the signal qua.lity~degrade~,.~epeaters am~tical for ampiify- .ing .the signa! .on long,haul, trans-ocea.nic sys~;ems. consequently repeaters, need .to.be spaced stra~egfca!ly.a!ong the cable depending on the length of a pa~cu!a~ segment,. Tt~.e~drY plaitt ls essentially the’fan.cling station,. ~r the pta~e ,where, the cable reaches ShO~, The ]an~ing s~atiO~.houses the i~cess~ry equipment tO terminate .the.sig~al anddeitver ..to WDM Terminating Equipment Other Equipment Dry Plant SDH/SONET Equipment To Terrestrial System a ~terre~al.~ys~erri. Tt~ewet plant elements Predetemline the maximum possible capacity of the system and do not.change after~ deployme~, ¯ The dry.p!ant.e]~me~ts, however, a~ upgradeab!e..Th~dtfference between existing arid maximum ~capadty is called the upgradeabietatio.. ,.. :~-. Tile ,l:tgrad~ble:x;~tio. comp~. (:he system’s curi~t =pa~.,. to )~ max(mum.poss)bl.e .capac(ty ~th ~ry plant ~pgrades: ¯ .For example, a ,80~640 ratio means that:the wet plant .can handle 640 :Gbps, .but! Only 80 Gbps is use~t: ’ In most cases. the~ uPg~adeable rat10 of. a. system .Is .notI publidze~.i Fu.rthermore,..new developmerLts.in,photoni.c technology may! .~.: enable anp]anned ~pgrades at a later da~;o i ¯ :i o ¯ ° ¯ .,,~ ~’ :,,,,,,; ,, ,~ ¯ ¯ ,. TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 3. Map of Major Trans-oceanic Submarine Cables © TeteGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Key Major International Submarine Cables Capacity (Gbps) 2,560 1,280 640 320 120 20 10 16,384 128 Capacity (STM-ls) Planned Capacity is measured in STM-ls (155,520 Mbps), the unit of measure commonly used in the purchase of cable transmission rights or IRU’s (Indefeasible Right of Use), and Gbps (Gigabit per second). "Planned" c ables are those announced to come into service on or before June 1, 2001. "In-service" cables are those operational by December 1999. Cable systems that have been included in this diagram are those that have a capacity of 10 Gbps or greater. Map excludes most regional cable systems¯ I © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeographg, Inc. 1999 Note; submarine cable maps on thefollowing ~S included cables Urdike th. mapon Rgur~ 3; b~en exclud~ from these maps. : .... Sourc= TeleGeography, Inc. : © TeleGeography, Inc. 1999 TeleGeography 2000 Figure 5. Map of Major Submarine Cables in the United States (Atlantic Coast) & Caribbean Boston Mastic Beach Brookhaven Manhattan Manasquan~ Pe.nant Poin~ SL David’s West Cancun Baraequi~ Pa~ama City Puerto Viejo C’~m~na ~ort of Spain Columbus-2 Geor~ Note: Please refer to the note on Figure 4 for an explanation of this map. © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. ! 999 8uenaventur a ( No~e: Please refer ~o the note on Figure 4 f~ an explanation of this map. Flodar~ootis Foralez~ Joao Pessoe Recife Maceio Salvador © TeleGeography, Inc. 1999 TeleGeography 2000 Norden ann Zee Genoa TeleOeography 2000 © TeleCeographg, Inc. 1999 Shanto~ © TeleGeograph~l, Inc. 1999 TeleGeo~,’aphy 2000 TeleGeography 2000 © TeleGeographg, Inc. 1999 Ready for System Cost Cable Service Capacity (Gbps) STM-ls (US$ millions) Trans-Atlantic PTAT 1 1989 1.68 10.8 $510 TAT-9 1991 1.68 10.8 $406 TAT-10 1992 1.68 10.8 $300 TAT- 11 1993 1.68 10.8 $280 Colu mbus 2 1994 1.68 10.8 $337 CANUS 1 1994 5.00 32.0 $302 CANTAT 3 1995 5.00 32.0 $414 TAT-12/13 1995 30.00 192.0 $750 Gemini 1998 30.00 192.0 $600 AC 1 1998 80.00 512.0 $750 Atlantis-2 1999 5.00 32.0 $20 Columbus 3 1999 10.00 128.0 $236 TAT- 14 2000 640.00 4,096.0 $1,500 Level 3 2000 1,280.00 8,192.0 $800 AC 2 2001 1o280.00 8,192.0 $500 FLAG Atlantic-1 2001 2,400.00 15,360.0 $1,000 Hibernia 2001 160.00 1,024.0 $630 Oxygen 2002 2,560.00 16,384.0 n.a Trans-Pacific NPC 1991 1.68 10.8 $425 TP C 4 1992 1.12 7.2 $373 PacRim East 1993 1.12 7.2 $233 PacRim West 1994 1.12 7.2 n.a. TPC 5 1997 20.00 128.0 $1,240 G-P Cable 1999 5.00 32.0 $100 China-US 1999 80.00 512.0 $1,400 PC 1 1999 80.00 512.0 $1,240 Japan-US 2000 80.00 512.0 $1,000 Southern Cross 2000 80.00 512.0 $940 Europe-Africa-Asia SEA-ME-WE-2 1994 1.12 7.2 n.a. FLAG 1997 10.00 64.0 $I ,500 SEA-ME-WE-3 1999 80.00 512.0 $1,173 SAFE/SAT-3 2000 20.00 128.0 $280 © TeleGeography, Inc. 1999 TeleGeography 2000 Ready for System Cost Cable Service Capacity (Gbps) STM-ls (US$ millions) Pan-American/Caribbean Americas-l* 1994 5.00 32.0 n.a. MAC 1999 40.00 256.0 $415 PanAm 1999 10.00 64.0 $214 Americas-2 2000 80.00 512.0 $375 ARCOS-1 2000 60.00 384.0 $400 Atlantica-1 2000 20.00 128.0 $1,000 Mercus-1 2000 80.00 512.0 $900 PAC 2000 40.00 256.0 $280 Sam-1 2001 40.00 256.0 $900 Magellan 2001 80.00 512.0 $500 Pan-Asian Tasman 2 1991 1.12 7.2 $110 APC 1993 1.68 10.8 $332 APCN 1996 10.00 64.0 $550 Tasman 3 1999 40.00 256.0 n.a. East Asia Crossing 2001 80.00 512.0 $1,280 J~r transmission over 0C-3 TeleGeography 2000 © TeleGeography, Inc. 1999 International Circuit Usage by U.S. Carriers N. and C. America South America Caribbean W. Europe E. Europe Middle East Africa Asia Oceania Tote I U.S. Carrier 64 Kbps Circuit Usage For Private For Public Switched Total Circuits Idle Total Lines Network In Use Circuits Available 1996 34,104 71,011 105,115 3,390 108,505 1997 64,230 88,989 153,219 34, 504 187,723 1998 78,601 94,952 173,553 126,197 299,750 1996 3,194 6,412 9,606 4,531 14,137 1997 5,928 6,900 12,828 3,986 16,814 1998 7,958 7,716 15,674 5,536 21,210 1996 872 5,378 6,250 1,511 7,761 1997 1,034 6,478 7,512 4,006 11,518 1998 1,439 7,026 8,465 4,494 12,959 1996 33,083 29,536 62,619 33,053 95,672 1997 43,784 34,476 78,260 46,245 124,505 1998 69,051 34,133 103,184 52,937 156,121 1996 478 3,344 3,822 1,704 5,526 1997 1,326 3,742 5,068 1,719 6,787 1998 1,004 4,418 5,422 1,231 6,653 1996 908 2,836 3,744 560 4,304 1997 1,432 3,096 4,528 479 5,007 1998 1,920 2,807 4,727 844 5,571 1996 406 2,416 2,822 327 3,149 1997 699 2,608 3,307 292 3,599 1998 1,080 2,712 3,792 320 4,112 1996 15,015 16,475 31,490 27,163 58,653 1997 23,545 19,567 43,112 30,830 73,942 1998 30,563 19,262 49,825 45,915 95,740 1996 3,302 3,110 6,412 2,523 8,935 1997 5,430 4,861 10,291 1,690 11,981 1998 6,753 4,023 10,776 6,095 16,871 1996 91,362 140,518 231,880 74,762 306,642 1997 147,408 170,717 318,125 123,751 441,876 1998 198,369 177,049 375,418 241,052 616,470 Note: Data baaed on year-end FCC circuitstatus reports filed by AT&T. MCt WorldCom, and Sprint for circuits originating in continental U.S. only. "Idle" circuits= are cir- cuits owned by s carrier at year end but not in use. Satellite capacity utiJization is general|y not reflected in this data because U.S. carriers do not acquire irrte~natiom at sstellite capacity in advance. The FCC estimates that 25-30 percent of total submarine cable capacity landed in the U.S. is controlled by foreign carriers and thus not reported here. ALso, up to t00 percent of used capacity goes unrepo~ed because it is reserved for restoration purposes on~y. There is a discrepancy with figures repo~ted last year in Te/eGeography t989. One carrier had over reported the number of active private |ines in Canada. The FCC has subsequently made an adjustment to these figures. The correction shifted the classification of 31,000 circuits from activated to idle. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Figure 2 Idle Circuits of UiS~ Carriers by Region; 1~6~1998 .... Circuit Usage and Idle Capacity, 1996-98 1996 N. and C. America 1997 1998 South America Caribbean 1996 1997 1998 1996 1997 1998 I ¯ I PLs [] PSTN r~ldle W. Europe 1996 1997 1998 1996 E, Europe 1997 1998 1996 Middle East 1997 1998 1996 Africa 1997 1998 excess capacity 1996 Asia 1997 1998 1996 Oceania 1997 1998 0% 10% 20% ~SliTl~IIlIiIi~iil I 30% 40% 50% 60% 70% 80% 90% 100% Percent of Total 64 Kbps Circuits iiits o~ed :b~ a iiiiisi at ~£ar end b~ not i~se Saieiii~e caB££i~ ~iii!~i~io~ !S gei£iiii~ n~i ref!~d i~ 1~i~ ~a bbi£iSe Ui~ ca!iie~s di n~i ~q~ire ihternati~nai : : TeleGeography 2000 © TeleGeography, Inc. 1999 For Private For Public Switched Total Circuits Idle Total Lines Network In Use Circuits Available Canada 1996 20,410 41,793 62,203 917 63,120 1997 37,383 50,343 87,726 32,178 119,904 1998 53,302 54,719 108,021 120,961 228,982 M exi c o 1996 13,312 27,784 41,096 1,329 42,425 1997 19,155 36,935 56,090 1,148 57,238 1998 24,463 38,301 62,764 4,080 66,844 Hong Kong 1996 1,921 961 2,882 840 3,722 1997 3,058 1,221 4,279 1,825 6,104 1998 4,685 1,027 5,712 3,623 9,335 J a p a n 1996 7,682 5,354 13,036 17,696 30,732 1997 10,087 6,149 16,236 17,178 33,414 1998 11,907 6,098 18,005 26,042 44,047 Singapore 1996 1,114 582 1,696 508 2,204 1997 1,617 570 2,187 571 2,758 1998 1,959 608 2,567 1,999 4,566 U.K. 1996 18,959 12,648 31,607 10,844 42,451 1997 23,008 14,662 37,670 20,118 57,788 1998 47,410 11,818 59, 228 27,671 86,899 Note: Data b~sed on year-end FCC ci~it ~atus ~eporLs ~iied by AT&T, MC WofldC0m~and spent, for circuits 0riginating in continental U.Si 0niy. "id!e" circuits are cir~ nternat on: ;; Circuit Usage and Idle Capacity, 1996-1998 1996 Canada 1997 1998 1996 Mexico 1997 1998 1996 Hong Kong 1997 1998 I I ~IPLs ~PSTN I n ldle I 1996 Japan 1997 1998 I I 1996 Singapore 1997 1998 100 I I 1996 1997 1998 0% 10% 20% 3O% 40% 50% 60% 70% Percent of Total 64 Kbps Circuits 80% 90% 100% © TeleGeography, Inc. 1999 International Communications TeleGeography 2000 Satellites: rummy) m. my my m. Key ¯ Six36-MHztransponder equivalents for the C-band (4/6 GHz) ¯ Six36-MHztransponder equivalents for the Ku-band (11-12/14 GHz) [] ~’ Outline denotes satellites under construction as of August 1998 The maps on this page show 46 of the approximately 180 commercial geostation- ary satellites scheduled to begin operation by Januan/ 2001. These satellites use C- and Ku-band transponders to carry voice, data, and video. On some satellites (such as Panamsat’s), a majority of capacity is used for video transmission services. A 36 MHz C-band transponder can generally carry about 2,000 calls, but requires a large dish for TV reception. Smaller Ku-band dishes are increasingly the technology of choice for direct broadcast satellite recep- tion. Intersputnik satellites carry only one (Gorizont) or two (Express) Ku-band transponders at 36 MHz, therefore the actu- al Ku-band capacity for these satellites is somewhat less than suggested on the maps. Source: TeleGeography, Inc., International Satellite Oirectory 1999, company reports ~’,~ mv mv [] ¯ my m. mv ,.o mm v m,,, I01 TeleGeography 2000 © TeleGeography, Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Transponders (36 MHz Equivalent) Satellite Bus Launch Date Orbital Slot C-band Ku-band Columbia Columbia 515 Ford Aerospace January 1989 332.5 E 42.0 21.4 TDRS-5 TRW August 1991 185.7 E 12.0 17.0 TDRS-6 TRW January 1993 319.5 E 12.0 17.0 Intelsat Intelsat 511 Ford Aerospace June 1985 330.5 E 42.0 21.3 I nte Is at 601 H S -393 0 cto be r 1991 325.5 E 64.2 24.6 Intelsat 602 HS-393 October 1989 62.0 E 64.2 24.6 Intelsat 603 HS-393 March 1990 335.5 E 64.2 24.6 Intelsat 604 HS-393 June 1990 60.0 E 64.2 24.6 Intelsat 605 HS-393 August 1991 332.5 E 64.2 24.6 Intelsat 701 FS-1300 October 1993 180.0 E 42.3 24.5 Intelsat 702 FS-1300 June 1994 177.0 E 42.3 24.5 Intelsat 704 FS-1300 January 1995 66.0 E 42.3 24.5 Intelsat 705 FS-1300 March 1995 342.0 E 42.3 24.5 Intelsat 706 FS-1300 May 1995 307.0 E 42.3 24.5 Intelsat 707 FS-1300 March 1996 359.0 E 42.3 44.9 Intelsat 709 FS-1300 June 1996 310.0 E 42.3 24.5 Intelsat 801 AS-7000 February 1997 328.5 E 64.2 16.7 Intelsat 802 AS-7000 June 1997 174.0 E 64.2 16.7 Intelsat 804 AS-7000 December 1997 64.0 E 64.2 16.7 Intelsat 805 AS-7000 June 1998 304.5 E 64.2 16.7 Intelsat 901 n.a. 4Q 2000 60.0 E 76.0 20.0 Intelsat 902 n.a. 1Q 2001 62.0 E 76.0 20.0 Intelsat 903 n.a. 2Q 2001 335.5 E 76.0 20.0 Intelsat 904 n.a. 3Q 2001 325.5 E 76.0 20,0 Intelsat 905 n.a. 4Q 2001-1 Q 2002 332.5 E 76.0 20.0 Intersputnik Express 2 Express October 1994 14.0 W 10.0 2.0 Express 6 Express September 1996 80.0 E 10.0 2.0 Express-A1 n.a. October 1999 80.0 E n.a. n.a. Express-A2 n.a. 1Q 2000 14.0 W n.a. n.a. Gorizont 38 Gorizont November 1992 50.0 E 6.0 1.0 G o rizo nt 41 G o rizont N ove m be r 1993 130.0 E 6.0 1.0 Gorizont 42 Gorizont May 1994 142.5 E 6.0 1.0 LMI-1 A2100 AX September 1999 75.0 E 35.0 30.0 SESAT NPO-PM 3Q 1999 36.0 E 36.0 103 TeleGeography 2000 © TeleGeography, Inc. 1999 Transponders (36 MHz Equivalent) Satellite Bus Launch Date Orbital Slot C-band Ku-band New Skies NSS 513 Ford Aerospace May 1988 183.0 E 42.0 21.3 NSS 703 FS-1300 October 1994 57.0 E 42.3 24.5 NSS 803 AS-7000 September 1997 338.5 E 64.2 16.7 NSS 806 AS-7000 February 1998 319.5 E 36.0 44.0 Orion Orion 1 Eurostar 2000 January 1995 37.3 W 48.0 Orion 2 Eurostar 2000 4Q 1999 12.0 W 45.0 Panamsat PAS-1 GE 3000 June 1988 45.0 W 24.0 12.0 PAS-1R HS-601 2Q 2000 45.0 W 36.0 36.0 PAS-2 HS-601 July 1994 169.0 E 14.0 30.0 PAS-3 HS-301 January 1996 43.0 W 25.1 25.1 PAS -4 H S-601 Aug ust 1995 68.5 E 25.1 24.6 PAS-5 HS-601 August 1997 58.0 W 24.0 24.0 PAS-6 FS-1300 August 1997 43.0 W 36.0 PAS-6B HS-601 HP December 1998 43.0 W 32.0 PAS -7 FS 13003 S e pte m be r 1998 68.5 E 14.0 30.0 PAS-8 FS-13003 November 1998 166.0 E 24.0 24.0 PAS-9 HS-601 HP 2Q 2000 58.0 W n.a. n.a. Note Tabie inoiudes major commerciai intercontinental Source; TeieGeographyi Figure 3: Ge The lnternet TeleGeography 2000 © TeleGeography, Inc. 1999 International Internet Backbones Who’s Who Approximately 300 liSPs, or International Internet Service Providers, operate long distance transmission networks that, together, form the global lnternet’s international backbone. They cluster around four groups: Global liSP A handful of liSPs operate networks that connect to multiple countries in more than one region. The global liSP sector is dominated by players with strong U.S. roots, either in origin (AT&T/IBM, Global OnelSprint, GTEIBBN, MCI WorldComlUUNet, PSlNet), or by association: Cable & Wireless, which acquired much of MCl’s Internet backbone as a result of the MCl-WorldCom merger, or Canadian incumbent Teleglobe, with its origins just north of the CanadalU.S border. Indeed, the pattern emerging in international telecom services--the twenty largest carriers carry 60 percent of voice traffic--is even more pronounced in international Internet services. The ten largest liSPs control 70 percent of international Internet bandwidth. Regional liSP An liSP specializing in operating inter-regional international backbones, like Carrier1, GTS/E-Bone, and Level 3, in Europe; SingNet and Telstra, in Asia; and iAfrica. An increasing number of entrants into the regional lISP space include former national incumbents seeking to leverage region- al infrastructure. National lISP Typically, this is an Internet provider which has acquired international connectivity as part of a national or local service; acts increasingly as an upstream provider for other ISPs with little or no international connectivity; and moves to expand into neighboring countries. Examples include Demon lnternet (U.K.), Wirehub (Netherlands), or lnternet Initiative Japan. Academic Research networks often act as international con- nectivity providers alongside commercial liSPs. In many envi- ronments, they operate high-capacity, leading-edge systems-- catalyzing regional Internet development. Examples are Europe’s DANTE TEN-155 or Asia’s AI3 initiatives. America in the Middle Last year, in TeleGeography 1999, we explored the geography of the Internet and discovered that although it was becoming global, steep regional bandwidth costs, a comparative lack of local con- tent, and limited regional coordination caused the U.S. to become the Intemet’s central switching office, even for data flows within a region. Our current survey shows that the Internet is still U.S.-centric. But America’s role is changing. In Europe, where liberalization has 106 been greatest and country-to-country bandwidth prices have dropped fastest, two-thirds of international Internet connectivity now remains in-region, with hub cities like Amsterdam and Frankfurt running nine times as much bandwidth to other European countries as to the U.S. Asian liSPs, in a less integrated and more linguistically diverse region, continue to rely heavily on the U.S. west coast to exchange much of the region’s traffic, but significant progress in building regional connectivity has occurred. The result is that West Europe and East Asia are becoming sec- ondary Internet hubs. Enormous differences still remain from coun- try to country--differences which are reflected in the larger pattern of global connectivity. In less developed regions, intra-regional links have grown more slowly. Latin America still relies on slower satellite links for international connectivity; in Africa, just three countries (South Africa, Morocco, Tunisia) are connected to the Internet at above 10 Mbps, with Egypt due to join them shortly. The move away from a U.S.-centric architecture depends on a well- developed local infrastructure, and on a regulatory environment which neither prohibits connectivity nor prices it out of reach. Methodology and Acknowledgements The maps and tables on the following pages are the product of a six month study completed in September 1999. The research focused on the network topologies of 300 ISPs with international Internet links (that is, routers or switches directly connected across an international border over an internal network}. These links (and their bandwidth) were then tracked through 200 cities in 100 countries, generating a database of over 1,000 city pairs. Each ISP’s network routes and capacities were derived from a combina- tion of public documents, confidential interviews, and computer- based network analysis tools. The route-by-route figures published here represent the aggregate bandwidth for all ISPs covered in the study. Also, the specific switch and router locations have been grouped according to Consolidated Metropolitan Statistical Area, Census Metropolitan Area, or equiv- alent. And, finally, a disclaimer: because network architectures are complex and always changing, omissions may have occurred. While conducting the research for this study, we received assistance from many individuals and organizations. We would like to give special thanks to Bob Cohen [Cohen Communications Group, tel: +1 212 986 7720, bcohen@mail.bway.netl and Anthony Townsend [Global Network Analysts, www.networkanalysts.com] for their advice and substantial contributions. © TeleGeography, Inc. 1999 TeleGeography 2000 152 Mbps ~" .... Canada Europe Region 1 Africa Asia/Pacific Europe Latin America/Caribbean U.S./Canada Region 2 Africa (intra-regional) Asia/Pacific Europe Latin America/Caribbean U.S./Canada Region total Africa Asia/Pacific (intra-regional) Europe Latin America/Caribbean U.S./Canada Region total Africa Asia/Pacific Europe (intra-regional) Latin America/Caribbean U.S./Canada Region total Africa Asia/Pacific Europe Latin America/Caribbean (intra-regional) U.S./Canada Region total Africa Asia/Pacific Europe Latin America/Caribbean U.S./Canada (intra-regional) Region total International Internet Bandwidth 7.5 Mbps 3.0 Mbps 68.5 Mbps n.a. 170.0 Mbps 249.0 Mbps 3.0 Mbps 398.4 Mbps 152.2 Mbps 5,915.5 Mbps 6,469. I Mbps 68.5 Mbps 152.2 Mbps 31,918.3 Mbps 62.8 Mhps 13,257.5 Mbps 45,452.9 Mbps 62.8 Mbps 48.3 Mbps 949.1 Mbps 1,060.2 Mbps 170.0 Mbps 5,915.5 Mbps 13,257.5 Mbps 949.1 Mbps 7,841.0 Mbps 28,13 I. I Mbps 107 TeleGeography 2000 © TeleGeography, Inc. 1999 Rank City 1 1 London, United Kingdom 2 London, United Kingdom 3 Amsterdam, Netherlands 4 Amsterdam, Netherlands 5 Brussels, Belgium 6 Geneva, Switzerland 7 Frankfurt, Germany 8 Seattle, U.S. 9 Amsterdam, Netherlands 10 Chicago, U.S. 11 San Francisco, U.S. 12 London, United Kingdom Washington, DC, U.S. 14 NewYork City, U.S. 15 Frankfurt, Germany 16 Frankfurt, Germany 17 Amsterdam, Netherlands 18 Frankfurt, Germany 19 Montreal, Canada 20 San Francisco, U.S. 21 New York City, U.S. 22 Cheyenne, U.S. Chicago, U.S. Seattle, U.S. 25 Milan, Italy 26 Copenhagen, Denmark 27 Helsinki, Finland 28 San Francisco, U.S. 29 Brussels, Belgium 30 London, United Kingdom 31 Frankfurt, Germany 32 New¥ork City, U.S. 33 Amsterdam, Netherlands 34 Oslo, Norway 35 Amsterdam, Netherlands 36 Frankfurt, Germany 37 Hong Kong, China 38 London, United Kingdom 39 Ankara, Turkey 40 Portland, U.S. 41 Geneva, Switzerland 42 Paris, France 43 San Francisco, U.S. 44 Munich, Germany 45 los Angeles, U.S. 46 Paris, France 47 Copenhagen, Denmark Copenhagen, Denmark 49 Amsterdam, Netherlands Munich, Germany City 2 Internet Bandwidth New York City, U.S. 5,212.0 Mb ~s Paris, France 4,092.3 Frankfurt, Germany 3,349.1 Brussels, Belgium 3,026.0 London, United Kingdom 2,733.0 Paris, France 2,684.0 Mb~s Mb~s Mb~s Mb~s MbJs MbJs Mb]s Mb }s Is Is )s )s 1,085.0 Mb ~s 957.0 Mbps 936.5 Mbps 910.0 Mbps 827.8 Mbps Geneva, Switzerland 2,677.0 Vancouver, Canada 1,889.0 London, United Kingdom 1,883.8 Toronto, Canada 1,754.0 Mb Tokyo, Japan 1,487.0 Mb Washington, DC, U.S. 1,449.5 Mb London, United Kingdom 1,449.5 Mb Toronto, Canada Paris, France New York City, U.S. New York City, U.S. London, United Kingdom New York City, U.S. Seoul, South Korea Stockholm, Sweden Vancouver, Canada Montreal, Canada Toronto, Canada New York City, U.S. Stockholm, Sweden Stockholm, Sweden Sydney, Australia Paris, France Stockholm, Sweden Stockholm, Sweden Palermo, Italy Stockholm, Sweden Stockholm, Sweden Paris, France Vienna, Austria San Francisco, U.S. Milan, Italy Washington, DC, U.S. Tokyo, Japan Milan, Italy Stockholm, Sweden Singapore, Singapore Vienna, Austria Tokyo, Japan Washington, DC, U.S. London, United Kingdom New York City, U.S. Geneva, Switzerland Stockholm, Sweden 820.0 752.3 668.0 622.0 622.0 622.0 580.0 562.0 560.O 541.0 536.0 5O6.0 499.0 465.0 448.0 443.0 403.0 384.0 361.0 345.0 Mb 300.0 Mb 270.0 Mb 257.0 Mb 234.0 Mb 229.0 Mb 205.0 Mb 203.5 Mb 201.5 Mb 200.0 Mb 200.0 Mb Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mb ~s Mb ~s Mb =s Mb ~s Mb~s Mb~s Mb~s 3S 3S 3S 3S JS )S )S )S )S )S IS 189.0 Mb }s 189.0 Mbps 108 © TeleGeography, Inc. 1999 TeleGeo{~raphy 2000 Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2O 21 22 23 24 25 26 27 28 29 3O 31 32 33 34 35 36 37 38 39 40 41 42 43 45 46 47 48 5O C~ty London New York Amsterdam Frankfurt Paris Brussels Geneva Stockholm Washington, DC San Francisco Toronto Chicago Seattle Vancouver Tokyo Montreal Milan Copenhagen Seoul Vienna Zurich Munich Los Angeles Sydney Helsinki Cheyenne Madrid Palermo Hong Kong Singapore Oslo S~o Paulo Taipei Portland Moscow Ankara Osaka Dublin Nova Scotia Portsmouth Dallas Kuala Lumpur Vancouver Island Monaco Cologne Prague Hanover Auckland Leeds Buenos Aires International Country Internet Bandwidth United Kingdom 17,969.1 Mbps U.S. 13,204.9 Netherlands 10,874.1 Germany 10,516.6 France 9,687.3 Belgium 6,213.0 Switzerland 5,947.0 Sweden 4,388.2 U.S. 3,998.4 U.S. 3,949.8 Canada 3,512.0 U.S. 2,666.0 U.S. 2,607.0 Canada 2,511.0 Japan 2,392.7 Canada 1,690.4 Italy 1,628.0 Denmark 1,274.0 South Korea 1,105.8 Austria 978.5 Switzerland 869.3 Germany 756.0 U.S. 740.4 Australia 697.9 Finland 670.0 U.S. 624.0 Spain 618.0 Italy 571.8 China 541.2 Singapore 497.3 Norway 490.0 Brazil 329.3 Taiwan 323.8 U.S. 315.1 Russia 302.3 Turkey 300.0 Japan 250.0 Ireland 239.0 Canada 232.3 United Kingdom 213.8 U.S. 202.5 Malaysia 188.2 Canada 188.0 Monaco 187.5 Germany 180.0 Czech Republic 177.0 Germany 159.1 New Zealand 155.0 United Kingdom 155.0 Argentina 147.3 Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps Mbps 109 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 4. Map of Major Asia/Pacific International Backbone Routes, 1999 (~ a to Vanvouver Seattle to Portland to Montreal Mumbai Key Aggregate Internet backbone bandwidth 750 Mbps 300 Mbps 75 Mbps 20 Mbps Auckland Note: Map includes international backbone routes with at least 20 Mbps of aggregate capacity. Figures represent estimated Internet bandwidth between Consolidated Metropolitan Statistical Areas or equivalents, Oomestic backbone routes are omitted. Data current to September 1999. Source: TeleGeography GlobalBackbone Database; additional research by Cohen Communications Group © TeleGeography, Inc. 1999 110 © TeleGeographg, Inc. 1999 TeleGeo~raphy 2000 Rank City 1 1 Tokyo, Japan 2 SeouI, South Korea 3 Sydney, Australia 4 Hong Kong, China 5 Tokyo, Japan 6 Singapore, Singapore 7 Tokyo, Japan 8 Tokyo, Japan SeouI, South Korea Osaka, Japan 11 Kuala Lumpur, Malaysia 12 Taipei, Taiwan 13 Taipei, Taiwan 14 Auckland, New Zealand 15 TelAviv, Israel 16 Tokyo, Japan 17 Seoul, South Korea 18 Singapore, Singapore 19 Haifa, Israel 20 Seoul, South Korea 21 Hong Kong, China Yokohama, Japan Sydney, Australia Sydney, Australia Singapore, Singapore Singapore, Singapore Seoul, South Korea Osaka, Japan Mumbai, lndia Bangalore, India Auckland, NewZealand 32 Taipei, Taiwan 33 Tel Aviv, Israel 34 Kuala Lumpur, Malaysia 35 Mumbai, India 36 Melbourne, Australia 37 Hong Kong, China 38 Bangkok, Thailand 39 Manila, Philippines 40 Bangkok, Thailand 41 Singapore, Singapore 42 Sydney, Australia Jakarta, Indonesia 44 Hong Kong, China 45 Rosh ha’Ayin, Israel 46 Taipei, Taiwan 47 Hong Kong, China 48 Bangkok, Thailand 49 Hong Kong, China 50 Taipei, Taiwan City 2 Internet Bandwidth San Francisco, U.S. 1,487.0 Mbps 752.3 541.0 361.0 270.0 229.0 203.5 155.0 155.0 155.0 115.0 111.0 101.5 100.0 96.0 73.0 67.9 59.0 58.5 San Francisco, U.S. San Francisco, U.S. San Francisco, U.S. Portland, U.S. San Francisco, U.S. Los Angeles, U.S. New York City, U.S. Vancouver Island, Canada New York City, U.S. San Francisco, U.S. San Francisco, U.S. New York City, U.S. San Francisco, U.S. New York City, U.S. Chicago, U.S. Los Angeles, U.S. Los Angeles, U.S. New York City, U.S. Tokyo, Japan Osaka, Japan Seattle, U.S. Montreal, Canada New York City, U.S. Chicago, U.S. Vancouver, Canada Portland, U.S. San Francisco, U.S. Los Angeles, U.S. Los Angeles, U.S. Los Angeles, U.S. Tokyo, Japan London, United Kingdom Singapore, Singapore Palermo, Italy San Diego, U.S. Tokyo, Japan San Francisco, U.S. San Francisco, U.S. Vancouver Island, Canada Tokyo, Japan Los Angeles, U.S. San Francisco, U.S. Singapore, Singapore Boston, U.S. Chicago, U.S. Seoul, South Korea Los Angeles, U.S. Taipei, Taiwan Los Angeles, U.S. Mbps Mbps Mbps Mbps Mb ~s Mb ~s Mbas Mbas Mb~s Mb~s Mb~s Mb~s Mb~s Mb~s Mb~s Mb~s Mb~s Mb ~s 48.0 Mb ~s 45.5 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb as 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 38.9 Mb as 36.0 Mb ~s 35.5 Mb ~s 34.0 Mb ~s 32.0 Mb ~s 31.8 Mb~s 27.5 Mb ~s 24.4 Mb~s 23.0 Mb ~s 22.5 Mbps 20.0 Mbps 20.0 Mbps 18.5 Mbps 18.0 Mbps 17.0 Mbps 16.7 Mbps 16.0 Mbps 14.0 Mbps 12.6 Mbps 111 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 6. Map of Major European International Backbone Routes, 1999 Helsinki Oslo Moscow to New York to San Francisco c Genev; Key Aggregate Internet backbone bandwidth 5Gbps 2.5Gbps 500Mb#s lOOMbps Ankara Note: Map includes international backbone routes with at least 100 Mbps of aggregate capaciW. Figures represent estimated Internet bandwidth between Consolidated Metropolitan Statistical Areas or equivalents. Domestic backbone routes are omitted. Data current to September 1999. Source: TeleGeography Global Backbone Database; additional research by Cohen Communications Group © TeleGeography, Inc. 1999 112 © TeleGeography, Inc. 1999 TeleGeography 2000 Rank City 1 1 London, United Kingdom 2 London, United Kingdom Paris, France 3 Amsterdam, Netherlands Frankfurt, Germany 4 Amsterdam, Netherlands Brussels, Belgium 5 Brussels, Belgium London, United Kingdom 6 Geneva, Switzerland Paris, France 7 Frankfurt, Germany Geneva, Switzerland 6 Amsterdam, Netherlands London, United Kingdom 9 London, United Kingdom Washington, DC, U.S. 10 Frankfurt, Germany Paris, France 11 Frankfurt, Germany New York City, U.S. 12 Amsterdam, Netherlands NewYork City, U.S. 13 Frankfurt, Germany London, United Kingdom 14 Stockholm, Sweden NewYork City, U.S. 15 Milan, Italy NewYork City, U.S. 16 Copenhagen, Denmark Stockholm, Sweden 17 Helsinki, Finland Stockholm, Sweden 18 Brussels, Belgium Paris, France 19 London, United Kingdom Stockholm, Sweden 20 Frankfurt, Germany Stockholm, Sweden 21 Palermo, Italy New York City, U.S. 22 Amsterdam, Netherlands Stockholm, Sweden 23 Oslo, Norway Stockholm, Sweden 24 Amsterdam, Netherlands Paris, France 25 Frankfurt, Germany Vienna, Austria 26 London, United Kingdom Milan, Italy 27 Ankara, Turkey Washington, DC, U.S. 28 Geneva, Switzerland Milan, Italy 29 Paris, France Stockholm, Sweden 30 Munich, Germany Vienna, Austria 31 Paris, France Washington, DC, U.$. 32 Copenhagen, Denmark London, United Kingdom Copenhagen, Denmark NewYork City, U.S. 34 Amsterdam, Netherlands Geneva, Switzerland Munich, Germany Stockholm, Sweden 36 Zurich, Switzerland NewYork City, U.S. 37 Frankfurt, Germany Milan, Italy 38 Frankfurt, Germany Washington, DC, U.S. 39 Madrid, Spain Paris, France Munich, Germany Paris, France 41 Amsterdam, Netherlands Vienna, Austria Copenhagen, Denmark Frankfurt, Germany Frankfurt, Germany San Francisco, U.S. Hanover, Germany NewYork City, U.S. Leeds, United Kingdom Washington, DC, U.S. 46 London, United Kingdom Zurich, Switzerland 47 Dublin, Ireland London, United Kingdom 48 Moscow, Russia NewYork City, U.S. 49 Milan, Italy Zurich, Switzerland 50 Vienna, Austria Zurich, Switzerland City 2 Internet Bandwidth New York City, IJ.S. 5,212.0 Mb )s 4,092.3 Mb )s 3,349.1 Mb )s 3,026.0 Mb ~s 2,733.0 Mb ~s 2,684.0 Mb ~s 2,677.0 Mb )s 1,883.8 Mb )s 1,449.5 Mb )s 957.0 Mb )s 936.5 Mb )s 910.0 Mb )s 827.8 Mb ~s 668.0 Mb ~s 580.0 Mb ~s 562.0 Mb: Is 560.0 Mb ~s 536.0 Mb ~s 506.0 Mb ~s 499.0 Mb ~s 465.0 Mbps 448.0 Mbps 443.0 Mbps 403.0 Mbps 384.0 Mbps 345.0 Mbps 300.0 Mbps 257.0 Mbps 234.0 Mbps 205.0 Mbps 201.5 Mbps 200.0 Mbps 200.0 Mbps 189.0 Mbps 189.0 Mbps 180.0 Mbps 177.0 Mbps 165.0 Mbps 159.0 Mbps 159.0 Mbps 155.0 Mbps 155.0 Mbps 155.0 Mbps 155.0 Mbps 155.0 Mbps 153.0 Mbps 148.0 Mbps 137.5 Mbps 113.0 Mbps 102.0 Mbps 113 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 8. Map of Major South American International Backbone Routes, 1999 to to to to to to Vancouver San Francisco Dallas Montreal NewYork Washington, DC to Miami Lima~ Bogot~ aquil Rio De Janeiro S~o Paulo to Paris Horizonte to Palermo Santiago Buenos Aire~ Montevideo ~orto Alegre Key Aggregate Internet ~acKDone uandw~dth 100 Mbps 50 Mbos 25 Mb~s I0 Mbos Note: Map includes international backbone routes with at least 2.5 Mbps of aggregate capacity. Figures represent estimated Internet bandwidth between Consolidated Metropolitan StatisticalAreas or equivalents. Domestic backbone routes are omitted. Data current to September 1999. Source: TeleGeography Global Backbone Database © TeleGeography, Inc. 1999 114 © TeleGeographyo Inc. 1999 TeleGeography 2000 Rank City 1 1 S~o Paulo, Brazil 2 Mexico City, Mexico 3 Buenos Aires, Argentina 4 Rio de Janeiro, Brazil 5 S~o Paulo, Brazil 6 Caracas, Venezuela Caracas, Venezuela Mexico City, Mexico Monterrey, Mexico 10 Santiago, Chile 11 Buenos Aires, Argentina 12 Monterrey, Mexico Buenos Aires, Argentina 14 Rio de Janeiro, Brazil 15 Lima, Peru 16 S~o Paulo, Brazil 17 Rio de Janeiro, Brazil 18 S~o Paulo, Brazil 19 Lima, Peru 20 Rio de Janeiro, Brazil 21 Bogota, Colombia Lima, Peru Rio de Janeiro, Brazil 24 Caracas, Venezuela 25 Panama City, Panama 26 Hamilton, Bermuda Panama City, Panama S~o Paulo, Brazil 29 Belo Horizonte, Brazil 30 Quito, Ecuador 31 Asunci6n, Paraguay 32 Guayaquil, Ecuador 33 Bogota, Colombia Havana, Cuba Havana, Cuba Medellin, Colombia S~o Paulo, Brazil S~o Paulo, Brazil Belmopan, Belize Bogota, Colombia Brasilia, Brazil Bridgetown, Barbados Buenos Aires, Argentina Havana, Cuba Port of Spain, Trinidad & Tobago Port-au-Prince, Haiti Porto Alegre, Brazil St. John’s, Antigua 49 Asuncibn, Paraguay 50 $8o Paulo, Brazil City 2 Internet Bandwidth New York City, U.S. 182.1 Mb~s Dallas, U.S. New York City, U.S. Montreal, Canada Montreal, Canada Miami, U.S. S~o Paulo, Brazil Houston, U.S. Dallas, U.S. Washington, DC, U.S. Dallas, U.S. Los Angeles, U.S. Palermo, Italy New York City, U.S. Washington, DC, U.S. Paris, France Paris, France Washington, DC, U.S. Montreal, Canada Washington, DC, U.S. Dallas, U.S. Vancouver Island, Canada Dallas, U.S. Los Angeles, U.S. Miami, U.S. New York City, U.S. Atlanta, U.S. San Francisco, U.S. New York City, U.S. Miami, U.S. San Francisco, U.S. Miami, U.S. Cheyenne, U.S. Montreal, Canada Washington, DC, U.S. Montreal, Canada Boston, U.S. Miami, U.S. Los Angeles, U.S. Miami, U.S. Washington, DC, U.S. Los Angeles, U.S. Montevideo, Uruguay Palermo, Italy Toronto, Canada Los Angeles, U.S. Washington, DC, U.S. Toronto, Canada Miami, U.S. Lisbon, Portugal 91.0 Mb ~s 68.2 Mb ~s 68.0 Mb Is 68.0 Mb Is 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 45.0 Mb ~s 42.3 Mb ~s 42.0 Mb ~s 34.0 Mb ~s 34.0 Mb ~s 25.5 Mb ~s 20.0 Mb ~s 13.0 Mb ~s 12.0 Mb ~s 11.6 Mb ~s 10.0 Mb ~s 9.5 Mb ~s 8.0 Mb ~s 8.0 Mb ~s 8.0 Mb ~s 6.8 Mb ~s 6.0 Mb ~s 4.0 Mb ~s 4.0 Mb ~s 4.0 Mb ~s 3.5 Mb ~s 3.0 Mb ~s 2.3 Mb ~s 2.5 Mb ~s 2.0 Mb ~s 2.0 Mb ~s 2.0 Mb ~s 2.0 Mb ~s 2.0 Mb,~s 2.0 Mb~s 2.0 Mb~s 2.0 Mb ~s 2.0 Mb ~s 2.0 Mb ~s 2.0 Mbps 2.0 Mbps 2.0 Mbps 2.0 Mbps 2.0 Mbps 2.0 Mbps 1.9 Mbps 1.5 Mbps 115 TeleGeography 2000 © TeleGeography, Inc. 1999 116 © TeleGeography, Inc. 1999 TeleGeography 2000 City 1 New York City, U.S. Seattle, U.S. Chicago, U.S. San Francisco, U.S. Washington, DC, U.S. New York City, U.S. New York City, U.S. city 2 London, United Kingdom Rank 1 2 3 4 5 6 7 8 NewYork City, U.$. 9 Montreal, Canada 10 Sen Francisco, U.S. 11 N ew York City, U.S. 12 Cheyenne, U.S. Chicago, Seattle, U.S. 15 New York City, U.S. 16 San Francisco, U.S. 17 NewYork City, U.S. 18 San Francisco, U.S. 19 Washington, DC, U.S. 20 Portland, U.S. 21 San Francisco, U.S. 22 Los Angeles, U.S. 23 Washington, DC, U.S. 24 NewYork City, U.S. 25 N ewYork City, U.S. 26 N ewYork City, U.S. 27 Washington, DC, U.S. 28 New York City, U.S. N ewYork City, U.S. N ewYork City, U.S. N ewYork City, U.S. San Francisco, U.S. Vancouver Island, Canada Washington, DC, U.S. 35 N ewYork City, U.S. 36 Washington, DC, U.S. 37 San Francisco, U.S. 38 San Francisco, U.S. 39 NewYork City, U.S. 40 San Francisco, U.S. 41 New York City, U.S. 42 Dallas, U.S. 43 Chicago, U.S. 44 NewYork City, U.S. 45 Montreal, Canada 46 Montreal, Canada 47 Los Angeles, U.S. 48 NewYork City, U.S. 49 Los Angeles, U.S. 50 New York City, U.S. Vancouver, Canada 1,889.0 Mb Toronto, Canada 1,754.0 Mb Tokyo, Japan 1,487.0 Mb London, United Kingdom 1,449.5 Mb Toronto, Canada 1,085.0 Mb Frankfurt, Germany 936.5 Mb Amsterdam, Netherlands 910.0 Mb New York City, U.S. 820.0 Mb Seoul, South Korea 752.3 Mb Stockholm, Sweden 668.0 Mb Vancouver, Canada 622.0 Mb Montreal, Canada 622.0 Mb Toronto, Canada 622.0 Mb Milan, Italy 580.0 Mb Sydney, Australia 541.0 Mb Palermo, Italy 465.0 Mb Hong Kong, China 361.0 Mb Ankara, Turkey 300.0 Mb Tokyo, Japan 270.0 Mb Singapore, Singapore 229.0 Mb Tokyo, Japan 203.5 Mb Paris, France 201.5 Mb Copenhagen, Denmark 200.0 Mb S~o Paulo, Brazil 182.1 Mb Zurich, Switzerland 180.0 Mb Frankfurt, Germany 165.0 Mb Hanover, Germany 155.0 Mb Osaka, Japan 155.0 Mb Nova Scotia, Canada 155.0 Mb Tokyo, Japan 155.0 Mb Frankfurt, Germany 155.0 Mb Seoul, South Korea 155.0 Mb Leeds, United Kingdom 155.0 Mb Moscow, Russia 137.5 Mb Amsterdam, Netherlands 125.5 Mb Kuala Lumpur, Malaysia 114.4 Mb Taipei, Taiwan 111.0 Mb Taipei, Taiwan 101.5 Mb Auckland, New Zealand 100.0 Mb Tel Aviv, Israel 96.0 Mb Mexico City, Mexico 91.0 Mb Tokyo, Japan 73.0 Mb Buenos Aires, Argentina 68.2 Mb Rio de Janeiro, Brazil 68.0 Mb S~o Paulo, Brazil 68.0 Mb Seoul, South Korea 67.9 Mb Madrid, Spain 61.0 Mb Singapore, Singapore 59.0 Mb Haifa, Israel 58.5 Mb Internet Bandwidth 5,212.0 Mb )s )s )s )s }s }s }s )s )s )s )s )s )s )s )s )s }s )s }s )s }s }s )s )s )s )s )s )s )s )s )s )s )s )s )s )s )s )s )s )s |s )s )s )s )s )s )s )s )s )s 117 TeleGeography 2000 © TeleGeography, Inc. 1999 118 © TeleGeographyo Inc. 1999 TeleGeography 2000 Figure 13. The Top 50 U.S. Domestic Backbone Routes, 1999 Rank City1 City2 Internet Bandwidth 1 NewYork City, NY Washington, DC 7,478 rvlbps 2 Dallas, TX Houston, TX 5,301 Mbps 3 Los Angeles, CA San Francisco, CA 5,145 Mbps 4 San Francisco, CA Salt Lake City, UT 4,566 Mbps 5 Indianapolis, IN St. Louis, MO 4,231 Mbps 6 Washington, DC Atlanta, GA 4,163 Mbps 7 Chicago, IL NewYork City, NY 4,003 Mbps 8 San Francisco, CA Seattle, WA 3,913 Mbps 9 Los Angeles, CA Dallas,’lX 3,780 Mbps 10 Chicago, IL Washington, DC 3,583 Mbps 11 Atlanta, GA Dallas,’lX 3,468 Mbps 12 Boston, MA NewYork City, NY 3,428 Mbps 13 Atlanta, GA Miami, FL 3,023 Mbps 14 Chicago, IL Atlanta, GA 2,913 Mbps Washington, DC Boston, MA 2,913 Mbps 16 Kansas City, MO Dallas, TX 2,888 Mbps 17 Atlanta, GA Houston, TX 2,803 Mbps 18 Chicago, IL Kansas City, MO 2,643 Mbps 19 Chicago, IL Indianapolis, IN 2,538 Mbps 20 San Francisco, CA Kansas City, MO 2,533 Mbps 21 Atlanta, GA Charlotte, NC 2,428 Mbps 22 Kansas City, M0 St. Louis, MO 2,363 Mbps 23 Cleveland, OH Buffalo, NY 2,338 Mbps 24 Oklahoma City, OK Dallas,’lX 2,293 Mbps 25 Los Angeles, CA Las Vegas, NV 2,273 Mbps 26 Cincinnati, OH indianapolis, IN 2,248 Mbps Phoenix, AZ Tucson, AZ 2,248 Mbps 28 El Paso, TX Tucson, AZ 2,203 Mbps St. Louis, M0 Tulsa, OK 2,203 Mbps 30 Columbus, OH Dayton, OH 2,183 Mbps 31 Denver, CO Salt Lake City, LIT 2,183 Mbps 32 Cincinnati, OH Dayton, OH 2,138 Mbps Las Vegas, NV Phoenix, AZ 2,138 Mbps 34 Akron, OH Cleveland, OH 2,093 Mbps Albany, NY NewYork City, NY 2,093 Mbps Atlanta, GA Jackson, MS 2,093 Mbps Cleveland, OH Akron, OH 2,093 Mbps 38 Baton Rouge, LA Jackson, MS 2,048 Mbps Buffalo, NY Rochester, NY 2,048 Mbps Charlotte, NC Winston-Salem, NC 2,048 Mbps Dallas, TX Midland, TX 2,048 Mbps Denver, CO Topeka, KS 2,048 Mbps Grand Junction, CO Prove, LIT 2,048 Mbps Kansas City, MO Topeka, KS 2,048 Mbps Roanoke, VA Washington, DC 2,048 Mbps Rochester, NY Buffalo, NY 2,048 Mbps Syracuse, NY Rochester, NY 2,048 Mbps Topeka, KS Kansas City, Me 2,048 Mbps Tulsa, OK Oklahoma City, OK 2,048 Mbps Winston-Salem, NC Charlotte, NC 2,048 Mbps Note: ~guresrepresentt~ta~ntemetbandwidthbet~veenC~so~idatedMetr~itanStat~sti~a~Ar~s~requiva~o~ts~ Alflnternatienal backboneroutesareo~ttad, Data curremtto September 1999, Source: Anthony Townsend © TefeGeography, ~nc~ 119 TeleGeography 2000 © TeleGeographg, Inc. 1999 lnternet Exchange Points lnternet eXchange (IX} points, also known as Network Access Points (NAPs) or Metropolitan Area Exchanges (MAEs), are physical installations created by third parties to facilitate traffic exchange between in, dependent Internet Service Providers {ISPs). The table below shows a selection of IX points operat- ing in October 1999, based on exchange size, geographical diversity, and location in a global lntemet hub city. Number of ISPs do not always correlate to traffic size, but do provide a rough benchmark for comparison. ISPs also exchange traffic at private locations which are not well publicized [for a review of private peering arrangements, see "The Old Boy’s Network," Data Communications, Oct. 1999). Figure 1. Major Internet Exchange Points, A-K Internet Exchange AIX (Athens IX) AMS-IX (AMSterdam IX) APE (Auckland Peering Exchange/CityLink) BCIX (British Columbia IX) BIX (Budapest IX,) BNIX (BelNet IX) BUHIX (BUcHarest IX) Chicago NAP (Ameritech) Chile NAP CINX (Capetown InterNet eXchange) CIXP (Cern IX Point) CORE (Common Routing Exchange) DeCIX (Deutsche Commercial IX) DGIX (Distributed Global Internet eXchange) DIX (Danish IX) ESPANIX]PNE (ESPANa IX]Puncto Neutral Espafiol) FICIX (Rnnish Commercial IX) HKIX (Hong Kong IX) IIX (Indonesia IX) IIX (Israeli IX) INEX (Internet Neutral EXchange) JPIX (Japan IX) KINX (Korea InterNet eXchange/PSINet) KLIX (Kuala Lumpur IX) Connected ISPs Location/URL 8 Athens, Greece www.aix.gr 71 Amsterdam, Netherlands www.ams-ix.net 3 Auckland, New Zealand www.citylink.co.nz 7 Vancouver, Canada www.bc.net/services.html 19 Budapest, Hungary www.nic.hu/bix 30 Brussels, Belgium www.belnet.be/bnix 16 Bucharest, Romania www.buhix.ro 93 Chicago, U.S. nap,aads.net/main.html 9 Santiago, Chile www.nap.cl 11 Capetown, South Africa www.jinx.n et.za/jinx/cinx 26 Geneva, Switzerland wwwcs.cern.ch/public/services/cixp/index.html n.a. Manila, Philippines www.ph.net/CORE.html 51 Frankfurt, Germany www.decix.de 16 Stockholm, Sweden www.netnod.se 21 Lyngby, Denmark www.uni-c.dk/dix 13 Madrid, Spain www.espanix.net 10 Helsinki, Finland wwvv.ficix.fi 49 Hong Kong, China www.hkix.net 35 Jakarta, Indonesia www.iix.net.id 18 Petah-’fiqva, Israel www.iso c.org.il/iix.html 6 Dublin, Ireland www.inex.ie 36 Tokyo, Japan www.jpix.co.jp 15 Seoul, Korea, Rep. www.kinx.net n.a. Kuala Lumpur, Malaysia www.klix.net SourCe: TeleGeography, ]no, ¯ .... © ~eleGeography~ Inc. 1989 120 © TeleGeography, Inc. 1999 TeleGeoi~raphy 2000 Figure 2. Major Internet Exchange Points, L-Z Icternet Exchange L-GIX (Latvian Global IX) L21X (Layer 2 Internet eXchange/DACOM-IX) LINX (London InterNet eXchange) Mg-lX (M9 Telehouse IX) MAE-East (MCI WorldCom) MAE-FFM (MCI WorldCom) MAE-Paris (MCI WorldCom) MAE-West (MCI WorldCom/NASA Ames) MIX (Milan IX) Connected ISPs Location/URL 16 Riga, Latvia www.nic.lv/gix.html 32 Seoul, Korea, Rep. www.dacomix.net 82 London, United Kingdom www.linx.n et 54 Moscow, Russia www.ripn.net/ix 92 Washington, DC, U.S. www.mae.net/east.html n.a. Frankfurt, Germany www.mfst.com/mfs-international/frankfurt.html n.a. Paris, France www.mfst.com/mfs-international/paris.html 83 San Jose, U.S. www.mae.net/west.html 11 Milan, Italy www.aiip.it/mixit.htTnl 35 Pennsauken, U.S. www.sprintbiz.c om/na p/ 20 Prague, Czech Republic wwvv.nix.cz 8 Oslo, Norway 193.156.90.0 New York NAP (Sprint) NIX (Neutral IX) NIX (Norwegian IX) NSPIXP2 (Ne’aNork Service Provider IX Project 2) n.a. Tokyo, Japan wwvv.wide.ad.jp NSPIXP3 (Ne’aNork Service Provider IX Project 3) n.a. Osaka, Japan wwvv.wide.ad.jp NYIIX (New York International IX) 23 New York City, U.S. wwvv.nyiix.net PacBell NAP (Pacific Bell) 50 SF/LA, U.S. wwvv.pacbel].c om/prod ucts/business/fastrak/ne’aNorking/nap PAIX (Palo Alto IX/AboveNet) 60 Palo Alto, U.S. wwvv.paix.net PARIX (France T~l~}com) 18 Paris, France wwvv.parix.net PhlX (Philippine IX) 9 Manila, Philippines list.infocom.seq uel.net PIX (Portuguese IX) 16 Lisbon, Portugal vwwv.fccn.pt PI-F-ANSP (PI-F Academic Ne’aNork at S~o Paulo) 6 S~o Paulo, Brazil wwvv.ansp.br QIX (Quebec IX) 6 Montreal, Canada wwvv.risq.qc.c a/rese au/table/brancher/brancher_01 .html SFINX (Service for French INternet eXchange) 47 Paris, France wwvv.sfinx.tm.fr SIX (Seattle IX) 24 Seattle, U.S. wwvv.altopia.com/six SIX (Slovak IX) 21 Bratislava, Slovakia wwvv.six.sk STIX (SingTel IX) 22 Singapore wwvv.stix.net SydNAP (AUIX/AusBONE) 15 Sydney, Australia wwvv.ausbone.net/syd ney.htm ThaiSARN PIE (Thai SARN Public Internet Exchange) 18 Bangkok, Thailand ntl.nectec.or.th/pie THIX (THai IX) 27 Bangkok, Thailand wwvv.c at.net.th/S e rvic esiTH IX/thix.html TorlX (Toronto IX) 11 Toronto, Canada wwvv.torix.net TWIX (TaiWan IX/Chunghwa) 23 Taipei, Taiwan wwvv.Wvix.net VIX (Vienna IX) 43 Vienna, Austria vwwv.vix.at Source; Te[eGeography, tnc. © TeleGeography, Inc. 1999 121 Tele(3eol~raphy 2000 © TeleGeography, Inc. 1999 Global Internet Indicators "Hosts" are computers which are connected to the Internet. Technically, this means that they respond when hailed at the IP address corresponding to their entry in the lnternet’s domain name registry. The tables on the following pages present two methods for examining Internet host growth: by Top Level Domain name and by country. Top Level Domains (TLDs) are the domain name extensions like ".corn" which are used by Internet hosts. Figure 2 shows data for 55 TLDs, including six generic TLDs (".corn", ".net’, ".or~", etc.) as well as 49 country code or ccTLDs (".ca", ".jp", ".fr", etc.). These TLDs represent over 98 pement of the Inter-net’s hosts. Although country code TLDs exist for 250 economies (see the list at www.iana.or~!cctld.htm]), most have such small host populations that data points are unlikely to be accurate. The data shown here are based on continuous sampling by the Netsizer project team at Telcordia’s Applied Reseamh ~oup (www.netsizer.com), ahd verified against surveys undertaken biannually by the Internet Software Consortium (ISC, www.isc.org)--formerly carded out under the aegis of Network W~zards (www.nw.com)--as well as monthly surveys of 96 countries by R~seaux IP Europ~ennes (www.ripe.net). Fi~oure 3 shows TLDs broken down by country. For ccTLDs, the conversion process is straight forward: host names ending in a country code such as ".ca’’ are assumed to be associated with the corresponding country (in this example, Canada). Allocating countries to gTLDs is more difficult---and more important: ".corn" alone represents more than a third of all hosts. The Netsizer project deals with this problem by taking numerical IP addresses as its starting point and constantly gen- erating random combinations of available four-octet IPv4 addresses to test (such as 128.96.4 I. I). Then Netsizer looks them up in the domain name reostries (this is called a "whois" query) to see what domain name they are associated with. The ’whois’ query returns the registered posta] address for the host’s administrator; the host is then associated with the coun- try included in that address. This makes it easy to allocate gTLDs to single countries. ~i~ Figure 1. Host Growth and Distribution by Region, 1997-1999 Host Distribution by Region, Sep. 1997 Host Distribution by Region, Sep. 1999 Africa Latin 0.5% America 0.6% Europe Europe 18.4% 15.7% Asia/Pacific 9.9% Asia/Pacific Africa Latin America 8.6% 0.3% 1.1% Note: Area of pies are scaled bytetal host growt~ between |997 and t999. :~ Source: Teloordia Tecfinologies (www.netsizer.com) Design © TefeGeography~ Inc. 1999 122 © TeleGeographU, Inc. 1999 TeleGeography 2000 Figure 2. Host Count by Top Level Domain~ 1997-1999 TLD Country Sept, 97 Sept. 98 corn -- 7,600,100 11,745,000 net -- 3,825,700 7,223,100 e d u -- 3,584,200 4,867,200 jp Japan 1,120,400 1,678,800 rail United States 902,100 1,595,900 u!~ United Kingdom 859,300 1,315,800 us United States 732,200 1,121,700 ca Canada 640,800 1,084,400 de Germany 834,700 1,173,500 au Australia 593,800 779,500 org -- 537,000 677,600 gov United States 428,900 614,500 nl Netherlands 332,500 543,800 fr Franc e 308,900 455,300 fi Finland 328,400 488,000 se Sweden 298,100 390,100 it Italy 206,500 351,300 tw Taiwan 121,700 n.a. no Norway 181,200 324,300 es Spain 161,000 239,700 dk Denmark 135,900 211,300 be Belgium 84,100 176,000 br Brazil 87,800 177,200 kr Korea (South) 100,400 177,200 c h ¯ Switze rla nd 143,600 222,700 nz New Zealand 130,600 195,600 at Austria 68,700 131,500 mx Mexico 22,400 81,600 ru Russia 56,800 123,200 za South Africa 103,300 n.a. pl Poland 74,000 112,600 il Israel 40,800 90,500 ar Argentina 17,500 41,300 cz Czech Republic 45,100 71,100 sg Singapore 73,300 78,400 hu Hungary 35,300 71,100 hk Hong Kong 55,800 73,800 g r G reece 29,400 43,900 tr Turkey 18,400 36,100 pt Portugal 29,700 51,000 my Malaysia 35,000 n.a. ie Ireland n.a. 39,800 cl Chile n.a. 25,300 cn China 19,700 n.a. co Colombia 10,100 13,800 th Thailand 14,200 23,500 ee Estonia 9,400 16,100 is Iceland 15,000 21,200 in India 3,100 9,700 ua Ukraine 9,400 15,000 si SIovenia n.a. n.a. sk SIovakia 6,200 16,000 id Indonesia 6,100 11,900 re Romania 5,000 10,200 uy U rug u ay 1,000 8,400 Sept.99 23,732,000 12,596,000 5,975,900 2,321,200 1,957,600 1,768,100 1,522,100 1,509,600 1,439,500 966 300 879 100 772 900 714 100 702 800 572 200 527 500 504 700 455 500 378 700 347 800 311 600 298 500 298 000 282 700 270 300 230800 191 300 190 700 181 000 165 900 154 100 136 700 109 400 105 200 102 600 102 200 96,300 69,400 67,700 64,700 58,800 48,600 45,700 38,400 32,800 29,500 27,000 26,700 21,500 21,500 21,000 20,800 19,500 16,600 13,800 CAGR, 97-99 77% 81% 29% 44% 47% 43% 44% 53% 31% 28% 28% 34% 47% 51% 32% 33% 56% 93% 45% 47% 51% 88% 84% 68% 37% 33% 67% 192% 79% 27% 44% 83% 150% 53% 18% 70% 31% 54% 92% 48% 30% 40% 8O% 44% 69% 33% 163% 51% 83% 79% 82% 271% Note: An tnternet hosttypicaily represents one computer connected to the Internet, ~lthough many users may be connected through a single host and some indlvid- ua| computers may act as multiple hosts~ CAGR is compound annual growth rate. Source: Telcordia Technalogies (www~netsizeccom) Design © TeleGeography, [nc. 1999 123 TeleGeography 2000 © TeleGeographg, Inc. 1 999 Figure 3. Host Count by Country, 1997-1999 Count~//Region Sept. 97 Sept. 98 United States 16,667,000 26,167,000 Japan 1,158,300 1,704,800 Canada 964,400 1,665,000 United Kingdom 976,200 1,487,200 Germany 900,900 1,278,100 Australia 616,200 825,000 Taiwan 181,800 n.a. Netherlands 363,000 592,800 France 330,700 492,400 Finland 361,900 528,000 Sweden 324,100 424,600 Italy 217,500 361,000 Norway 189,600 327,900 Spain 168,600 256,300 Korea, Rep. 108,900 195,200 Denmark 145,400 217,700 Switzerland 157,200 252,900 Belgium 87,500 179,400 Brazil 91,600 180,200 New Zealand 134,300 196,500 I sra el 48,000 124,300 Austria 71,800 145,400 Hong Kong 96,800 133,500 South Africa 120,700 n.a. Mexico 23,400 84,200 Russia 61,200 135,900 Pola nd 75,700 112,500 Hungary 39,500 82,480 Argentina 18,600 44,100 Sing apore 78,000 82,400 Czech Republic 48,000 75,100 Turkey 19,500 42,200 Greece 31,100 44,800 Portugal 31,600 53,000 Malaysia 36,500 n.a. Chile n.a. 32,200 Ireland n.a. 43,300 Colombia 13,200 21,500 China 21,000 n.a. Thailand 16,100 27,600 India 4,600 14,900 Ukraine 16,200 23,100 Iceland n.a. 21,000 Estonia 8,300 14,700 Slovenia n.a. n.a. Slovakia 7,700 16,200 Indonesia 6,900 13,200 Romania 6,500 10,300 Uruguay 1,100 8,900 Sept. 99 44,23O,000 2,372,900 2,345,500 2,072,700 1,675,500 1,037,000 841,700 817,400 778,300 634,300 614,700 533,600 390,800 381,900 318,400 316,600 315,000 302,400 301 700 241 400 234 900 229 300 208 100 206 800 200 100 196 900 155,400 115,800 114,000 109,100 108,300 79,400 69,700 65,200 59,400 53,600 52,200 43,200 42,500 34,200 31,9OO 31,700 26,900 25,600 22,400 20,300 19,500 17,000 14,300 CAGR, 97-99 62.9% 43.1% 56.0% 45.7% 36.4% 29.7% 115.2% 50.1% 53.4% 32.4% 37.7% 56.6% 43.6% 50.5% 71.0% 47.6% 41.6% 85.9% 81.5% 34.1% 121,2% 78.7% 48.6% 30.9% 192.4% 79.4% 43,3% 71.2% 147.6% 18.3% 50.2% 101.8% 49.7% 43.6% 27.6% 80.9% 42.3% 45,7% 163.3% 39.9% 75.6% 62.4% 68.1% 61.7% 260.6% U.S./Canada 17,942,200 28,574,200 48,029,300 63.6% Europe 4,699,700 7,218,200 10,049,300 46.2% Asia/Pacific 2,507,490 3,317,400 5,551,000 48.8% Latin America/Caribbean 162,900 371,100 726,900 111.2% A fric a 120, 700 n.a. 206,800 30. 9 % Total 25~12,824 40~17, 459 64,177,265 59.5% Note: An |ntemet host ~,pJc ally represents one computer connected to the Internet,altflough many users may be connected through a~ sipgle host and some i~divid- ua[ computers may act as multiple hosts, Generic Top Level Domain (gTLD) hosts ere dlstributed between courrb’ies according to regist~al~ons in Septomber 1.899. CAGR is compound ann~ual growth rate, : Design © TeleGeography,!tnc. t999 Source: Teicordia Technologies (www, netsizer, com) © TeleGeography, Inc. 1999 TeleGeography 2000 VolP Traffic and Settlements The New Rules Although the vast majority of international voice and fax traffic is still carried over the public switched telephone net- work (PSTN) and subject to per minute settlement pay- ments between carriers, the rapid spread of lnternet proto- col (IP)-based networks has opened up new methods for transmitting international calls. Even as the old settlement rules become obsolete, however, voice over IP (VolP) carri- ers are actively seeking new rules to provide a stable service and compensation scheme. The need for new standards is being driven largely by two factors. First, as with the international PSTN, VolP carriers typically must rely upon other networks to complete a call; but on the Internet, the number of potential correspondents may num- ber in the thousands. Thus, a common set of ground rules is required for handing-off and metering traffic as well as for distributing revenues. Designing a viable method for shar- ing revenues presents a special challenge on the Internet, because large Internet service providers (ISPs) currently exchange traffic on a sender-keeps-all or peedng basis, and smaller ISPs generally pay larger ISPs for carrying their traf- tic by volume, regardless of the service (e-mail, phone calls, web pages) transmitted. As well, because most Internet phone calls terminate on the PSTN, the terminating ISP may need to pay local access charges to complete a VolP call. Thus to provide VolP carriers with an incentive to carry Internet calls and cover their PSTN termination costs, new Internet-based compensation schemes are needed. The second driver of VolP standards is related to the first: Quality of Service (QoS). To compete with the switched net- work, VolP carriers must offer comparable service in the longer run and that too requires cooperation. End-to-end VolP performance standards depend upon the quality of each network along the way. In order to see how the new interconnection and QoS rules are developing, this essay looks at two basic questions: how are VolP carriers currently exchanging voice traffic and how are they settling accounts with each other? Back~ound The first voice conversations carried on the Internet were low quality and relied upon rudimentary PC sottware for Figure 1. Major Clearinghouse Networks and Standards Settlement Clearinghouse Core Network Standard URL Arbinet Global Clearing Network none proprietary www.agcn.net AT&T Global Clearinghouse managed public Internet OSP/proprietary www.ap.att.com/clearinghouse Delta Three private IP links proprietary www.deltathree.com/business PSTN/Internet failover GRIC private IP links OSP www.gric.com GTE Internetworking private IP links proprietary www.bbn.com/products/voip.htm PSTN failover iBasis managed public Internet OSP www.ibasis.net PSTN failover iPass none proprietary www.ipass.com/services/clea ringhouse ITXC managed public Internet iNow www.itxc.com private IP/PSTN failover POPstar (IP fax only) none OSP www.pop-star.net Source; Company reports ©TeloGe~graphy, t~c, 1999 125 TeleGeography 2000 © TeleGeography, Inc. 1999 encoding speech at each end of the link. And although com- pression technology and faster computers have helped improve the quality of the experience, "PC to PC" phone calls remain a relatively low volume niche. But the idea had caught on. First, ISPs (and their vendors) built gateways that could terminate PC-originated calls to PSTN phone numbers. Then ISPs and alternative carriers set up gateways to originate phone calls on the PSTN, carry them on the Internet, and terminate them on the PSTN through another gateway. These companies, which could connect two PSTN end users via the Internet, became the first "Intemet Telephone Service Providers" (ITSPs). Unfortunately for the ITSPs, finding a way to terminate traf- fic on the PSTN is much more difficult than linking two PC users. In order to complete a call in another country, you need a point of presence (POP) there. So ITSPs began building relationships with ISPs, ITSPs, and carriers in places they did not already have a network node. This process soon became time-consuming and expensive, however, as the cost for establishing and maintaining hundreds of bilateral relationships was not something many newcomers could effectively do. Furthermore, monitoring the quality of con- nections was nearly impossible. One solution to the call termination problem was to create a self-contained, privately managed IP network which looks and acts like the public Internet but is capable of high-qual- ity voice transmission. Delta Three (now a subsidiary of RSL Corn) was one of the first companies to follow this model (although it relies on additional affiliate relationships as well). But whereas Delta Three uses a combination of facil- ities leased from its parent company along with leased lines from other carriers, the next generation of IP carrier will own most of its capacity, offering an end-to-end transport layer Figure 2. IP]PSTN Equipment Makers In order to exchange voice or fax traffic on IP networks, carriers need a tot of new equipment. Not surprisingly, the two biggest p]a~yers in next-generation switching te.ch~nolo~y are Cisco, the dominant maker of high-end IP touters, and Lucent, the dominant maker of hi,h-end PSTN switches. Both have ~everaged their existing products and mar- ket valuation to pay for buying sprees of smaller niche vendors that have had a jump start in the highly specialized needs of |P/PSTN gateways and associated equipment. So far Cisco has acquired an impressive group .of new companies, including Summa Four (for pro~ammabte switches), Lightspeed (for SS7 sisnatin~, and Transmedia [for next generation media switches), as well as Sel~ius and Calista. Lucent has likewise gobbled up Ascend, Nexabit (for terabit routing), and Excel [for pr~Irammab]e switches). Selected Vendor/Service Provider Alignments Vendor IP Service Provider 3Corn USA Globalink Cisco GTE Equant iBasis Telstre Telecom Italia Clarent AT&T Global Clearinghouse Ericsson Delta Three Lucent AT&T Global Clearinghouse GRIC Level3 VocalTec AT&T Global Clearinghouse Deutsche Telekorn ITXC Source: Philip Mutooni, iBaszs, Inc. Air’hough both companies have used acquisitions to bolster their product offerings, the two have diverged a bit on standards. Lucent is primarily backing the iNow profile with VocaiTec and ITXC. Cisco, however, has joined the growing raft of vendors and service providers backing the.Open Settlements Protocol. But Cisco and Lucent are certainly not the only forces in the nascent market for ]P!PSTN equipment. A small group of "soffswitch" mak- ers [a softswitch can handle multiple protocols and stahdards~, including Sonus, Salix, and Conve~ent Network~ have yet to be picked up by bigger players. Also, VocatTec and Clarent, two of the primary 1P-PSTN gatewaytgatekeeper vendors, remain (mosldlyi inde- pendent. No,tel, 3Com, Siemens, and other established players remain on the prowl for new acquisitions. VolP Access Concentrator Market Shares Source: Synergy Research Group, Jeremy Duke ©TeleGeography, Inc. t999 126 © TeleGeography, Inc. 1999 TeleGeography 2000 for IP services. Level 3 has based much of its business plan on this concept, and Concert, the AT&T/BT international joint venture, will likewise rely on its own network for IP ser- vices-although Concert has yet to lay out exactly how and when it will migrate its PSTN traffic to an IP network. The Clearinghouse A clearinghouse provides an alternative to building a net- work of global POPs. A clearinghouse can take many shapes, but at its core is a common approach: connect call originators with call terminators; keep track of the minutes; and create a standard billing format for shadng the revenue. The clearinghouse model enables anyone with the appropri- ate gateway equipment to originate or terminate calls in their local service area. The multilateral nature of the clear- inghouse also simplifies billing relationships and centralizes call monitoring. Some clearinghouses, like those developed by GRIC and iPass, originated from alliances formed to provide global roaming for lnternet access customers. Others, like those operated by iBasis (formerly VIP Calling) and ITXC, were specifically designed to offer VolP interconnection and wholesale services. Clearinghouses differ from one another primarily in the implementation of standards and equipment as well as in the role that the clearinghouse network plays (see Figure 1, "Qeadnghouse Networks and Standards"). Standards and Protocols On the PSTN, service providers rely on call detail records (CDRs) to resolve settlements issues. At this point, no sin- gle standard yet exists for IP call detail records, but they are still a key element to clearinghouse success. Furthermore, not all gateways used to connect IP networks to the PSTN are interoperable~which means that not all IP carriers can interconnect (see Figure 2, "IP/PSTN Equipment Makers"). Figure 3. How a VolP Clearinghouse Works Originating ITSP ",~!~,~ PSTN Simplified Clearinghouse Call Model Key Signaling Transpor~ =ssssssszss Billing PSTN 1. Originating Internet Telephony Service Provider (ITSP) accepts call, determines which clearinghouse to use. 2. Clearinghouse provides list of termination choices. 3. Originating ITSP signals terminating ITSP, call is initiated. 4. When call is finished, terminating ITSP signals originating ITSP and clearinghouse. 5. Both originating ITSP and terminating ITSP send call detail record to clearinghouse for reconciliation. 6. The originating ITSP (which charges its customer a retail rate) pays the clearinghouse either periodically or in advance; clearinghouse compensates terminating ITSP. Note: F~gum omits most ~TSP ~etwork details, such as hubs, switches, gateways, gatekeepers~ tmrder elements, radius servers, and setSernent servers. ©TeleGeography~ lnc, t999 127 TeleGeography 2000 © TeleGeography, Inc. 1999 Two prominent "open" clearinghouse mode{s are iNOW {pioneered by Lucent, VocalTec, and ITXC} and OSP {backed by Cisco, iBasis, GRIC, and TransNexus}. The iNOW {inter- operability NOW} model {which ITXC has used with successl relies on homogenous gateway components and a limited range of services (due to its reliance on the ITU’s H.323 standard). Alternatively, the OSP {Open Settlements Protocol) model is being standardized by the ETSI (European Telecommunications Standards Institute). The OSP system differs from iNOW primarily in its aim to host non-voice, next generation (XolP) services and its ability to route calls between clearinghouses. In addition, the two models use different protocols to transmit calls and call detail records {see Figure 3, "How a Clearinghouse Works"). Figure 4. A Guide to XolP Standards, Protocols, and Vendor Forums Standards Body URL Notes ETSI/’I’IPHON www.etsi.org Open Settlements Protocol provides ×ML-based I P traffic settlements Internet Engineering www.ieff.org Task Force (IETF) Major XolP Standards/Protocols OSP SIP RSVP RTP/AVT MGCP LDAP T.120 Session Initiation Protocol enables voice over IP gateways and client end-points Resource Reservation Protocol przoritizes packettraffic by use Real Time Protocol enables real-time transmission of audio and video (but doesn’t promise it) Media Gateway Control Protocol defines how different media (e.g., voice and video) will control data packets Lightweight Directory Access Protocol provides a universal address database for networks International www.itu.int Real Time Data Conferencing IAudiographics) Telecommunication Union (ITU) H .320 H .323 H .324 ISDN Videoconferencing Video (Audiovisual) communication on Local Area Networks Video and audio communications over low bit rate connections such as dial-up modem conneclJons Industry Forum URL Membership Notes International www.imtc.org Founded 1993, IMTC covers H.323 (and other ITU standards), iNow, and Multimedia currently 145 others Teleconferencing members Consortium (IMTC) Softswitch www.softswitch.org Founded 1999, Focused on SIP/MGCP and other internetworking Consortium currently 50 members technologies Internet & Telecoms itel.mit.edu Academic/corporate Covers technical, economic, and policy issues Convergence Consortium Industry Initiative URL Founders Notes Interoperabilib/Nowl www.imtc.org/ ITXC, Lucent, Standards-based IP telephony interoperability profile for (iNow) act_inow.htm VocalTec vendors and service providers based on H.323 IP Call Detail Record www.ipdr.org Jerry Lucas and 19 Goal is to define record protocol for IP traffic exchange and Initiative IIPDR) charter members billing and submit to standards bodies for discussion VON Coalition www.von.org Jeff Pulver and 22 Seeks to keep IP services as unregulated as possible and charter members educate consumers and the media about relevant technologies 128 © TeleGeography, tn~, t999 © TeleGeographg, Inc. 1999 TeleGeography 2000 In addition to "open" clearinghouses, some IP carriers have set up proprietary systems for call routing and reconcilia- tion. Obviously, the downside to a proprietary clearing- house lies in its lack of interoperability with other clearing- house systems. However, some major IP carders like Delta Three have been willing to take that risk in order to speed their market entry. Networks Another factor in differentiating clearinghouses is network architecture. Some clearinghouses act only as brokers for bilateral and multilateral traffic exchange. Others add more value (and cost) by backing up the clearinghouse with man- aged public Internet connections or private IP networks (again, see Figure 1, "Clearinghouse Networks and Standards"). Three basic forms of clearinghouse network strategies have emerged so far: 1. No core network: Some clearinghouses take a "hands off" approach to the business of physically connecting clearinghouse participants, sometimes known as media- tion. That is, the clearinghouse sets up the means for interconnection and reconciliation but lends no core net- work infrastructure to back up the clearinghouse. The Arbinet Global Clearing Network, for example, sets up VolP minutes deals between buyers and sellers in the same way it sets up PSTN exchanges (although IP traffic only constitutes two percent of the total traffic transiting its trading floor on average). 2. Managed public Internet network: In addition to set- tlement services, other clearinghouses add value to the system with managed Internet backbones. For instance, although they differ in method, both ITXC and iBasis put emphasis on routing traffic over monitored public Intemet links (i.e., voice calls are transported simultane- ously with web pages and other types of traffic). But because commercial quality VolP calls can only be reli- ably terminated in a handful of countries, both compa- nies have "failover" plans for completing calls using backup private lines, refile arrangements, or traditional PSTN correspondent relationships. In some cases, the effect of PSTN termination costs can be considerable. ITXC and iBasis have both cited the "failover" cost of ter- minating PSTN calls as a major short term obstacle to profitability. 3. Private IP network: Another, slightly different model, puts private IP links at the center of the network. In the networks implemented by Delta Three, GRIC, and GTE, traffic is primarily routed over private network infrastruc- ture which does not compete with public Internet traffic. (Level3 and Concert may follow this course too.) Although this method is more expensive than relying on the public Internet for transport, private links are easier to monitor for network congestion problems. Again, however, supplemental PSTN routing plans are essential to account for possible private network outages. Call Rating Since the early days of the Internet, ISPs have exchanged traffic at public and private peedng points where their net- works interconnect. And, as the Net has evolved, the com- pensation model has moved from a "sender keeps all" sys- tem to one where network quality, breadth, and bandwidth adjust the terms of interconnection (i.e., small ISPs usually pay big ones to connect). This trend has become acute in Figure 5. IP Carriers, Traffic, and IPOs Reporting IP Traffic Total Market Company Period (minutes) Revenues IPO Date IPO Goal Cap. Delta Three 1/98 6/98 n.a. $2.2 rn pending $55.7 rn -- (NASD: DDDC) 1/99 6/99 n.a. $4.3 m GRIC 1/98 6/98 n.a. $2.5 rn pending $50.0 rn -- (NASD: GRIC) 1/99 6/99 n.a. $3.1 m iBasis 1/98 6/98 1 m $0.2 m 11/10/99 -- $1.2 bn (NASD: IBAS) 1/99 6/99 48 m $6.0 m ITXC 1/98 12/98 5 m $1.9 m 9/28/99 -- $1.4 bn (NASD: ITXC) t/99 9/99 78 m $14.3 m Net2Phone 8/97 7/98 42 m $12 m 7/28/99 -- $582.2 m (NASD: NTOP) 8/98 7/99 229 m $33.3 m Note; T~tal, revenue figures include services, software, and ~quipmentin add~on to per minut~ cEargesfor IP voice ~nd fax services, Manet caps current to November 11,1999. Source; Company reports © TeleGeo~graphy, Into, 1999 129 TeleGeo{~raphy 2000 © TeleGeography, Inc. 1999 the case of voice traffic delivery. Because voice traffic is extremely sensitive to the speed and accuracy with which IP packets are delivered, the typical "best effort" transport of lntemet traffic is insufficient to replace switched service net- works. Traditionally, long distance service providers have defined call rating within the terms of usage and distance. In order for large volumes of voice traffic to move on to IP networks, a call rating system which takes quality of service (QoS) into account is a prerequisite. And, because IP networks can also transport non-voice but equally sensitive usage-based XolP traffic (i.e., multicast and video), some rating systems also take into account "class of service". Currently, howev- er, standards for exchanging XolP traffic are still in develop- ment (see Figure 4, "A Guide to XolP Standards, Protocols, and Forums"). The End Game In relative terms, VolP is still a small business. In 1998, IP calls accounted for approximately 300 million minutes of international traffic--less than 0.5 percent of the world’s traffic--and will probably not exceed one billion in 1999. AJthough this high growth rate may continue in the short term, the significant technical obstacles to IP-based voice services (i.e., quality, reliability, and interoperability} ensure that VolP will not erode PSTN traffic as fast as many ana- lysts have predicted. AJso, it is hard to ignore the fact that the world’s established base of telephone companies have sunk billions of dollars into equipment which uses SS7---a protocol for call set up and record keeping that is not designed to work with IP net- works. This means that some enhanced services integrated into SS7 (like caller ID, call forwarding, and local number portability) will not work if part of a call is routed over exist- ing IP gateways. But these problems, while significant, are concerned with what happens at the ends of the IP network, and are out- side the scope of this essay. We began with a question con- cerning what happens in the middle: How do IP carriers exchange traffic and settle accounts? Briefly, this can be answered thus: VolP carriers are using a combination of self-contained networks, bilateral correspondence, and multilateral clearinghouses to exchange traffic. Service providers settle with one another based on non- standardized VolP call detail records, which can (in some cases) encompass call quality, as well as distance and usage. Current VolP settlement systems are still evolving, and as yet, aro not profitable largely due to the cost of port- ing traffic to the PSTN where IP termination options don’t exist or aro unreliable. Because VolP carriers still carry relatively low volumes of traffic, are not always interoperable with each other, and have no standard method for transmitting next generation XolP services, much of this subject matter remains unpre- dictable. One thing is clear, however: the market believes IP will work, and its constituents have already put in enough time, money, and effort to guarantee that it will. TeleGeography gratefully acknowledges the technical advice of Philip Mutooni of iBasis, Inc. in developing this article. 130 © TeleGeography, Inc. 1999 TeleGeography 2000 Internet Governance: Then and Now Governance at the Core Dramatic growth has transformed the Internet, once an experi- mental research network, into both a quasi-public global infra- structure and a very big business (see Figure 1 ). By going glob- al and by going commercial, the lntemet’s management priori- ties have shifted. The result: innovation is giving way to stabil- ity and experimentation is sacrificed for predictability. Companies are less likely to take risks on untried, even if tech- nologically superior, designs. The lntemet, now comprising thousands of distinct private net- works, has always required some core management layer to track its existing interconnection standards and improve them when necessary. This "core" also has been responsible for allo- cating scarce resources. The registration of names, like www.telegeography.com, and network numbers, like 209.8.150.94,’as well as the operation of central "root serv- er" databases to link the names and numbers together, are part of a centrally-managed global address space. The management of these critical internetworking processes is at the heart of Intemet governance. Standards Despite the pressure for change, the Intemet still adopts tech- nical standards much as it has since the 1970s, when a forum called the lnternet Configuration Control Board (ICCB; see Figure 3) was formed to coordinate the process. The lntemet Activities Board (lAB, now the lnternet Architecture Board), which took the ICCB’s place in 1983 and has orchestrated the lntemet since, is a standards forum for the engineers which make the Net run. A somewhat shaky scaffolding for Internet governance has ~’own up around the lAB. First, in 1989, the ever-growing and increasingly unwieldy Board was subdivided. The lAB, still responsible for Inter-net standards, yielded increased responsi- Figure 1. Growth of lnternet Users and RFCs 160 Internet Users and RFCs, 1994-1999 3,000 140 120 100 80~ 60- 4O Requests For Comments (RFCs) Internet Users 1994 1995 1996 1997 1998 1999 ¯ 2,500 2,000 1,500 1,000 5OO Source: Computer Industry Almanac, RFCarchive © TeteGeography, lnc, 1999 Them am many ways to chart the intemet’s growth. Because they can be statistically sampled, coun~ting Intemet-connected host com- puters is one of the more popular ones; since 1994, the worldwide Intemet user population has leapt from just under 4 million to ove~ t 60 million. The fntemet’s routing table, provides another growth indicator. A routing table is a constantly updated road map which guides traffic between all those computers; in October t 999, the table contained as many as 70,000 routes---more than four times larger than it: was in I994. And a third handle on the expanding Intemet is via the venerable UseNet feed which distributes public dis- cussion groups around the wodd. UseNet has grown from 25 to 70 gigabytes in the last year alone, a far cryfrom the 700-odd megabytes (that’s under 0.7 GB) that a full news feed demanded five years ago. But to understand the consequences of fntemet scaling, take a took at who’s involved: the ratio of builders to usershas changed, in the t990s, the ]nternet’s user population, appears to have been growing at a compound annual rate of 80 percent, f~r outstripping the lntemet’s working papeis, called Requests For, Comments, "which" have.accumulated at t0.8 percent annually. As a new medium, the tntemet was a bubbling community of technical inventors, constant~ ly minventing the network. With widespread adoption, the goal for the Net’s manager is to ensure a robust operating environment. 131 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 2. The lnternet Governance Stack Internet Governance Stack, 1968-1999 Standards lAB RootAuthoriW IANA(don Postel) IANA U.S. G ov’t IANA (Jon Postel) Number Space NSl(InterNIC) ICANN RIPE, regional A’~NIC, regional ;~ ARIN, regional Name ~pace GSI-NSI (InterNIC) ICANN @TeleGeography, Inc. 1999 Them was a time when the Intemet was accused of being anarchic. That was never quite true. Anarchy seeks to abolish central author- ity; the lnternet disaggregated it, dividing the governance of its core functions into a sedes of fayers--a counterpart to networking’s seven-la3~er Open Systems Interconnect [OSf): stack. To critics, tCANN doesn’t have the look and feel of the ]nl~ernet, The Intemet governance stack shows why: ICANN’s scope works against the trend, aggregating multiple governance functions :within a single organiza~ tion. Good corporate strategy in a competitive market, perhaps--but will it play on the Net? Here are the layers: Backbones information moves across the 1nternet via !ong-haul networks, known as backbones. The first backbone was funded by the U.S. military’s Advanced Research Projects A~ency (ARPA], which earned it the name ARPANET.. The backbone fun .ction was taken Over by the U.S." National Sdence Fohndation (NSF)’s NSFNEF in 1987. Following the creation ’Of the" £~ommerdal lntemet exchange in -]991 ~.’commercia[traffic began .to.play an increasing.role, and by 1995, the NSFNETwent back to being a research network; commercial progiders had taken over the back- bone function. Today, the lilternet’s backbon~ consists of a series of inten:0n:necting:pdvate networks; Terms for in~erconnection are likewise a pdvate matter, although some governments have begun to take closer look at the market power of the largest backbone operators. Standards ~The lntemet’~ core feature is interoperability, which depends on agreeing on standards for implementing new technical features. Until Vint Cerf. then at ARPA, instigated the tnternet Configuration Control Board [ICCB)’s formation in }979, "the lead role in Internet s~ndards was played by the ARPANET Network Working Group, later known as the lnternet Workin~ Group. A series of reorganizations have replaced the.ICCB with ~e lntemet Activities Board {lAB, later the lnternet Architecture Board, !985), which was joined in ! 989 by the intemet Engineering Task Force (fETF) and Internet Engineering Steerir~ Group {IESG). Root Authority At ~e root of the lnternet’s number addressing and naming system iF the responsibility for coordinating theh" allo- cation and propagation through a distributed system of computer directories. For most of the Intei-net’s life span, that responsibili- ty fell to the late Dr. Jon Postel, first~ informally from his base at the University of "Southern California’s It/formation Sciences institute (tSI}, and then, following an attempt to aggregate this function into the activities of the SPJ Network Information Center "(see "Name Space", beiow), formally as director of the [nternet (continued) 132 © TeleGeography, Inc. 1999 TeleGeography 2000 bility to a subcommittee called the lntemet Engineering Task Fome (IETF), made up of eight functional areas, and to an lnternet Engineering Steering Group (IESG), made up of IETF area leaders. This symbiotic process--the IETF nominates twelve of the IAB’s thirteen members, and populates the IESG with its area leaders--took on its current institutional form in 1992 when the IAB was formally "chartered", or recognized, by the newly-conceived lnternet Society (ISOC) as the lnternet Architecture Board. ISOC is a U.S.-based professional membership society created to act as a non-governmental institutional and fund-raising umbrella group for the lnternet. A self-described "internation- al o~anization for global coordination and cooperation on the lnternet," ISOC’s creation marked the Internet’s first non-gov- ernmental coordinating body. As the seat for the standards process, ISOC has hinted over the years that it would like to be responsible for other areas of Internet governance, too. Its annual "Inet" conference (www.isoc.orB/isoc/conferences/inet) is meant to be the wodd’s yearly face-to-face Internet meeting. And it has sponsored key initiatives to enable Internet access to unconnected parts of the world. But ISOC’s constituent stan- dard-setting bodies continue to be the Society’s main contribu- tions to Interne[ governance. So how do technical standards get adopted? The process isn’t too different from an academic review process, and relies on a document series called Requests For Comments (RFCs, search- able at www.rfc-editoro~). There are four steps. 1. Informal circulation RFCs begin life as Internet-Drafts, allowing the internetworking community to provide com- ments and criticism pdor to formal submission. Anyone can propose an Internet-Draft by outlining their specification according to a well-defined format (detailed in RFC2223) and submitting it to internet-drafts@ietf.org>. The draft is then circulated on a general e-mail list, abstracted on www.ietf.org/lid-abstracts.txt, and placed for a six-month period in a public archive (ftp.ietf.org/intemet-drafts). After that, the author may decide to revise and resubmit the pro- posed standard as a new Internet-Draft, or to take it to the next step--submission. 2. RFC submission An author who wants to escalate the Draft to an RFC must specify an RFC category. There are two. A standards-track RFCs may either be approved as a Technical Specification (TS) or as an Applicability Statement [AS), depending on its scope. A non-standards-track RFCs must seek approval as Experimental, Informational, or Historic. A third type of RFC is referred to as Best Current Practice (BCP). BCPs are queue-jumpers: as a mechanism for area group heads (IESG members) or lAB members to make pro- grammatic statements, they cut through the process and head straight to approval and publication. Standards- and non-standards-track RFCs need to be reviewed first. Figure 2. Internet Governance Stack (continued) Assigned Numbers Authod~ (1ANA; see www.iana.or~/postel), which was chartered by the U.S. ~overnment (see the International Forum on the White Paper, www.ifwp.or~). ICANN acquired responsibili~ for root authod~ functions in t 999, with a U.S. gov- ernment transitionary oversight period to end in September 2001. Beyond numbering and naming, one of the functions ICANN absorbed from IANA is to coordinate the identifiers and other con- stants used by the standardized protocols developed by the IETF and similar bodies. That function-is helped along through the advice of the ICANN’s Protocol Supportin{~ 0r~anization [PSO~, subcommittee, .whose members include relevant standards bodies ¯ such as the lEE, W3C, 133j,. and ETSI. For details on ICANN’s pro- tocoi parameters role, see www.iana.or~Jnumbers.hEnl.and RFCs 2/.54 and 1700. Number Space The number~eg registry, assigns and records .Internet addresses. Until. t 992, numbering had be~n a~re~ated into the root authori~-y role played by ~on Poster, SRI, and IANA. "Network Solutions Inc. (NSI), a spin off of U.S. government con- tractor. Government Systems Inc. (GSt), was then contracted to hand them out, part of NSI’s operation of the ln~rNIC {Intemet Network Information Center, see "Name Space"). A series of regional regis~Ties were created to delegate numbers in Europe (RIPE, 1989), Asia (APN1C, 1993), and the Americas (ARtN, 1997}. ICANN was assigned the central numbering role in 1999. Name Space For most of the Internet’s lifetime, generic top level domain names like .corn and .net--and the host: tables that.pre- ceded them--were managed by government contractor SRI international (formerly the Stanford Research Institute) through the Network information Center it operated, usually known as NIC. This was delegated to a corporation called Government Systems Inc. in 1991, who span off Network Solutions Inc. as a subsidiary, continuing to operate the tnterNIC until 1998, when it re-branded the "service as Network Solutions (www.networksolu- tions.com) in anticipation of coming market competition ahd ICANN’s central role in the, process. Controversy on’tCh~N’s roie in domain names has run rampant. But the name space is only the most visible layer of the lnternet’s core governance stack. ¯ © TeleGeography,]flc, t999 133 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 3. ICANN: The Bottom Up ICANN’s Organizational Structure DNS0 Domain Name Supporting Organization 3 board members 1 board member AS0 Address Supporting Organization 3 board membe~ PS0 Protocol Supporting 0rganization 3 board members 9 board members "At-Large" Members (elected) Task Force on Funding Governmental Advisory Commitee DNS Root Server System-’~ Advisory Committee "__._Advison/Committee on .independent Review "~ ~Ad vi S°MWe mC ~emr~ih~t; e ° n "~ Note: Diagram shows Advisory Committees mandated by ICANN by-taws. It does not show Committees of the Board of Directors. It.ANN Organizational Glossary ICANN Internet Corporation! for Assigned Names and Numbers. Non-profit corporation acting ~s the tnternet’s central coordinating body; took over from the Internet Assigned Numbers Authority (www.icann.org). Dedsion-makin8 power lies with the 19-member Board of Directors. Three Supporting Organizations (SOs) provide substantive policy recommendations f~r specialized matters, and each sends three representatives to the Board.¯ These SOs are: DNSO Domain Name Supporting Organization Members include generic and country code domain name registries, gtobal registrars; ISPs, trademark and intellectual property interests, business inter- ests, and non-commerdat domain name holders, ASO Address Supporting Organization Members include: ARIN American Registry for.lnternet Nun~bers. Allocates addresses in the Amedca~ and southern Africa. Based in Chantilly, Virginia (www.arin.net). ¯ RIPE N¢.¢:- R6seaux IP Eur0. p~ens Network Coordination C~nter. Allocates IP addresses in Europe, former.Soviet Asia, the Middle East, ahd northern Africa [see list at www.ripe.net/lir/registdes/indices). Based in Amsterdam. APNIC Asia-Pacific Network Information Center, Allocates fP addresses in most of Asia and Oceania. Based in Brisbane, Australia [www.apnic.net]. PSO Protocol Supporting Organization Members indude: ~’lnternet Engineed ng Task Force (www.ietf.org): W3C World Wide Web Consortium, Develops standards for the Web. Based at Massachusetts institute of Technology in Boston, institut national de recherche en informafique et eq automatique in France, and K€=o’ University in Fujisawa,.Japan [www.w3:org). .flu international Telecommunication union. Specialized agency of the United Nations for coordination of:global telecom networks and services. Based in;Geneva (www, ituJnt). "ETSI -"European Telecommunication standb~~s tnstltute. Stakeholders’ forum .for coordinating’telecommunications equip- ment and services standards in Europe. Based in Sophia Antip~tis, France (www.etsLorg). "" ©TeleEeography, inc. 1999 134 © TeleGeography, Inc. 1999 TeleGeography 2000 3. Review Here the area group heads who form the IESG, or even an IETF working group itself, swing into action. If reviewing a standards-track (TS or AS) RFC, they designate it as a Proposed Standard, Draft Standard, or full-fledged lntemet Standard, based on the level of implementation and experience that stands behind the document. A non-stan- dards track proposal is adopted under the Experimental or Informational categories, or designated as Historic. 4. Publication Once approved, the document is edited and published on-line as an adopted RFC. For RFCs with the highest adoption level, this step involves a new name. lnternet Standards documents are streamed into the STD series; Best Current Practices documents are designated as BCP. There’s more to it, of course. As lnternet Protocol (IP) networks become a basic element of the global public network, vested interests are putting pressure on the process. Controversy erupted in October 1999 when the agenda for an upcoming IETF meeting included, for example, action on a U.S. Government proposal to use the sixth version of lntemet Protocol (IPv6)--the standard for the Net’s next operating sys- tem (see "Scaling," below)--to facilitate phone-tapping by law enforcement officials, based on a 1994 U.S. law which requires telecommunications network operators to accommodate lawful electronic surveillance. Corporate maneuvering is introducing pressure, too. Witness the dramatic rise in vendor participation in the process. Nonetheless, an RFC entitled "The Tao of IETF" is still a widely read reference point on the group’s spirit (RFC1718). And the ISOC-governed standard-setting process which configures the Internet remains open, asynchronous, informal, and collegial-- a native Intemet application. Address Space Allocation IP numbers are the 32-bit numeric addresses to which domain names correspond; they let routers know where traffic is com- ing from, or going. When a network device has an IP number, it is globally addressable, and when it doesn’t, it’s not techni- cally on the Internet. While the IAB/IETF and its predecessors were taking care of standards, massive blocks of address space comprising tens of millions of numbers were being handed out by the address space’s then-central governor, Jon Postel of the Intemet Assigned Numbers Authority (IANA), to a few compa- nies and universities who asked for them. The Net’s odginal architects first imagined around 100 hosts on the Internet, so earmarking enough space in each IP packet header to allow for about four billion addresses seemed more than safe. Yet as Net use surged in the early 1990s, engineers feared a quick depletion of addresses: not only had the Intemet grown beyond all expectations, but numbers had been assigned in large blocks, many of which went unused. To compensate, allocation of the address space is now severely restricted by the three regional re~stries which distribute them from headquarters in the world’s three telecommunication axes: Asia (www.apnic.net), the Americas (whose registry is also responsible for sub-saharan Africa, www.arin.net), and Europe (also responsible for North Africa and the Middle East, www.ripe.net). The current distribution of IP numbers under- scores the imbalance of power between the United States and the rest of the world, and between the large firms who were lnternet pioneers and newer companies. Governments like China’s complain that relative latecomers to the Net have fewer address resources, notwithstanding their population and eco- nomic needs. Others retort that address allocation is based on the geography of cyberspace, not physical topography or geopolitics. But everyone acknowledges that IP addresses are hard to come by, even with techniques like Classless Inter-domain Routing (CIDR), which disaggregates some of those huge blocks of num- bers, or judicious use of Network Address Translators (NATs), which hide entire private networks behind a single address. Address space is still a scarce and valuable resource. Enter the lntemet Corporation for Assigned Names and Numbers, with the reassuring acronym ICANN (see Figure 2). It is IANA’s successor, and the oversight body meant to referee the address space debate and set things straight for the future. It was created in 1998 as a U.S.-chartered non-profit organiza- tion with an international board. And although it was original- ly funded by loans from powerful Net companies like IBM and MCI, it is committed to representing a broad Internet con- stituency. ICANN manages IP addressing policy, domain name policy, and the central root servers linking them. Its mandate is to be international, bottom-up, and non-governmental, a counterpart to the already-existing Internet Society. Already, though, ICANN is facing challenges to the naming and numbering system. In August 1999, the European Telecommunications Standards Institute (ETSI) made a bold bid for a lead role on ICANN’s Address Supporting Organization, proposing to harmonize 1P addresses, domain names, and tele- phone numbers. ICANN responded with the time-honored tac- tics of diplomacy everywhere: it created a committee to study the issue and report back in a year. Until now, the Internet’s success has been based on its ability to continually develop in radical ways. When Jon Postel was the sole arbiter of names and numbers--a role he, and IANA, played until Postel’s death in October 1998~ecisions were made quickly and new ideas encouraged. The lack of formal structure freed the process from bureaucracy, and decisions were based on en~needng arguments, not lobbying. But while the process was relatively straightforward as an engineering matter--little vanity or power politics were involved--it was never institutionalized until the Internet was a billion dollar TeleGeography 2000 © TeleGeograph~l, Inc. 1999 industry. ICANN’s fundamentat contradiction is in seeking to retain that fast-moving and free-wheeling process even as it is controlled by big business, which has an inherent interest in seeing that radical developments come slowly so that com- merce can adapt. As 2000 rolls around, corporate maneuvering and vitriolic pol- itics have taken center stage. ICANN’s Government Advisory Committee was savaged by critics who denounced it as a plan to regulate cyberspace [see summary at www.cookreport.com/icannreBulate.shtml). And the United States’ House Commerce Committee in July 1999 held hearings under the highly-charged title, "Is ICANN Out of Control?" Scaling the Network Remaining usable address space continues to shrink, while the huge routing tables which contain traffic directions continue to grow. If international IP networks are to continue to interoper- ate--and the resulting lnternet to keep running--then ICANN, or its successor, must quickly provide an answer to the Commerce Committee’s question. Partly to deal with address and routing problems, the IETF developed a new, sixth version of the Internet Protocol in 1994. Lacking incentive to compen- sate for the cost and time it takes to upgrade, most networks have balked at adopting it: networks on the public lntemet are struggling just to keep up with growth. Guiding the difficult implementation process, and working collaboratively with its ISOC organization counterparts to do so, will be one of ICANN’s biggest challenges. IPv6 will replace the current IPv4 (IPv5 exists experimentally) with a 128-bit numbering space. That opens up more than 2 28 addresses (that’s 340 trillion trillion trillion], which should be enough to go around. IPv6 also includes a flexible, hierar- chical numbering scheme to simplify routing table management, as well as security mechanisms, quality of service and multicast provisions, and other improvements. While some research net- works participate in the 6bone (www.6bone.net) and other test beds, few expect widespread IPv6 deployment before 2003. Microsoft’s influential Windows 2000 product, for example, is scheduled to ship without an IPv6 implementation. Around the corner is a fleet of networks and businesses ready to deploy huge blocks of IPv6 space, however. As more and more information and communication devices link to the lnternet to reap the benefits of interoperability, mobile tele- phony operators are in a golden position, deploying data appli- cations over hybdd phone/computing devices. With more than 300 million mobile telephone subscribers--dwarfing the lnternet’s 65 million hosts, or even its 200 million users--a simple move to equip all mobile customers would necessitate the switch to the newer protocol. And a deluge of IP addresses could be used elsewhere, like tele- vision video game consoles and cable-passed households, automobiles, and various already-existing appliances. A lot of companies will be scrambling to get hold of huge swathes of IP address space; the process began in July 1999, when IANA del- egated the first IPv6 address space to the three regional reg- istries (see www.ipv6.org). The trend will snowball: as machine- to-machine traffic begins to move onto the lnternet, the killer app for IPv6 might turn out to be the ]east intuitive, such as net-enabling some alarms in every hotel room, or locks on win- dows and doors. Provided the cost of obtaining IP address space is not excessive, its deployment is finally in sight. For network operators, that holds out hope for seeing the squeeze on address space loosen, routing table overload lessen, and the U.S.-centric name space diffuse. For telegeog- raphers, it is a numbering explosion that will complete the fusion of physical and virtual layers~an internetworked globe where electronic terrain is part of the built environment. If this scenario is to take place successfully, management of the lnternet’s core must adapt to its new operating environment. That will mean learning to scale the legacy governance models of the standards process as new bodies like ICANN come online. Representing the new and ever-growing cast of Intemet stake- holders, ICANN was to be an entity that, unattached to tradi- tional geopolitical diplomacy, could move more quickly than governments do. But priorities have shifted--rather than a perpetual revolution, today’s Internet is moving toward slower evolution. The question is whether governance of the Intemet’s shared resources can or should evolve with it. 136 Traffic Analysis TeleGeography 2000 © TeleGeography, Inc. 1999 138 © TeleGeography, Inc. 1.999 TeleGeography 2000 Overview of International Traffic Trends Telecom executives and government policy-makers heralded 1998 as a year of great change. Country commitments placed with the Wodd Trade Organization (WTO) and liberalizing direc- tives of the European Union (EU) were to open two dozen national markets to competition. Most believed that new car- ders would drive down pdces and stimulate traffic volumes. Although prices did drop in 1998, call minutes did not increase apace. Globally, the annual growth rate of international call minutes actually fell--from 15 percent in 1997 to 12.5 percent in 1998. But this global statistic masks significant regional variations: Regional Patterns ¯ Europe. For Western Europe, growth predictions did hold true. Each of Europe’s major economies experienced higher rates of call volume growth in 1998 with the introduction of full scale competition. Germany served as a striking exam- ple: after a paltry 1.1 percent outgoing minute increase in 1997, the burst of competition in 1998 helped usher in growth rates above 14 percent. ¯ Asia. Compared to the wide open markets introduced to continental Europe, competition in Asia remained relatively limited. More importantly, however, the 1997 Asia financial crisis finally hit the international telecom service industry. Nearly every major market experienced noticeable declines in outgoing traffic growth rates (see Figure 1). Hong Kong, where outbound call minutes grew by nearly 10 percent, stands out as a significant exception. Almost all of this growth occurred dudng the fourth quarter of the 1998/1999 fiscal year--immediately after the introduction of competition for voice International Simple Resale (ISR). Still, overall traffic volumes to and from Hong Kong barely changed. Instead, most of the new outgoing call minutes were drawn from Hong Kong’s incoming traffic flows, as cus- tomers switched from call-back to Hong Kong-based leased line resellers at the beginning of 1999. ¯ Americas. Telephone traffic from Latin America and the Caribbean increased 20 percent, mostly due to price cuts by incumbents. Call minutes from the U.S. and Canada, how- ever increased by a relatively anemic 9 percent. The rea- sons for this apparent drop are not entirely clear. Traffic sta- tistics from the U.S. and Canada include some refile traffic; many countries may have shifted their reodBination traffic away from North America to other hubs as new markets opened in 1998. More likely, the slower growth rate reflects maturing telecom markets. With call charges already low, Figure 1. International Traffic in Asia, 1996-1998 Annual Traffic Change for Selected Countries 30.0% ¯ 1996-1997 25.0% 1997-1998 20.0% 15.0% 10.0% 5 0% 0.0% Indo Korea -5.0% -10.0% China Singapore Hong Kong Hong Kong Traffic, April 1998-August 1999 250 Outgoing Traffic 200 = 150 Introduction of Competition 5o- o Apr-98 Jul-98 0ct-98 Jan-99 Apr-99 Jul-99 Source: Office ofthe Telecommunications A~athorit7 (Hong Kong), TeleGeogrsphy, inc. © TeleGeography, tnc, t999 139 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 2. Traffic Base of Incumbent and New Carriers Traffic Base of Top Carriers vs. Rest of World, 1998 Traffic Base of Carriers Beginning Operations after 1989 12.0 13% 10.0 Global Market 8.0 Share of New Carners 8.0 "~ 7% 5% 4.0 2.0 0.03% O.O8% 0.0 1990 1991 3% 2% 0.31% 9% 1992 1993 1994 1995 1996 1997 1998 Note: Top carrfiers are ranked by outgoing international call minutes. AT&T, M Ct WorldCom, and Deutsche Telekom car(ted 27 percent of global trafficin 1997. "New Carrier Traffic" i~c[udes only carriersthat began facilities-based operation after t989. : Source: TeleGeography, Inc. © TeleGeography, Inc. t998 end users are no longer responding to price cuts with the same degree of increased demand as they did in previous years. Sluggish growth in the U.S. and Canada thus damp- ened global growth considerably. From 1990 to 1997, U.S. carriers’ share of worldwide international call minutes grew from 22 to 28 percent. In 1998, the U.S. share receded to only 26 percent of global traffic. Global Trends Despite regional variations, a number of worldwide trends did emerge: ¯ Shrinking elasticit~t of demand. The phenomenon of call volume growth lagging behind the pace of price cuts was not confined to North America. Many international carders experienced significant decreases in international service revenues as they slashed prices to meet competition. Telecom New Zealand’s experience stands out as an extreme example. Although lower tariffs stimulated a 45 percent increase in outbound call minutes, Telecom’s revenues on these calls dropped 24 percent. ¯ Shift to new carriers. One segment of the international telecom service industry experienced unambiguous growth in 1998: new competitors (see Figure 2). By offedng low rates on specialized routes, second tier carders began to take minutes--not merely relative market sham~away from incumbents. For the first time ever, AT&T and MCI WorldCom carded fewer international outgoing minutes on their own networks in 1998 than in 1997. Meanwhile, traf- fic carded by competitors and resold by AT&T and MCl WorldCom increased by 1.5 billion minutes. ¯ Alternative Traffic. The much anticipated Internet Protocol (IP) telephony boom did not happen in 1998. TeleGeography estimates that only 300 million international minutes were carded as IP traffic. ISR represented a much more significant departure from the public switched network. In 1998, carders sent heady 5 billion minutes over leased pdvate lines; much of this traffic was refile. The U.S. route stands out as particularly prominent. Canadian carriers sent over 40 percent of the country’s outbound call minutes as ISR or refi]e via the United States. It remains to be seen whether these minutes reflected a temporary phenomenon prior to Canada’s full market liberalization in October 1998 or a permanent trend to be replicated elsewhere. The following pages provide further detail on international traf- fic trends. Charts, maps, and tables review 1998 traffic on a global, regional, and country-by-country basis. An analysis of international calling to and from mobile networks examines some of the variables ddving international traffic on cellular phones. Finally, an overview of alternative settlement arrange- ments offers a snapshot of refile, call-back, ISR, and IP tele- phony traffic. 140 © TeleGeographyo Inc. 1999 TeleGeography 2000 Global Traffic Review Figure 1. International Traffic and Main Line Growth 2,000 200 1,750 1,500 1,250 1,000 750 500 250 Fixed main lines Cellular subscribers --o- International call minutes 175 150 125 100 75 5O 25 EE 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Note: Data include out-bound international traffic on public networks only, Projections assume 14% traffic growth, 6% main line growth, and 30% mobile sub- scriber growth annually. Source: tTU, TeleGeography, Inc. ©T~deGeographyo]nc. ~999 Figure 2. International Traffic, Revenue, and Subscriber Growth Historical Trend Slow Growth Same Growth Fast Growth CAGR CAGR CAGR CAGR Indicator 1989 1998 1989-98 2002 1998-2002 2002 1998-2002 2002 1998-2002 Calls (bn) 6.1 28.8 18.7% 48.8 14.0% 52.4 16.1% 56.1 18.1% Minutes (bn) 28.7 93.0 14.0% 146.3 12.0% 157.1 14.0% 168.4 16.0% per main line subscriber 52.9 110.1 8.5% 139.4 6.1% 146.1 7.3% 153.1 8.6% per main line plus mobile 52.2 80.0 4.9% 79.1 -0.3% 79.5 -0.1% 80.2 0.1% Revenue (US$ bn) 30.4 69.0 9.5% 73.2 1.5% 70.7 0.6% 67.4 -0.6% Assumptions Call length (rains) 4.3 3.2 -3.1% 3.0 -1.8% 3.0 -1.8% 3.0 -1.8% Price per minute (US$) 1.16 0.74 -4.8% 0.50 -9.4% 0.45 -11.7% 0.40 -14.3% Main lines (m) 497 845 6.1% 1,050 5.6% 1,075 6.2% 1,100 6.8% Mobile subscribers (m) 7 318 52.3% 800 25.9% 900 29.7% 1,000 33.2% Total subscribers (m) 504 1,163 9.7% 1,850 12.3% 1,975 14.2% 2,100 15.9% Note: 1989-98 based on reported data. 1999-2002 based on ITU and TeleGeography forecasts. Scenanos are as follows: 1. Slow Growth: Traffic growth slows as minutes move offthe public switched network (PSTN) and large markets mature. 2. Same Growth: Traffic growth continues at s~milar rate of last five years, assuming that faster rates of price-cutting keep traffzc on the PSTN. 3. Fast Growth: Traffic growth increases, assuming a faster growth rate of network subscribers and faster rates of pr~ce-cutting, plus a s~gnificant component of new demand created by international traffic generated from mobiles. Source: ITU World Telecommunication indicators Database, ]TU estimates, Telel]eography, Inc.. ©TeleGeography, Inc. 1999 141 TeleGeography 2000 @ TeleGeography, Inc. 1999 Figure 3. Intercontinental Traffic Flows, 1995 & 1998 1995 1998 Key Traffic Flows 2,500 1,000 500 700 (Mil/ions of MI’/’Ts) Each band is proportional to the total annuat traffic on the public network in both directions between each pair of countries. These maps shows all inter-conti- nental routes with an annual volume of more than 100 million minutes. The total volume of these routes in 1998 was 27.3 billion minutes, approximately 29% of global interna- tional traffic. ¯ © TeteGeogr~phy,’]nc, 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 International Traffic by Region Figure 1. Interregional Traffic Flows, 1998 Key All figures are g=ven in bdhons of Mi- nutes of Telecommunications Traffic forthe pubhc telephone neWvork. 0.5 5 Each band is proportional to the total annual traffic from one reg=on to another. Arrows =ndicate direction of traffic beWveen regions. Note: Those interregiona]~ traffic fiowstota124,8 billion minotes, That sum does not equal globaltotatof 93.0 billion minotes because (1)data set based on top 20 out- going routes for 110 largest countries only; (2) traffic within regions account for a further 45,8 bitlion minutes based on data set; (3) interreg[onal routes below ~g0 mil- lion minutes not show~. © TeleGeegraphy, tnc, 1999 Figure 2. International Traffic by Origin, 1998 Latin America & Caribbean Oceania Africa 1.9% Note: Global traffic was 930 bn minutes in 1998. © TeleGeography, Inc. t999 Figure 3. Traffic Growth by Region, 1997-1998 25% : 20% L 15% 10% 5% 0% Global Average U.S.& Asia Europe Oceama Africa Latin Canada America & Caribbean Note: Global traffic wes 93.0 bn minutes in ~998. © ~TeleGeegrsphy, tnc. 1999 143 TeleGeo{~raphy 2000 © TeleGeography, Inc. 1999 Figure 4. European Telecommunications Traffic Flows, 1998 SWEDEN FINLAND CZECH HUNGARY Key ~Traffic Flows, All figures are given in millions of minutes of te|ecommunications traffic for the public telephone network. The map shows all intra-European routes w~th a ~ombined 1998. volume of more than 120 million minutes. 600 400 200 Each band is, proper~onat to the total annual traffic on the public telephone network in "both directions between each parr of countries. (~)Total Outgoing Traffic The area of each circle proportional to the volume of the total annual outgoing traffic from each country, Balance of Traffic 0e ~outes where traffic in one dire~ct~on accounts .for more tha~ 60% of the tetel~ an arrow shows the direction ,mOst of ¯ e traffic f|ews~ © TeleGeograph¥, |no. 1999 144 © TeleGeography, inc. 1.999 Figure 5. Latin American Telecommunications Tralfi¢ Flows, 1998 TeleGeography 2000 MEXICO ELSALVADOR BOLIVIA ARGENTINA URUGUAY ©TeleGeography, Inc. 1999 Key All figures are given in millions of minutes of telecommuilicetions traffic, for the. public telephone network= The map shows al]’~outes within Latin America with a combined ~998 volume of more than ~0 million rain utes. Traffic Flows 100 50 10 Each band is proportionat to tite total annual traffic on the public telephone network in both directions between each pair of countries~ (~Total Outgoing Traffic ¯ ,0g0 ¯ Minutes The area of each circte is proportional to the volume "of the total annual outgoing t~offic from each country, Balance,of Traffic On routes where t~affic in one direction accounts for more than 60% of the total, an arrow ¯ shows the direction most Of thevaffic tiows. © TeleGeography, Inc. 1999 145 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 6. Asian Telecommunications Traffic Flows, 1998 AUSTRALIA Key All figures are g~ven in millions of minutes of telecommunications traffic for the public telephone netw6rk. The" map shows all intra-Asian routes with a combined 11998 ,volume of, more than 50 million ¯ minutes. © Te|eGeography~ fnc. t999 Traffic Flows ~200 50 Each band is proportional to the totsf annual t~ffic on the publi~c telephone n~etwork "i~ both ¯ directions between each pair of countries. OTotal Outgoing Traffic. The area of each circle is ¯ prop0~onaLto the volume of the tote| annual outgoing traffic from each country. Balance of Traffic 0n routes where t~affic in one direction accounts for more than 60% of the total, an arrow shows the direcl~on z~0St,,of, the traffic flows. 146 © TeleGeographg, Inc. 1999 TeleGeography 2000 International Traffic by Country Figure 1. Outgoing International Telephone Traffic Growth for Selected Countries, 1997-1998 30% 25% 20% 15% 10% Global Average ’, © TeleGeography, tnc. t999 Figure 2. Telephone Traffic Balances for Selected Countries, 1998 Countries with a traffic surplus (incoming exceeded outgoing minutes) Chart to Scale U.S. Deficit -2,000 -1,000 0 Mexico India ~. ...~-~ Germany Philippines Dominican Republic ~~ Pakistan Italy Egypt Turkey Jamaica Brazil Vietnam Indonesia Hong Kong [] Taiwan Korea, Rep. Australia Japan Switzerland United Kingdom United States Countries with a traffic deficit (outgoing exceeded incoming minutes) 1,000 2,0OO International Telephone Traffic Balance (Millions of Minutes) ©~TeleGeography, Inc. I999 147 TeleGeography 2000 © TeleGeography, Inc. 1999 Figure 3. International Traffic Indicators, 1998 International Traffic Outgoing Incoming (m minutes) (m minutes) Albania (a, c) 49.1 52.2 Algeria (c) 121.3 n.a. Andorra 47.4 32.2 Angola 27.3 22.3 Argentina 358.7 n.a. Armenia 56.6 94.3 Aruba 21.1 31.0 Australia (b) 1,690.0 n.a. Austria (c) 1,160.0 n.a. Azerbaijan (a, c) 42.9 46.0 Bahamas 63.5 90.0 Bahrain (a) 124.4 102.1 Bangladesh 41.8 196.2 Belarus (a, c) 176.1 193.5 Belgium (a) 1,460.0 n.a. Bolivia 31.6 76.4 Bosnia-Herzegovina (c) 94.9 159.2 Brazil (a) 545.8 806.9 Bulgaria 96.0 201.0 Canada (a) 4,805.0 n.a. Chile 259.4 n.a. China 1,711.5 n.a. Costa Rica (a) ’, 82.7 112.9 COte d’lvoire 57.3 46.6 Cyprus (a, c) 182.0 120.8 Czech Republic (a) 317.4 466.9 Denmark (a) 710.0 n.a. Dominican Republic 157.5 730.5 Egypt 127.3 475.3 El Salvador 43.1 149.2 Estonia (a) 75.1 79.2 Finland 410.8 n.a. France 4,115.0 n.a. Germany (a) 5,495.0 n.a. Ghana 28.9 100.8 Greece 681.3 710.1 Hong Kong (a, b) 1,879.8 1,833.0 Hungary (a) 298.3 374.5 India (a, b, c) 438.2 1,498.8 Indonesia (a) 324.5 434.2 Iran 177.0 185.7 Ireland (b) 885.0 n.a. Israel (a) 861.0 424.0 Italy (c) 2,846.0 n.a. Jamaica (a) 60.1 349.8 Japan (b) 1,895.0 1,575.0 Jordan (a) 122.8 178.9 Kazakhstan 118.9 137.5 Kenya 29.2 72.5 Korea, Rep. 907.7 719.4 Kuwait (d) 160.5 n.a. Latvia (a) 55.4 87.2 Luxembourg 293.8 242.8 Surplus (Deficit) Population Minutes (Out) Main Lines Minutes (Out) (m minutes) (m) per Capita (thous.) per Main Line 3.1 3.1 15.7 116 424.0 n.a. 30.1 4.0 1,600 75.8 (15.2) 0.1 632.0 33 1,432.6 (5,0) 12.1 2.3 72 378.0 n.a. 36.1 9.9 7,323 49.0 37.7 3.5 16.0 556 101.8 9.9 0.1 224.3 33 634.6 n.a. 18.7 90.3 9,580 176.4 n.a. 8.1 142.5 3,999 290.1 3.2 7.7 5.6 680 63.1 26.5 0.3 214.6 104 609,1 (22.3) 0.6 193.7 158 789.2 154.4 124.8 0.3 n.a. n.a. 17.3 10.3 17.1 2,490 70.7 n.a. 10.1 144.0 5,073 287.8 44.8 8.0 4.0 535 59.1 64.3 3.7 25.8 332 285.9 261.1 165.9 3.3 19,987 27.3 105.0 8.3 11.5 2,742 35.0 n.a. 30.3 158.6 19,241 249.7 n.a. 14.8 17.5 2,977 87.1 n.a. 1,255.7 1.4 87,421 19.6 30.2 3.8 21.5 660 125.3 (10.7) 14.3 4.0 170 337.1 (61.4) 0.7 263.2 405 449.8 89.5 10.3 33.8 3,741 93.0 n.a. 5.3 134.0 3,496 203.1 573.0 8.2 19.1 764 206.2 348.0 66.0 1.9 3,972 32.1 106.1 6.0 7.1 483 89.3 4.1 1.5 51.7 499 150.6 n.a. 5,2 79,7 2,855 143.9 n.a. 59.7 68.9 34,000 121.0 n.a. 82.1 66.9 46,500 118.2 72.0 19.2 1.5 106 273.7 28.8 10.6 64.3 5,538 123.1 (46.8) 6.7 281.1 3,729 504.1 78.2 10.2 29.1 3,423 86.6 1,062.6 982.2 0.4 21,594 20.2 109.7 206.3 1.6 5,572 58.2 8.8 65.8 2.7 7,355 24.1 n.a. 3.7 246.4 1,600 553.1 (237.0) 6.0 110.5 2,800 236.1 n.a. 57.7 45.8 25,986 101.6 289.7 2.5 23.7 419 1 43.3 (320.0) 126.2 15.0 60,381 31.4 54.4 6.0 20.5 403 304.4 18.6 17.1 6.9 1,818 65.4 43.3 29.0 1.0 272 107.4 (188.3) 46.4 19.6 20,089 45.2 n.a. 1.8 90.1 412 389.9 31.8 2.5 22.5 741 74.7 (51.2) 0.4 693.4 293 1,002.4 Notes: Data are in millions of minutes of public switched traffic. a. International minutes based on bilhng point of traffic. b. International traffic for year ending 31 March. Australia and Pak~sten ends 30 June. c. Traffic data exclude some carners or routes. (See country table for details.) d Traffic data is for 1997, © TeleGeography, Inc. 1999 148 © TeleGeography, Inc. 1999 TeleGeography 2000 Figure 3. International Traffic Indicators, 1998 (continued) International Traffic Outgoing Incoming (m minutes) (m minutes) Macau (a) 125.2 95.8 Macedonia (c) 37.1 91.7 Malaysia (a, b) 685.0 n.a. Malta 37.3 n.a. Mauritius (b) 29.7 38.5 Mexico (a) 1,310.0 3,060.0 Moldova (a) 55.8 90.3 Monaco 234.0 140.0 Morocco 158.0 460.0 Namibia 61.9 45.3 Netherlands 1,805.0 n.a. Netherlands Antilles 43.0 113.8 New Zealand (b) 580,0 n.a. Nicaragua (a, d) 40.4 52.5 Norway (a) 546.0 n.a. Oman (a) 92.9 74.2 Pakistan (b, c) 87.5 640.4 Palestinian Authority 27.6 16.6 Panama (a) 50.0 97.6 Paraguay 37.8 57.4 Peru (a, b) 90.4 272.6 Philippines (a, b) 295.0 n.a. Poland (a) ’ 602.4 1,144.2 Portugal (a) 460.2 711.5 Qatar 112.5 70.0 Romania (d) 110.8 278.6 Russia (c) 1,038.3 1,029.8 Saudi Arabia 932.6 n.a. Singapore (a, b) 1,235.0 n.a. SIovak Republic (a) 151.8 186.4 SIovenia 129.6 137.0 South Africa 405.0 n.a. Spain (a) 1,605.0 n.a. Sri Lanka 39.3 146.8 Sudan (c) 18.4 88.0 Sweden (a) 1,205.0 n.a. Switzerland 2,425.0 n.a. Syria (a) 103.0 n.a. Taiwan (a) 862.0 781.8 Thailand (a, c) 296.4 358.6 Trinidad & Tobago (a, b) 64.4 141.5 Turkey (c) 644.1 955.9 Ukraine 465.9 n.a. United Arab Emirates 874.8 n.a. Surplus (Deficit) Population Minutes (Out) Main Lines Minutes (Out) (m minutes) (m) per Capita (thous.) per Main Line (29.4) 0.4 294.5 174 720.1 54.6 2.0 18.6 408 91.0 n.a. 22.2 30.9 4,384 156.3 n.a. 0.4 97.1 192 194.6 8.8 1.1 25.9 245 121.1 1,750.0 95.8 13.7 9,927 132.0 34.4 4.4 12.8 657 84.9 (94.0) n.a. n.a. n.a. n.a. 302.0 27.9 5.7 1,515 104.3 (16.6) 1.7 37.3 101 615.2 n.a. 15.7 114.7 9,337 193.3 70.9 0.2 192.5 n.a. n.a. n,a, 3,9 148.8 1,868 310.5 12.1 4.4 9.3 128 316.2 n.a. 4.4 121.5 2,935 184.0 (18.7) 2.4 39.0 220 422.5 552.9 142.3 0.6 2,557 34.2 (11.0) n.a. n.a. n.a. n.a. 47.6 2.8 18.1 n.a. n.a. 19.6 5.2 7.2 289 131.0 182.3 24.8 3.6 1,655 54.6 n.a. 72.9 4.0 2,700 109.3 541.8 38.7 15.6 8,812 68.4 251.3 10.0 46.2 4,117 111.8 (42.5) 0.6 194.3 151 747.5 167.8 22.6 4.9 3,777 29.3 (8.5) 147.7 7.0 26,875 38.6 n.a. 20.2 46.2 2,878 324.0 n.a. 3.2 390.4 1,778 694.6 34.6 5.4 28.2 1,539 98.6 7.4 2.0 65.0 722 179.4 n.a. 44.3 9.1 5,075 79.8 n.a. 39.4 40.8 16,289 98.5 107.5 18.5 2.1 524 75.1 69.6 28.3 0.7 162 113.6 n.a. 8.9 136.1 5,965 202.0 n.a. 7.1 341.0 4,803 504.9 n.a. 15.3 6.7 n.a. n.a. (80.2) 21.9 39.4 11,500 75.0 62.2 60.3 4.9 4,827 61.4 77.1 1.3 50.2 266 242.3 311.7 66.7 9.7 16,960 38.0 n.a. 50.9 9.2 9,410 49.5 n.a. 2.4 371.8 915 955.8 58.9 139.5 32,072 256.5 270.5 92.2 172,452 144.7 3.3 23.8 824 95.1 23.8 3.9 1,490 61.6 23.2 7.1 2,712 60.6 77.6 0.7 2,000 28.0 10.6 20.6 2,319 94.7 12.7 4.2 212 249.0 United Kingdom(a, b) 8,225.0 6,400.0 (1,825.0) United States (a) 24,949.3 10,598.8 (14,350.5) Uruguay 78.3 97.0 18.7 Uzbekistan 91.7 74.7 (17.0) Venezuela (a) 164.5 298.1 133.6 Vietnam 56.0 334.0 278.0 Yugoslavia (a) 219.5 423.3 203.8 Zimbabwe 52.8 53.2 0.4 Notes: Data are in milhons of minutes of public swztched traffic. a. International mznutes based on billing point of traffic. b. International traffic for year ending 31 March. Australia and Pakistan ends 30 June. c. Traffic data exclude some carriers or routes. (See country table for details.) d Traffic data =s for 1997. © TeleGeography, Inc. 1999 1 49 TeleGeography 2000 © TeleGeography, Inc. 1999 International Traffic by Route Figure 1. The Top 50 International Routes, 1998 Countries Minutes each way Total Minutes 1. U.S. Canada 3,988.0- 3,400.0 7,388.0 2. U.S.- Mexico 3,117.6- 1,095.0 4,212.6 3. U.S. U.K. 1,531.4 1,260.0 2,791.4 4. Hong Kong China 1,165.0- 895.0 2,060.0 5. U.S. Germany 1,194.2- 395.0 1,589.2 6. U.K. Ireland 815.0- 605.0 1,420.0 7. U.S. Japan 833.3- 336.0 1,169.3 8. U.K. France 650.0- 485.0 1,135.0 9. U.K. Germany 725.0- 400.0 1,125.0 10. Austria Germany 535.0- 435.0 970.0 11. Switzerland o Germany 550.0 395.0 945.0 12. France Germany 470.0- 415.0 885.0 13. U.S. France 580.2- 290.0 870.2 14. U.S. India 772.9 52.4 825.3 15. Singapore Malaysia 437.0 350.0 787.0 16. Germany- Italy 410.0 385.0 795.0 17. U.S. Brazil 592.3 179.1 771.4 18. Netherlands Germany 395.0 375.0 770.0 19. U.S. Hong Kong 635.6- 70.0 705.6 20. U.S., Australia 459.7- 240.0 699.7 21. U.S.-’ltaly 532.2- 134.0 666.2 22. U.S. Philippines 573.1 85.0 658.1 23. Switzerland o France 380.0- 276.0 656.0 24. France Belgium 335.0- 315.0 650.0 25. Germany- Poland 405.0- 233.8 638.8 26. France Italy 329.0- 306.0 635.0 27. U.K. Spain 360.0- 270.0 630.0 28. U.S. Korea, Rep. 397.3 230.0 627.3 29. Belgium Netherlands 300.0 300.0 600.0 30. U.K. Italy 360.0 220.0 580.0 31. Germany Turkey 385.0- 186.4 571.4 32. Russia Ukraine 290.8 280.3 571.1 33. U.S. Dominican Rep. 452.9- 112.0 564.9 34. Switzerland Italy 320.0- 226.0 546.0 35. Canada U.K. 230.0- 310.0 540.0 38. U.K. Netherlands 305.0- 225.0 530.0 37. France Spain 260.0- 265.0 525.0 38. U.S. China 434.5- 80.0 514.5 39. New Zealand Australia 260.0 245.0 505.0 40. Spain Germany 265.0- 235.0 500.0 41. Australia U.K. 240.0 240.0 460.0 42. U.S. Taiwan 327.9 143.3 471.2 43. Taiwan China 243.1 180.0 423.1 44. U.S.- Israel 223.2- 185.0 408.2 45. U.S. Netherlands 273.2- 120.0 393.2 46. Japan China 253.7- 115.0 368.7 47. Korea, Rep. Japan 172.3 169.4 341.7 48. U.S. Jamaica 282.1 40.2 322.3 49. Belgium Germany 175.0- 145.0 320.0 50. U.S. Spain 230.9 76.0 306.9 Note: All data in millions of m=nutes of telecommunications traffic. The country which generates more traffic on each mute is listed first. The routeslisted above tota147.4 bilhon mmutes, equalto 51 percentof allinternationaltraffic. Data forAustraha, Hong Kong, lreland, Japan, Malaysia, New Zealand, Singapore, Ta=wan, and the U.K. are for fiscal year 1998/99. 150 © TeleGeographyo Inc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 u.s.- Mexico ~iii U.S.- Dominican Republic ~lli U.S.-Germany ~it~!i~ u.s.-sou U.S.- Malaysia ~;~;~ U.S.- Netherlands Antilles ~~~ U.S.-Hon, Kong I ~.S. Philippines ~~~ Japan China Germany Turkey France Morocco Germany Yugoslavia Switzerland Portugal Japan Philippines Korea, Rep. China Taiwan Japan Sweden Poland Qatar- India Canada Jamaica U,K. Pakistan Korea, Rep. Vietnam Canada India UAE Pakistan Italy Chile Saudi Arabia Egypt UAE Bangladesh Saudi Arabia Pakistan 12:1 Ratio of Outgoing to Incoming Traffic 151 TeleGeography 2000 © TeleGeography, Inc. 1999 100% 9o% 80% 70% 60% ~= 50% °’0% Ill o 30% || ~- o% Note: Country originating traffic listed first; country terminating traffic listed second. Some data based on billing point of ca ,and may not reflect actual route growth rates due to refile and c~H-back. © TeleGeography, Ir~c. t999 0% o -5% -- c) -10% o -15% ._o .~ -2o% -25% -30% 152 © TeleGeography, Inc. 1999 TeleGeography 2000 International Traffic on Mobile Networks: Current Trends and Pricing Few communications innovations of the last quarter century have seen as much success as the cellular telephone. Although introduced as a practical consumer device only in the late 1980s, mobile phones are now used by over 300 million peo- ple worldwide. At current growth rates, mobile phones will exceed fixed lines by 2007 (see Figure 1; for a country-by- country comparison of mobile subscribers, see the Blue Pages). As the number of wireless handsets increase, mobile networks play an ever larger role in originating and terminating interna- tional calls. Unfortunately for international carriers, high fixed- to-mobile termination fees make calling mobiles an expensive proposition. This article evaluates how international calls are made to and from mobile phones and how carriers and mobile operators are settling accounts. International Call Volumes and Mobiles Approximately seven percent of outgoing international calls were placed from mobile phones in 1998; a slightly higher per- centage of incoming international calls terminated on cellular networks. By 2001, one in ten international calls will originate on mobile terminals. A number of factors determine call vol- umes to and from mobile phones. Subscribership. The recent explosion of new cellular sub- scdbers has driven the increase of calls to and from mobiles. But far more important than the overall level of cellular tele- density in explaining this trend is the "substitution rate"--that is, the ratio of mobile to fixed subscribers. Many developing countries have built mobile networks as a less expensive means of providing access. For example, Cambodia now reports more cell phones than fixed-line sets. In developed countries, where individuals often perceive wireless as an optional supplement to their fixed-line set, mobile terminals are beginning to take their place as the primary instrument for voice telephony. Telephone numbers allocated to cell phones now outnumber those to fixed-line terminals in Finland, Italy, and Japan. In fact, the number of fixed connections has declined in each of these coun- tries. Regulation. While only 30 percent of the world’s countries have introduced competition for basic fixed-line services, over 60 percent permit competition for mobile services. In many countries, then, mobile operators serve as the sole competition to the fixed network incumbent. Subscribers in these countries appear to use mobile phones more intensively for basic tele- phony-including international calling--than users in countries with fixed network competition in place. In Portugal, for exam- ple, where the regulator granted mobile operators internation- al licenses one year before fixed network competition began, Telephone Network Subscribers at Current Growth Rates 1,600 1,4O0 1,200 Fixed 8O0 600 4OO 200 1990 1995 Mobile 2000 2005 2010 Percent of Total Subscribers and Outgoing International Traffic from Mobiles for Selected Countries, 1998 Finland 123.2 Italy 448.8 20.5 Japan 142.1 47.3 411.2 14.9 Switzerland Brazil Subscribers (m) ¯ International Traffic (m mins) Numbers show absolute levels of subscribership and call mins. 0.0% 10,0% 20.0% 30.0% 40.0% 50.0% 60.0% 153 TeleGeography 2000 © TeleGeography, Inc. 1999 Total Subscribers Fixed Mobile L. America & Caribbean Africa Total Outgoing International Traffic Originated from Fixed Originated from Mobile Asia & Oceania U.S.& Canada Europe 0.9 bn 0.3 bn fixed subscribers mobile subscribers (73 % of total) (27 % of total) 88.4 bn minutes 6.6 bn minutes from fixed from mobile (93 % of total) (7 % of total) the pementage of international calls placed from mobiles is double the world average. Price. The relatively high cost of calling from cellular phones often limits mobile-originated international traffic. In most countries, calls from fixed-line telephones remain far cheaper than those from mobile terminals. In the U.S., for example, a three minute call to the U.K. from a fixed line phone using AT&T’s One Rate plan would cost $0.30. With AT&T’s dis- counted high volume wireless package, that same call would cost $1.29. As mobile operators amortize the buildout costs of their rela- tively new networks and new licensees generate increased com- petition, mobile call prices are expected to fall. In some coun- tries, users can already find some mobile packages with inter- national call rates lower than standard fixed line plans (see Figure 3). As a result, these countries tend to generate a rela- tively high percentage of international calls from mobile phones. Fraud. Illegal "cloning" also contributes to international call traffic from mobiles. Cloning occurs when one cellular tele- phone’s electronic serial number (ESN) is programmed into another cell phone, so that call charges incurred on the imposter telephone are charged to the original owner Cloned phones thus offer individuals a handy tool for placing interna- tional calls free of charge. Brazil’s Embratel estimated that nearly half of its mobile-based international traffic originated from phones with cloned serial numbers. Worldwide, however, fraud likely contributes to no more than ten percent of outgo- ing international calls from cellular phones. Roaming. Mobile telephony gives subscribers the ability to use their handsets while roaming outside of their home region. At first glance, roaming should not contribute to mobile-based international calling. When a foreign visitor places a call on his or her cell phone, the call is picked up by the host operator’s network like any other domestic call. However, local calls account for less than 15 percent of calls made by roamers. Many mobile operators have found that over three quarters of calls go back to the subscriber’s home country. Roaming also increases international phone calls terminating on mobiles, as travellers leave their mobile number as a primary contact num- ber, and set calls to their home fixed-line phones to forward to their mobile sets while abroad. Roaming is particularly significant in Europe, Africa, and Asia, where the common Global System for Mobile Communications (GSM) permits interconnectivity. In Europe, around one-third of international calls from mobiles are generated by roamers while abroad. Worldwide, calls to and from roamers account for approximately a quarter of cellular-based international traf- fic. Widespread implementation of the planned third genera- tion (3G) wireless protocols would erode technical barriers to roaming and boost that percentage further. Cross-Border Mergers. The creation of multinational mobile operators by recent mergers could further stimulate interna- tional traffic to and from mobiles. These tie-ups--such as the 154 © TeleGeography, Inc. 1999 TeleGeography 2000 Vodafone-Airtouch merger and a potential Mannesmann- Vodafone deal--give operators the ability to offer end-to-end provisioning of international calls from handset to handset, cir- cumventing any termination payments. (For a list of top mobile operators by subscribership, see Figure 4.) As long as they pass along cost savings to their customers, the new multina- tionals will boost cross-border traffic volumes. Functionality. Over a third of the world’s phones are cellular, mobile call prices are falling, and international roaming serves as a central feature for mobile phones. So why do less than ten percent of international calls originate on mobiles? The answer lies in how mobile phones are used. Most individuals still employ wireless handsets as a supplement--rather than replacement--to their traditional fixed-line phone. Except in developing countries with limited fixed network deployment, cellular’s traditional niche has been casual, "on the go" con- versations; many people prefer to place relatively expensive international calls from fixed-line phones, where they have more control over call quality and background noise. In addi- tion, virtually all faxes originate from fixed lines rather than wireless sets--a significant gap, given that fax traffic accounts for at least 20 percent of international call minutes, depending on the country route. For the foreseeable future, then, the fixed-line telephone will remain the dominant medium for orig- inating and terminating international calls. Average Charges for Calls to European Union Countries, 1999 $4.00 ¯ Calls from Fixed-Line Phones Calls from Mobile Phones $2.50 Sl.5O $1.00 Italy Nether- Finland France Germany U.K. lands $0.00 International Calls to Mobiles: Who Should Pay? The spectacular growth of mobile subscribership has not occurred without cost. To a large degree, cellular users have Subscribers (m) Revenues (US$ bn) Operator Country 1998 1999 1998 1. NTT DoCoMo Japan 25.2 26.3 26.2 2. China Telecom China 23.6 n.a. 8.0 3. Telecom Italia Mobile Italy 14.3 17.0 n.a. 4. VodafoneAirTouch(b,c) U.K./U.S. 13.5 n.a. n.a, 5. SBCTelecom (d) U.S. 8.3 n.a. 3.7 6. AT&T U.S. 7.2 7.6 5.4 7. Bell Atlantic NYNEX Mobile (c) U.S. 6.6 7.0 3.5 8. Omnitel Pronto Italia Italy 6.2 9.0 2.5 9. SKTelecom Rep. of Korea 6.0 8.9 2.6 10. Mannesmann Mobilfunk Germany 5.0 5.9 4.3 11. T-Mobil Germany 5.8 7.7 3.5 12. DDI Japan 5.6 n.a. 7.0 13. FranceTOlOcom Mobiles France 5.5 7.0 2.8 14. Telef6nica Moviles Spain 4.9 6.7 3.1 15. BellSouth U.S. 4.8 5.1 2.7 16. GTE (c) U.S. 4.8 5.0 3.0 17. BT Cellnet U.K. 4.5 5.6 2.3 18. SFR France 4.2 6.0 2.9 19. AIItel U.S. 4.0 n.a. 2.1 20. Telstra Australia 3.1 3.4 2.5 Notes: a. Subscriber and revenue figures are for domestic operations only. 1998 figures are for year end 1998; 1999 subscribers are t~e most recent figures released by the company, b. Vodafone AirTouch merger was completed July 1999. 1998 subscribers are a pro forma estimate. c. Bell Atlantic and GTE Corp expect to complete merger by March 2000. Upon completion of merger, the wireless assets of both companies will be combined with Vodafone Airtouch’s U.S. wireless assets. Bell Atlantic-GTE will own 55% and Vodafone Airtouch 45% of the new company. The companies expect to complete the t~ansaction by 2001. d. SBC Wireless/Ameritech merger was completed October 1999 to form SBC Telecom. 1998 revenues are a pro forma estimate. 155 TeleOeography 2000 © TeleGeography, Inc. 1999 Figure 5. The Tromb of evading high fixed-to-mobile interconnect fees. Rather than sending calls directly to mobile operators, they have disguised their cellular- bound minutes as foreign calls by reoriginatint~ this traffic via another country. Like air through the slide of a trombone, calls are routed back to the country where they began, tn exchange for pay- ment of the standard international settlement rate--plus a small fee for sending the catls abroad---a domestic carrier can stick the home country international gateway operator with an expensive mobile interconnect bill. Somewhat anachronistically, this unusual form of refile has boosted the volume of traffic settled under the accounting rate regime at time when that regime is crumbling. But tromboning appears to be a short-lived phenomenon. Government authorities in the United Kingdom and France--where international carriers have complained vociferously about tromboning--mandated lower interconnect rates when international settlement rates fell below mobile termination charges. In addition, both British Telecommunications and France T~l~com have introduced higher settlement rates for calls bound for mobile networks, © TeJeGeography, Inc. 1989 paid the price of network buildout through subscription fees and relatively high per-minute call tariffs. More controversial- ly, fixed-line carriers also have paid via high fixed-to-mobile interconnection fees. Falling international settlement rates and stubbornly high inter- connect fees to mobiles have placed a squeeze on internation- al gateway operators seeking to pass on incoming internation- al calls to mobile operators. Settlement rates actually have fall- en below fixed-to-mobile termination fees in some cases, creat- ing unusual distortions of traffic flows (see Figure 5, "The Tromboning Effect"). International carriers face a number of options for alleviating this price squeeze: ¯ Negotiate lower fixed-to-mobile termination charges. Undoubtedly, this option is the first choice for international operators. However, interconnect payments are an impor- tant revenue source for mobile operators, and they have lit- tle incentive to lower termination fees. Lower charges often have required regulatory action. ¯ Eliminate the Calling Party Pays (CPP) system for inter- national calls. Fixed-to-mobile fees tend to be higher in markets (such as in Europe) where callers must pay for traf- fic to mobiles than in countries where individuals receiving calls pay. Removing CPP would remove mobile operators’ incentive to prop up interconnect fees. As both cellular com- panies and their subscribers no doubt would greet any manipulation of CPP with hostility, however, such proposals stand little chance of success. Indeed, it seems possible that CPP could spread to new markets. In the U.S., for example, where mobile users currently pay for incoming calls, the cel- lular industry is pressuring for a move to Calling Party Pays. ¯ Establish separate fixed-to-mobile termination charges for domestic and international calls. Some international car- riers have entered into informal agreements with mobile operators, permitting cellular companies to maintain a rela- tively high termination charge for domestic calls while obtaining interconnection at a steep discount on calls from international destinations. In Finland, for example, some international gateway operators pay mobile carriers an interconnect fee at one-fifth the official rate to connect calls from abroad. A drawback to these unofficial agreements is that they require the cooperation of all parties involved. Domestic fixed network carriers could undermine the system by tromboning calls to mobiles via third party countries. ¯ Implement a settlement rate averaging costs for calls to fixed and mobile networks. Although carriers lose money on incoming traffic to mobiles, they recoup their losses on calls to fixed networks. The drawback is that this strategy would create an incentive for foreign carriers to use the settlement rate system only for calls to mobiles, but to abandon the system in favor of direct interconnection when sending calls to fixed-line numbers. ¯ Introduce two different settlement rates for calls to fixed and mobile networks. Most carriers in countries with sepa- rate mobile operators have turned to this option in 1999. 156 © TeleGeography, Inc. 1999 TeleGeography 2000 Domestic Termination Rates to Fixed and Mobile Networks ¯ Fixed-to-Fixed Fixed-to-Mobile Mexico Germany Australia $0.01 $0.02 $0.03 $0.04 $0.05 $0.06 Termination Rate (U.S.$ per rain) Sweden Nether- lands U.S. $o.oo U.S. Wholesale Prices to Fixed and Mobile Networks $0.50 $0.30 $0,20 $0.~o ¯ Calls to Fixed Networks ~l Calls to Mobile Networks $0.00 Mexico Hong Kong Italy Germany U.K. Australia gle fee to send traffic to any phone within a country, held as long as national monopolies controlled access to all public net- works. The introduction of new network operators has forced incumbents to adapt. Former monopolies have responded to direct interconnection--made available to foreign carriers at major Points of Presence thanks to fixed network competition-- by offering discounted settlement rates on calls to principal cities. As options for terminating traffic increase, so too does the complexity of the international settlements regime. Under the new agreements, international operators charge foreign carriers at the normal settlement rate as long as traf- fic to mobiles remains at or below a certain benchmark level (usually, five percent of total call minutes). Beyond that benchmark, foreign carriers must pay a higher settlement fee. For example, U.S. carriers must shell out 25¢ per minute--10¢ per minute over the fixed network rate--to send traffic to mobile phones in Australia. Differentiated settlement rates protect carriers receiving international traffic from the price squeeze effect. They also pass on the high cost of fixed-to-mobile interconnection to foreign carriers. In turn, these carriers are transferring the higher costs to customers. In June 1999, one major U.S. wholesaler charged resellers a higher rate to send calls to mobiles in virtually every country where a mobile operator existed outside the fixed network incumbent (see Figure 6). Retail service providers thus face a quandary: do they incur the marketing and billing hassles required to pass on higher costs to end users, or do they absorb these costs on their own? The demise of the traditional settlement rate system reflects an even larger trend: termination fees on all networks are becom- ing more cost-based as competitors arbitrage away each oth- ers’ margins. Mobile is no exception. In the short run, mobile operators probably benefit as they face less competition than fixed-line carriers. But as regulators license more alternative mobile players, the comparatively high termination costs on wireless networks will come under pressure too. ~i~ One Country, Many Rates The introduction of dual settlement rates for calls to fixed and mobile networks reflects a larger trend in international telecom- munications-the demise of the "one country, one rate" settle- ment scheme. That system, whereby foreign carriers paid a sin- 157 TeleGeography 2000 © TeleGeography, Inc. 1999 Alternative Traffic Arrangements 1996 Total International Traffic ISR 1.5% 71.7 billion minutes 1997 Total International Traffic Refile 4.2% Call-Back 2.5% ISR 2.1% 82.5 billion minutes 1998 Total international Traffic Call-Back 2.2% ISR 5.4% Internet 0.3% 93.0 billion minutes 5O% 4O% 30% 20% 10% 0% 8O 87 United Kingdom ISR Traffic from U.S. ¯ 1997 1998 336 2,954 World Germany Average 1,556 Canada 50% 40% 30% 20% o% 95 71 ISR Traffic to U.S. ¯ 1997 ~1998 106 1,239 1,652 United World Germany Canada Kingdom Average 158 © TeleGeography, Inc. 1999 TeleGeography 2000 Cus o er dials international to terminating carrier in desti- I. Customer dials international 4. Customer calls hub country or number, nation country, access number, waits for ring third country. Hub country Call is routed over the Public 3. Call is delivered to its final des- then hangs up. makes settlement payment to Switched Telephone Network tination by terminating carrier. 2. Call-back carrier in hub coun- carrier in customer’s home (PSTN) by originating carrier try calls customer, country and to terminating car- which pays settlement charge 3. Customer answers phone and rier. has dialtone from call-back country. Refile 1. Customer dials international number. 2. Originating carrier sends call to hub country via PSTN or over international private line, known as International Simple Resale (ISR). 3. Refile carrier re-originates call over PSTN. 4. Call is delivered to final desti- nation via refile carrier, which pays settlement charge to ter- minating carrier. Internet Telephony 1. CuStomer dials international number. Call is routed over PSTN to gateway computer. 2. Call is converted from analog voice to Internet Protocol lIP) format and sent over the Internet to a gateway in termi- nating country. 3. Call is converted back to ana- log format. 4. Call is completed as a local call on PSTN. No international set- tlements are paid by the origi- nating carrier. 159 TeleGeography 2000 © TeleGeography, Inc. 1999 160 International Traffic Statistics TeleGeography 2000 © TeleGeography, Inc. 1999 Albania Destination Minutes (millions) Percentage of Outgoing Traffic ........................ ....... ,.-~ .... .... ...~ ~,,, ..... ~:::,~,~:~ ........... 38.2% ...................... ~*~ ",’~’I.’,.~ ~I~ : ~4 ~-~ ~-".~ .... ::*: ~:: ::~ ~’~: 35,3% 3. Germany 2.2 i i~I 4 6% ..................... 4. Yugoslavia ..................... 2.2 ~4 4.6% 5. United States .................. 2.0 4.1% 6. Slovenia ...................... 0.9 1.8% 7. Switzerland .................... 0.8 ~ii 1.6% 8. United Kingdom ................ 0.8 ~i 1.5% 9. Turkey ........................ 0.7 ~! 1.3% 10. France ........................ 0.6 ~i 1.3% 11. Macedonia .................... 0.5 ~il 1.0% 12. Austria ........................ 0.5 ~i 0.9% 13. Bulgaria ....................... 0.2 i 0.5% 14. Croatia ........................ 0.2 i 0.4% Other ......................... 1.5 ~i 3.0% Total ......................... 49.1 Minutes 1996 1997 1998 ............... !nc°~!ng ..................................................................................................................................... 4&0 ...................................... n.a.: ...................................................... 522 ............ ....... Outgo.i.ng ............................................................................................................................. 2&6 ..................................................... 4.!~:8 ......................................... 4& 1 ................................. .............. ~u[plus...(OefiF!t) ..................................................................................................... 25~4 ...................................... n,a, ................................................................... 3,.! ............... Total Volume 66.6 n.a. 101.2 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. 162 © TeleGeography, Inc. 1999 TeleGeography 2000 Algeria Minutes (millions) France ....................... 62.0 United Kingdom ................ 7.9 Italy .......................... 5.5 Morocco ...................... 5.1 Spain ......................... 4.7 Germany ...................... 4.0 Canada ....................... 3.9 United States .................. 3.4 Belgium ....................... 3.3 Switzerland .................... 2.5 United Arab Emirates ........... 1.9 Saudi Arabia ................... 1.7 Syria ......................... 1.3 Egypt ......................... 1.2 Lebanon ...................... 1.1 Netherlands ................... 1.1 Sweden ....................... 0.4 Greece ........................ 0.2 Denmark ...................... 0.2 Australia ...................... 0.1 Other ......................... 9.8 Percentage of Outgoing Traffic 4.2% ~!!i 3.9% 3.3% ~i 3.2% ~li 2.8% ~i 2.7% ~i 2.1% ~i 1.6% ~il 1.4% ~li 1.1% ~ii 1.0% ~i 0.9% !i 0.9% 0.3% O.2% 0.2% 0.1% Total ........................ 121.3 1997 163 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Andorra Minutes (millions) 33.0 8.1 2.2 1.0 0.3 0.3 0.2 0.2 0.2 0.2 1.6 Percentage of Outgoing Traffic 17.2% ~i 4.6% 2.1% 0.7% !ii 0.6% iii 0.5% ii! 0.5% i 0.4% i 0.4% ~! 3.4% 47.4 © TeleGeography, Inc. 1999 TeleGeography 2000 Angola Destination Minutes (millions) 1. PortuGal ...................... 12.2 2. South Africa ................... 4.1 3. United Kingdom ................ 1.2 4. United States .................. 1.1 5. Brazil ......................... 1.1 6. France ........................ 1.0 7. Namibia ....................... 0.7 8. Germany ...................... 0.6 9. Belgium ....................... 0.5 10. Spain ......................... 0.4 11. Netherlands ................... 0.3 12. Switzerland .................... 0.3 13. Italy .......................... 0.3 14. Zimbabwe ..................... I).3 15. Cuba ......................... 1).2 16. Mozambique ................... 0.2 17. Russia ........................ 0.2 18. Gambia ....................... 0.2 19. Canada ....................... 0.1 21). Nigeria ........................ 1).1 Other ......................... 2.3 Total ......................... 27.3 Mi[]utes 1996 1997 1998 .............. Surplus (Oeticit) ...................................................................................... (1~1) ................................ (3:1) ................ (+5+0) ............. Total Volume 35.1 40.7 49.6 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. 165 TeleGeography 2000 © TeleGeographg, Inc. 1999 (~Argentina 166 © TeleGeography, Inc. 1999 TeleGeography 2000 Armenia Destination Minutes (millions) Percentage of Outgoing Traffic 1. Russia ....................... 43.8 ~%~iI~ii~I~‘~iI~!~ii~i~i~iI~ii~i~iiiii~!~i~ii~i~ii!~i~i 77 4% 2. Ukraine ....................... 3.3 ~i 5.9%° 3. Georgia ....................... 2.3 ~1 4.1% 4. Belarus ....................... 0.6 ii 1.1% 5. Kazakhstan .................... 0.5 0.8% 6. Uzbekistan .................... 0.4 ~i 0.7% 7. Turkmenistan .................. 0.3 i 0.5% 8. Moldova ...................... 0.2 i o.3% Other ......................... 5.3 ~i~i~!~.~ 9.4)~ Total ......................... 56.6 1998 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Aruba Destination Minutes (millions) Percentage of Outgoing Traffic ................. ~=~ ~" ¯ ~=~ "~..~ ~’~,~ ~;~ ,-’."~ ;~ ~’~;~, 29.3% 2. Netherlands Antilles ............ 4.2 ~::~::~::~ ~~::~::~::~p~ 20.0% 3. Venezuela ..................... 2.8 .~:~ ........ 4. Netherlands .2.7 ~~ ~:~ .................. 5. Colombia .2.0 6. Dominican Republic ............. 0.9 7. Jamaica ...................... 0.2 0.9% 8. Suriname ...................... 0.2 0.9% 9. Canada ....................... 0.2 ~.;~ 0.9% 10. United Kingdom ................ 0.1 0.4% 11. Brazil ......................... 0.1 0.4% 12. Trinidad & Tobago .............. 0.1 0.4% 13. Haiti .......................... 0.1 0.3% 14. Mexico ....................... 0.1 0.3% 15. Peru .......................... 0.1 0.3% Other .1.3 .... ::~:~= 6.0~ Total ......................... 21.1 168 © TeleGeog~a phy, In Ci t999 © TeleGeography, Inc. 1999 TeleGeography 2000 Australia Destination Minutes (millions) Percentage of Outgoing Traffic 1 New Zea and 245 0 ~t~ ::!Ii:: ~i~%~~i~::::~:::::~:~ ::::~ ° 2 Un ted K ngdom 240 0 ~:::::~ ~~;~ ~;~%~ 14 2% 3. Un ted States 240 0 14 2 o ..... 4 S nga pore 75 0 ;;~ 4 4~. ....... 5. Hang Kong ................... 65.0 3.8Vo 6. Indonesia .................... 55.0 33% 7 Ch na 45 0 ~:::: ~)~;~ 27% .......................... ; ¯ 8 Japan 45 0 .... : 27Vo ~:.~:.~;~.~ ¯ 9 Canada 400 ~::::~::::~ 24% 10 Ph ppnes 400 24% ........................ ~::~:: ::~::~::~::;~ ¯ :.~:.~ ~:.~:.~:.~ ¯ 11 Ta wan 40 0 24% ......................... ~:: ~::~::~::~::~::~ ¯ 1 ’ ...... 13. Argentina .................... 35.0 2.1% ......... 2 Ma ays a .............35.0 .. 21 ~0 14 Germany 35 0 %~ 21% ....................... FY 1996/97 FY 1997/98 169 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Austria Destination Minutes (millions) Percentage of Outgoing Traffic 2. Italy ......................... 70.0 3. Yugoslavia .................... 46.0 ~+ 4+0% 4. Hungary ...................... 45.0 ~+ 3.9% 5. United Kingdom ............... 35.0 ~ii 3.0% 6. Poland ....................... 31.0 2.7% 7. Netherlands .................. 29,0 ~I 2.5% 8. Turkey ....................... 29.0 ~l 2.5% 9. Croatia ....................... 28.0 ~! 2.4% 10. United States ................. 27.0 2.3% 11. France ....................... 26.0 ~i 2.2% 12. Czech Republic ................ 26.0 ~I 2.2% 13, Slovak Republic ............... 17.0 ~! 1.5% 14. Romania ..................... 16.0 1,4% 15. Slovenia ..................... 16.0 Other ....................... 184.0 ~:~::+:~;~!~i~::~:::i~:~:::.~+:-~.:+ :: ~" +~ ........ .... 15.9% 170 Total ...................... 1,160.0 1997 1998 © TeleGeography, Inc. 1999 TeleGeography 2000 Azerbaijan Destination Minutes (millions) 1. Russia ....................... 22.8 2. Turkey ........................ 4.1 3. Ukraine ....................... 2.3 4. Georgia ....................... 1.3 5. Iran .......................... 1.2 6. United Kingdom ................ 1.0 7. Kazakhstan .................... 0,8 8. United States .................. 0.7 9. Germany ...................... 0.7 10. United Arab Emirates ........... 0.6 11. Belarus ....................... 0.5 12. Uzbekistan .................... 0.5 13. Turkmenistan .................. 0.3 14. France ........................ 0.2 15. Switzerland .................... 0.2 16. Netherlands ................... 0.1 17. Italy .......................... 0.1 18. Poland ........................ 0.1 19. Moldova ...................... 0.1 20. Norway ....................... 0.1 Other ......................... 4.7 42.9 1998 171 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Bahamas Destination Minutes (millions) Percentage of ~ut~oing Traffic 1. Un ted States 49 8 1~ ~i !!i~.~::ti::tiiiiifiiii::iit~iiiili~iii~i~i~i ~;~!::~::~::~::~ 78 3% 4. United Kingdom ................ 1.4 2.2% 5. Swi~erland .................... 0.6 0.9% 6. Haiti .......................... 0.4 0.6% 7. Turks & Caicos Islands .......... 0.4 0.6% 9. France ........................ 0.4 ~0.6% 10. Germany ...................... 0.4 ~0.6% 11. Trinidad & Tobago .............. 0.2 ~0.3% 12. Cayman Islands ................ 0.2 O.3% 3. Italy .......................... 0.2 0.3~o 14. Brazil ......................... 0.2 ~0.3% ,. S,a,n ......................... ...... 16. Bermuda ...................... 0.2 0.2% 17. Dominican Republic ............. 0.1 0.2% ~8. China ......................... 0.~ ~o.~% ~. Colombia ...................... 0.~ ~o.~% 20. ~uva.a ....................... o.~ ~o.=% Other ..... 3 3 5 2% 172 1998 © TeleGeography, Inc. 1999 TeleGeography 2000 Bahrain ,i es 198 124.4 1998 TeleGeography 2000 © TeleGeography, Inc. 1999 Bangladesh .... ’~ ’~ ~" ~@ ~: ; o 2 Un ted Kingdom 5 1 ,~ ¯ ..-~ .... ~: ..... .~: ’ ,~ ..: ~.:~::~ 122Vo .... 3. United States ...... 5 0 ~;~i~i~~;~::~;~;~;i~i 4 Saudi Arabia 3 5 8 5% I .................... "~~~~ .... ,~ ..... 5 S ngapore 2 8 ~~~~~ 6.8% ........... 6. Un~Ar~m~r~ ........... 2.~ ~~ 7. Hon Kon .......... 1.8 ~:~:-~,~:. 43~. ..... 8. Pakis{an ...................... 1.6 ~:~ 9. Malaysia ...................... 1.3 3.0% 10. Korea, Rep ..................... 1.2 2.8% ¯ ....................... ~;~L ......... 11 Ch na 1 0 23~o ~. ,,a,~ .......................... o., 13. Japan ........................ 0.9 2.1% 14. Canada ....................... 0.9 2.0% 15. Germany ...................... 0.8 ~;~ 2.0% 1 8°0 ~. ~,~ ...................... o.~ .~ ~7. ........................ 0.7 ~.~ ,. ..................... .............................. ~., ~:~:.~ ¯ .. 174 1998 © TeleGeographyo Inc. 1999 TeleGeography 2000 Belarus Minutes (millions) Russia ...................... 104.6 Ukraine ...................... 25.1 Poland ........................ 5.8 Latvia ......................... 4.0 Estonia ........................ 3.3 Lithuania ...................... 3.3 Moldova ...................... 2.2 Kazakhstan .................... 2.2 Armenia ....................... 1.2 Azerbaijan ..................... 1.2 Czech Republic ................. 1.1 Uzbekistan .................... 1.0 Georgia ....................... 0.8 Turkey ........................ 0.3 Kyrgyzsta n .................... 0.3 Turkmenistan .................. 0.3 SIovak Republic ................ 0.2 Tajikistan ...................... 0.2 Cyprus ........................ 0.1 Slovenia ...................... 0.1 Other ........................ 18.7 Percentage of Outgoing Traffic ~iiiiii ~s.4% !~~I~ 14.2% ~ii 3.3% ~i 2.3% ~!i 1.9% ~i~ 1.9% ~i 1.3% 0.7% 0.7% ~0.6% 0.5% o.4% o.2% o.2% 0.~% :~ o.~% o.1% o.1% ~o.~% Total ........................ 176.1 1998 175 TeleGeography 2000 © TeleGeography, Inc. 1999 f Belgium Destination 176 Minutes (millions) 1. France ...................... 315.0 2. Netherlands ................. 300.0 3. Germany .................... 175.0 4. United Kingdom .............. 140.0 5. Italy ......................... 75.0 6. Luxembourg .................. 50.0 7. United States ................. 50.0 8. Spain ........................ 48.0 9. Switzerland ................... 31.0 10. Sweden ...................... 18.0 Other ....................... 258.0 Total ...................... 1,460.0 Percentage of Outgoing Traffic ~l~ .~ 3 4% ~! 2.1°/o ~.2% © TeleGeography, Inc. 1999 TeleGeography 2000 Bolivia Destination Minutes (millions) 1. United States .................. 6.1 2. Argentina ..................... 5.3 3. Peru .......................... 3.7 4. Brazil ......................... 3.6 5. Chile .......................... 3.2 6. Ecuador ....................... 1.4 7. Colombia ...................... 1.0 8. Spain ......................... 0.6 9. Italy .......................... 0.6 10. Mexico ....................... 0.5 11, Germany ...................... 0.5 12. Paraguay ...................... 0.4 13. Canada ....................... 0.4 14. Venezuela ..................... 0.4 15. ,Japan ........................ 0.3 16. United Kingdom ................ 0.3 17. Uruguay ....................... 0.2 18. France ........................ 0.2 19. Switzerland .................... 0.2 20. Cuba ......................... 0.2 Other ......................... 2.4 Total ......................... 31.6 TeleGeography 2000 © TeleGeography, Inc. 1999 Bosnia-Herzegovina Minutes (millions) Croatia ....................... 25.5 Germany ..................... 19.4 United States .................. 5.2 Austria ........................ 5.1 Switzerland .................... 2.7 Italy .......................... 2.3 United Kingdom ................ 2.2 Sweden ....................... 2.2 Netherlands ................... 1.5 France ........................ 1.5 Norway ....................... 0.8 Denmark ...................... 0.8 Belgium ....................... 0.7 Canada ....................... 0.7 Spain ......................... 0.4 Australia ...................... 0.4 Russia ........................ 0.2 Greece ........................ 0.2 Luxembourg ................... 0.1 Poland ........................ 0.1 Other ........................ 22.8 Percentage of Outgoing Traffic !i 20.4% 2 oo, 2.4% 2.3% 1.6% ~i 1.6% o.7% 0,7% 0.4% o.4% 0.3% o.2% 0.1% 0.1% Total ......................... 94.9 178 1998 © TeleGeography, Inc. 1999 TeleGeo{~raphy 2000 Brazil Destination Minutes (millions) 1. United States ................ 179.1 2. Argentina .................... 41.8 3. Italy ......................... 28.2 4. Gerrnany ..................... 22.9 5. Portugal ............ ......... 22.3 6. United Kingdom ............... 21.0 7. France ....................... 20.0 8. Spain ........................ 16.9 9. ,Japan ........................ 14.3 10. Uruguay ...................... 11.6 11. Canada ...................... 11.3 12. Paraguay ..................... 10.4 13. Chile .......................... 9.6 14. Moldova ...................... 7.7 15. Switzerland .................... 7.6 16. Mexico ....................... 6.7 17. Bolivia ........................ 6.1 18. Netherlands ................... 5.6 19. Sao Tome & Principe ............ 5.2 20. Colombia ...................... 4.7 Other ........................ 92.8 © TeleGeography, !nc "1999 1998 179 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Bulgaria Destination Minutes (millions) Percentage of Outgoing Traffic I 3 Turkey 96 ~. .... .... ~.~ ....... ~: ¯ .. .~... ~. ¯ ..,,...~.~-,~ ............ 88~ 5. United States .................. 6.9 ~~~; 7.2% 6 ......................... 5.2 ~, ~~ 5.4~ 10 Romana 34 ~N" @ N35% 12 Ukra ne 3 0 NN~N~ 3 1% 15. Czech ~epub c 1 6 ~@~ 17% :/ ~6. Swi~erland .................... 1.4 1.5% 18 Cyprus 1 3 1 3% 180 © TeleGeography, Inc. 1999 TeleGeography 2000 Canada Destination Minutes (millions) Percentage of Outgoing Traffic 2. United Kingdom .............. 230.0 ~i 4.8% 3. Germany ..................... 75.0 ~i 1.6%o 4. Hong Kong ................... 65.0 1.4% 5. India ......................... 65.0 ~i 1.4% 6. France ....................... 60.0 t! 1.2% 7. Australia ..................... 40.0 !il 0.8% 8. Italy ......................... 40.0 0.8% 9. Philippines ................... 40.0 ii 0.8% 10. Japan ........................ 35.0 Ii 0.7% 11. Netherlands .................. 30.0 !i 0.6% 12. Taiwan ....................... 30.0 0.6% 13. Vietnam ...................... 30.0 0.6% 14. Jamaica ..................... 25.0 ! 0.5% 15. Korea, Bep .................... 25.0 il 0.5% 16. Mexico ...................... 25.0 ii 0.5% 17. Switzerland ................... 25.0 t 0.5% 18. Sri Lanka ..................... 25.0 0.5% 11.2% Other ....................... 540.0 Total ...................... 4,805.0 1998 181 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Chile ’ Destination Minutes (millions) Percentage ~f O~tgoing Traffi~ ..... .... 1 Un ted States ...... 95 8 2 Ar ent na .......... 29 6 .... ........ 3. Spain .............. 138 ,.3% 4. Peru ............. 10.3 5 Brazil ............ 9 6 .............. ~:.~ _.7 ,_ ..... 6. Germany .............. 6.2 ~...~ 24% 7. Canada ............. 6.0 23Fo Japan .......... 57 ~; 2.2% 9. Bo=ivia ........................ 5.2 10. France ........................ 4.2 1.6% 11. Mexico ....................... 4.2 1.6% 12. United Kingdom ................ 3.6 1.4% 13. Venezuela ..................... 3.4 1.3% 14. Ecuador ....................... 3.4 1.3% 5. Italy .......................... 3.3 ~;~;~ 1.3% 16. Sweden ....................... 3.0 1.1% 17. Colombia ...................... 2.1 0.8% 18. Australia ...................... 1.9 0.8% 19. Uruguay ....................... 1.8 ~’~ 0.7% 20. Paraguay ...................... 1.4 0.5% Other a~ n ~~:;~: ~::::~ .... ~, ............... ........................ ~.u ~::~::~::~::~ ~;~;~ To= ........................ zs9.4 © TeleGeography, Inc. 1999 TeleGeo{~raphy 2000 China Destination Minutes (millions) Percentage of Outgoing Traffic ...... 2. Ta~wan ...................... 180.0 ~II==i~t~;~l==~ 10.5~ 3. Japan ........ 1150 i~:~:~:~ ;;~:~:~:~:~:~:~ 6.7% 4. United States ................. 80.0 ~,~,,~i! 47% 5. Korea, Rep .... 400 6 Macau 350 .... 20 o ...... ......................... :~:~ 7. Singapore .................... 31.0 1.8% 8. ~ustralia ..................... ~6.0 ~. c~.~ ...................... ~.~ ~.~% ~. ~m~ ..................... ~., ~. u~t~, ~m ............... ~.o ~,.~ 1 2. France ....................... 11,0 0,6% ~3. Russia ........................ 9.0 0.5% ~4 ~.~ ................. 9.~ ~.~% I ~’"~ ...................... ~" ~~% Other ........ 2a6 5 ............................................... 1998 183 TeleGeography 2000 © TeleGeography, Inc. 1999 Costa Rica Destination 1. United States 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) ................. 30.6 Nicaragua .................... 15.2 Panama ....................... 5.2 Mexico ....................... 4.9 Guatemala ..................... 4.5 El Salvador .................... 3.5 Honduras ..................... 2.6 Colombia ...................... 2.0 Canada ....................... 1.3 Spain ......................... 1.1 Italy .......................... 1.0 Germany ...................... 0.9 Cuba ......................... 0.8 Venezuela ..................... 0.7 Argentina ..................... 0.7 Peru .......................... 0.7 Dominican Republic ............. 0.7 Brazil ......................... 0.6 France ........................ 0.5 Chile .......................... 0.5 Other ......................... 5.0 Total ......................... 82.7 1998 184 © TeleGeography, Inc. 1999 TeleGeography 2000 C te d’Ivoire Destination Minutes (millions) Percentage of~utgoing Tragic 2 Senegal 4 4 ~!:~~~ 7.6% ......................... 3 Un ted States ...... 3 6 4 Italy 3 1 5. Burkma F~sa ................... 2.~ 7. Nigeria ................ 2 a 7% ~0. Benin ......................... ~.4 ~;~,,~ ~.4g~ .......................... 12 Gabon 1 1 ::~ 2 0°/ .......................... 13. Belgium" 1.0 o ....................... ......... 14. Germany ...................... 1.0 15. Congo, Rap ..................... 0.9 ~;;~;; 1 6% 16 Cameroon ...... 09 ~?. Swl~erland .................... 0.9 1.6N ~8. 6hana ................ 0 8 ~. ~r~c~ ...................... 0.~ 20. Canada ....................... ~.8 Other ......................... 8.1 Xot~Z ......................... s7.3 1998 185 TeleGeo{~raphy 2000 © TeleGeography, Inc. 1999 ( Cyprus Destination Minutes (millions) Pe~centage of Outgoin~ Traffi~ 1. Greece ....................... 39.8 ~i~llllt~ii~liiiiltli~l!!~iI~1iii 21.9% 2 Umted K~n dom 36 6 3. Lebanon ............... 204 4. Egypt ........................ 16.4 :~:~ ~; 90~ 5. Russ=a ................. 8.3 6. Germany ......... 5 2 ~;~ 2.9% 7 Un ted States 4 9 %~ 2 7% .................... ::~ ¯ 8. Jordan ........................ 4.7 ~::~ 2.6~o 9. Romaine’ ...................... 3.3 ~=~1.8~oo .... 10. Bulgaria ....................... 3.1 ~% 1.7% 11. Syria ......................... 3.0 1.6% 12. Ukraine ....................... 2.9 1.6% 13. Italy .......................... 2.5 1.4% 14 France 2 2 ;~ % 1 2% .......................... 15. Yugoslavia. 1 9 ~:~ 11% 6. Sweden ....................... 1.9 17. Swi~erland .................... 1.7 0.9% 18. Netherlands ................... 1.7 0.9% 19. Austria ........................ 1.1 0.6% 20. Australia ...................... 1.0 Other ........................ 19.5 ~~~ 10.7% Tom~ ........................ 1~o ......... M iput~s ................................................................................................................................................. !9~ .................................................... !997 ................................................. !998 .................. ................. !nc.(~,m! ng .......................................................................................................................................................... 92~0 .............................................. 1.~ 5~ 2 ............................................... ! 20 ~6 ............................... .......... 0~g.o.i~g ....................................................................................................................................... !.2.8~ ........................................ ~!54.A ............................................ !.8~(~ ................ ................. ~ur p!~ s...( D e ~i cit) ................................................................................................... (36~ 6) ..................................... (.39~ 2) ............................................... (.6 !.A ) ........................... Total Volume 220.6 269.6 302.7 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Data exclude traffic to Turkey and some traffic to Greece. 186 © TeleGeography, Inc. 1999 TeleGeography 2000 Czech Republic Minutes (millions) 78.2 65.6 22.7 17.3 13.3 11.7 10.6 9.8 9.4 8.2 7.5 4.9 4.5 4.4 3.9 3.3 3.0 2.5 2.3 2.2 32.1 317.4 1996 1997 1998 187 TeleGeography 2000 © TeleGeographg, Inc. 1999 ( Denmark Minutes (millions) 120.0 110.0 89.0 67.0 33.0 30.0 27.0 21.0 15.0 15.0 14.0 13.0 12.0 11.0 8.0 125.0 710.0 Minutes 1996 1997 1998 ............ !n.€.0ming ............................................................................................................................ 600~0 .......................................... ,68).~0 ......................................... n:a~ .............................. .......... Outgoing ..................................................................................................................................................................... 573:2 .................................................. 607~5 ............................................ ?.!.0:0 ................ ........ S U [P!.~s (D.efic i0 ............................................................................................................... ~.~8 ......................................... 7~5 ....................................................... n~ a~ ............................ Total Volume 1,173.2 1,289.5 n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Totals from 1996 are for Tele Danmark only. 188 © TeleGeography, Inc. 1999 TeleGeography 2000 Dominican Republic Destination Minutes (millions) ~ercentag~ of ~utgoing ~ffic 2. Spain ......................... 5.3 ;!iiiiiii 3.4% 3. Germany ...................... 2.9 ~i~ 1.8% 4. Italy .......................... 2.7 ~1.7% 5. Canada ....................... 2.5 lt~ 1.6% 6. Venezuela ..................... 1.9 !ilii 1.2% ?. Ch,e .......................... 1.~ Ii’,11.o% 8. Switzerland .................... 1.5 ~i 1.0% 9. Netherlands Antilles ............ 1.2 ~i 0.8% 10. Argentina ..................... 1.1 ii 0.7% 11. France ........................ 1.0 10.6%o 12. United Kingdom ................ 1.0 ii 0.6% I o 3. Cuba ......................... 1.0 0.6~o 14. Haiti .......................... 1.0 0.6% ~. ~:o ....................... 0.9 0.6% le. Colo~bi~ ...................... 0.8 0.5% ~. ....................... o.e o.~% ~e. co~t~ ~i:~ ..................... o.~ ~o.3% ~. ~,~,~ ................... o.~ ~o.~ Other ........................ 17.2 ~::~ Total ........................ 157.5 1998 TeleGeography 2000 © TeleGeography, Inc. 1999 (~Egypt 190 © TeleGeography, Inc. 1999 TeleGeography 2000 El Salvador Destination Minutes (millions) 1. United States ................. 19.2 2. Guatemala ..................... 6.2 3. Honduras ..................... 2.2 4. Costa Rica ..................... 2.2 5. Mexico ....................... 2.1 6. Nicaragua ..................... 1.4 7. Canada ....................... 0.7 8. Panama ....................... 0.7 9. Colombia ...................... 0.3 10. Spain ......................... 0.2 11. 6ermany ...................... 0.2 12. France ........................ 0.2 13. Argentina ..................... 0.2 14. Italy .......................... 0.1 15, Venezuela ..................... 0.1 16. Chile .......................... 0.1 17. Brazil ......................... 0,1 18. Korea, Rap ..................... 0.1 19. Peru .......................... 0.1 20. United Kingdom ................ 0.1 Other ......................... 8.5 Percentage of Outgoing Traffic ~i 14"3°/° ~ii 5.2% 5.0% 4.9% ~4~ 3.3% 1.6% 1.6% 0.7% 0.5% 0.4% 0.4% 0.4% 0.3% 0.3% ~o.3~ 0.3% 0.3% 0.2% 0.2% Total ......................... 43.1 Minutes 1996 1997 1998 .......... !n cD~i. ~.g .......................................................................................................................... ! 60:5 ................................................................... 68:2 ......................................................................... 9:2 ....................... ..... 0!~!~go! ng .............................................................................................................................................................. 2&.6 ............................................. 343 ............................................. 43, ! ........................... Su[plu~ (Def!£it) .............................................. !31.9 .............................. 133.9 ............... 106.1 ....... Total Volume 189.1 202.5 192.3 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data are for ANTEL only. Other carriers originated an estimated 4 million minutes of traffic in 1998. 191 TeleGeol~raphy 2000 © TeleGeography, Inc. 1999 ( Estonia Destination Minutes (millions) Percentage of Outgoing Traffic ...................... ~~~t~::~!~ ~:.~~~f~::i~::~f::~i~ 28.0% 2. Russia ....................... 16.6 ~~~~ ~’~’~:: ~~ ~’~’~ 22.0% 3. Sweden .5.8 7.8% ...................... ;~ ........................ ~;~ ~. ~/o 5. Germany ,..4.7 ~~ 6.3% ................... ..................... ~,,~ o 6. Ukraine .3.5 ~;~;~;~:~ 4.6~o 7. Lithuania ...................... 2.9 3.8% 8. United Kingdom .2.1 2.8% 9. Denmark ...................... 1.8 2.4% 10. United States .................. 1.4 1.8% 11. Belarus ....................... 1.1 1,5% 12. Norway ....................... 1.1 1.4% 13. Netherlands ................... 1.0 1.3% 14. Italy .......................... 0.8 1.1% 15. France ........................ 0.8 1.0% 16. Poland ........................ 0.8 1.0% 17. Belgium ....................... 0.5 0.7% 18. Swi~erland .................... 0.4 0.6% 19. Spain ......................... 0.4 0.5% 20. Austria ........................ 0.4 0.5% Other ......................... 2.9 ~::;~ 3.9% 192 Total ......................... 75.1 1997 1998 © TeleGeography, Inc. 1999 TeleGeo~raphy 2000 Finland Minutes (millions) Sweden ..................... 130.0 Germany ..................... 37.0 United Kingdom ............... 36.0 Estonia ....................... 30.0 Russia ....................... 27.0 Norway ...................... 17.0 United States ................. 15.0 Denmark ..................... 12.0 France ....................... 12.0 Netherlands ................... 8.0 Other ........................ 86.8 Percentage of Outgoing Traffic 31.6% ~~ 6.~% I t o ~il 1.9% Minutes 1996 1997 1998 ......... !n~o~!.~g ................................................................................................ n.a~ ............................... n:a~ ................................ n.a. ......... Sur~plus (Deficit) ............................................................. n~.a.: ........................................... n.~.a.~ ....................................... n~a~ ......... Total Volume n.a. n.a. n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. 193 TeleGeography 2000 © TeleGeography, Inc. 1999 ( France Destination Minutes (millions) Percentage of Outgoing Traffic 1. United Kingdom .............. 485.0 ~~ii~~ 11.8% 3. Belgium .335.0 4. Italy .329.0 ....................... ~~~ 8.0% 5. United States ................ 290.0 6. Swifferland .276.0 ~~~ 6.7% 7. Spain ....................... 260.0 ~~~ 6.3~ 8. Morocco .................... 170.0 ~~ 4.1% 9. Po~ugal .189.0 4.1% 10. Netherlands ................. 141.0 3.4% 11. Algeria ...................... 135.0 ~E~ 3.3% 12. Canada .100.0 o .................... 2.4~ 13. Tunisia ....................... 92.0 2.2% 14. Turkey ....................... 52.0 1.3% 15. Poland ....................... 49.0 1.2% 16. Luxembourg .................. 46.0 1.1% 17. Sweden ...................... 34.0 0.8% 18. Austria ....................... 28.0 o.7% 19. Greece ....................... 28.0 0.7% 20. Denmark ..................... 27.0 0.7% Other ....................... 599.0 ~; ~;~ ~;~ ~~%~::~~:.~ 14.6% Total ...................... 4,115.0 Minutes 1996 1997 1998 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. © TeleGeography, Inc. 1999 TeleGeography 2000 Germany Destination Minutes (millions) 1, Austria ...................... 435.0 2. France ...................... 415.0 3. Italy ........................ 410.0 4. Poland ...................... 405.0 5. United Kingdom .............. 400.0 6. Switzerland .................. 395.0 7. United States ................ 395.0 8. Turkey ...................... 385.0 9. Netherlands ................. 375.0 10. Spain ....................... 235.0 11. Belgium ..................... 145.0 12, 6reece ...................... 135.0 13. Yugoslavia ................... 130.0 14. Czech Republic ............... 120.0 15. Denmark .................... 115.0 16. Croatia ...................... 105.0 17. Hungary ...................... 75.0 18. Portugal ...................... 67.0 19. Romania ..................... 65.0 20. Sweden ...................... 58.0 Other ....................... 630.0 5,495.0 1997 1998 TeleGeol~raphy 2000 © TeleGeography, Inc. 1999 Ghana 1. United Kin dom ............... 51 ~, ~. .~. .~ .~ ..~.,,~,~- .~ ,~ ,.~ ~, ~:~::: :,~. ~: 2 Unted States .................. 2.9 ~~~~::~ 100% 3. a y .......................... 2.8 ....... ~~~ 98% ......... 4 Lebanon 2 5 ~.~t~.~,,~~~,~ ==,=%~; ,, ...... 5. Germany ........ 1 9 ~;~~1~" 66% 6. Canada .. 17 ........... 7. South Africa ................... 1.3 ~~ 45% ..... ~. ~ra.c~ ........................ ~.~ ~. ~,an ........................ 0 ~~.~ 10. Netherlands 1 0 11. Nigeria ........................ 0.5 1.7% 13. Singapore ..................... 0.4 ~15% 15. Israel ......................... 0.3 O.9% ........................ ~.~% ~. c~. ......................... a.~ N~% 20. Zimb.b~e ..................... Other ....... 4 6 u.~ Minutes 1996 1997 1998 ............. 6~ig~i~ ......................................................................................................................................................... i~]~ ........................................................................... ~i:~ ................................................................ ~:~ ......................... .... ~i~ (~ ~i~i~ ................................................................................................................... ~:~: .................................................. ~:~: .................................................... )~:~ ................... ............. ................................................................................................................................................. ................................................................... ................................................. ............................... Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. 196 © TeleGeography, Inc. 1999 TeleGeography 2000 Greece Destination Minutes (millions) ~rc~ntaoe ~f Outgoing Tra~,ic~ 2 Un ted K ngdom 96 8 ~~ ’~;~’~ .~ o ....................... :-~ .... .~ ..... ,~.~*~.~.,. *~ ................... ~: ::~::::: 14 2Yo ......... 3. Italy ......................... 58.6 4. United States 40 3 .... ~:~:;~:~:~:~:~:~:~:~:~:~:~,,~:~:~ 5.9% 6 France .................. 28.0 41% 7. Cyprus ....................... 25.6 ~. Bu ~ar a .......... 2 ~,~; ~% 9 Boman a 19 4 ~’~ 28% ....................... I 12. Sw~erland .................. 12.2 ~::~::~::~ 1.8Yo ~s. Sw~on .................... ~’; ;;~-~ ~’~%o ~. U~ai~ ...................... ~0.~ ~.~ 1 5% 15, Canada ...................... 10.5 ~::~ ¯ ~. ~uss,a ....................... ~o.~ ~.~ 17. Australia ........... 98 1.4% 18. Austria ........................ 9.5 1,4% 19. Turkey ........................ 9.2 1.4% 2o. ~o~a.~ ........................ ~.~ ~.~% Total ........................ 681.3 1998 197 TeleGeography 2000 © TeleGeography, Inc. 1999 Hong Kong Destination Minutes (millions) 1. China ...................... 1,165.0 2. Philippines ................... 75.0 3. United Kingdom ............... 70.0 4. United States ................. 70.0 5. Singapore .................... 55.0 6. Canada ...................... 50.0 7. Japan ........................ 50.0 8. Taiwan ....................... 50.0 9. Macau ....................... 50.0 10. Australia ..................... 45.0 Other ....................... 199.8 Minutes FY 1996/97 FY 1997/98 FY 1998/99 ............. I.n.€om!.!~g ............................................................................................... 1~940~8 ........................................ ~.!00~3 ..................................... 1,833~0 ........................ .......... Out. g.~i.ng .................................................................................................................. !.., 7.38~ 6 ................................. !, 718:0 .......................... ~879:8 ............. .......... S U.~P !us (.Deft ci!;) .................................................................................................... 202:2 ..................................................... 38 3 ........................................ (46.~8) Total Volume 3,679.4 3,818.3 3,712.8 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Fiscal year ends 31 March. 198 © TeleGeographgo Inc. 1999 TeleGeography 2000 Hungary Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Germany ..................... 73.0 2. Austria ....................... 36.2 3. Romania ..................... 17.9 4. United Kingdom ............... 17.4 5. Italy ......................... 16.8 6. United States ................. 13.1 7. France ....................... 10.5 8. Netherlands ................... 8.2 9. Switzerland .................... 8.1 10. Yugoslavia ..................... 8.1 Other ........................ 87.0 Percentage of Outgoing Traffic 24.6% 6.0% ~~ 5.9°/o 5.7% 4.4% 3.5% 2.8% 2.7% 2.7% 29.4% Total ........................ 296.3 © TeleGeography, Inc. 1999 National Traffic Balance Minutes 1996 S~,~p,.! s,, ( B,e f!~ i,t! ......... !~6.0 Total Volume 546.0 1997 1998 324.6 374.5 287.1 296.3 37.5 78.2 611.7 670.8 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. © TeleGeography,]nc. 1999 199 TeleGeography 2000 © TeleGeograPhU, Inc. 1999 India Largest Telecommunications Routes, F’Y 1998/99 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Saudi Arabia .................. 91.1 United States ................. 52.4 United Kingdom ............... 36.2 United Arab Emirates ........... 35.0 Singapore .................... 18.2 Ornan ........................ 14.7 Kuwait ....................... 13.6 Germany ..................... 13.3 Canada ....................... 9.9 Hong Kong .................... 8.4 Japan ........................ 8.1 Australia ...................... 7.3 Sri Lank’a ...................... 6.9 France ........................ 6.7 Italy .......................... 6.6 Bangladesh .................... 6.1 Qatar ......................... 5.3 Malaysia ...................... 4.9 Pakistan ...................... 4.8 Philippines .................... 4.3 Other ........................ 82.8 Percentage of Outgoing Traffic 20.9% .. 12.0% 8.3% 8.0% 4.2% 3.4% ~3.1% ~3.1% 2.3% 1.9% 1.8% 1.7% 1.6% 1.5% 1.5% 1.4% 1.2% 1.1% 1.1% 1.0% 19.0% Total ........................ 436.2 ©TeleGeography, inc. 1999 National Traffic Balance Minutes FY 1996/97 FY 1997/98 FY 1998199 ¯ I,n,~m~g~ .............................................. 1,_0~:,0~, ~~ 1,256.6 1,498.8 0 ut,g,0i~og~ ....................... 384.2 ~;~ .......... 436.2 Total Volume 1,384.2 1,677.1 1,935.0 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Fiscal year ends 31 March. Data exclude some cross-border traffic wzth Bangladesh, Nepal, and Pakistan. 200 ©TeleGeograph~, Inc~ t999 © TeleGeograph~l, Inc. 1999 TeleGeoi~raphy 2000 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Indonesia Largest Telecommunications Routes, 1998 Minutes (millions) Singapore .................... 76.9 Australia ..................... 33.5 Malaysia ..................... 32.9 United States ................. 30.0 Japan ........................ 25.0 I-long Kong ................... 13.2 Taiwan ....................... 12.5 Korea, Flep .................... 11.0 United Kingdom ................ 9.1 Philippines .................... 7.2 Germany ...................... 7.1 China ......................... 5.6 Thailand ...................... 5.6 Netherlands ................... 5.2 France ........................ 4.4 India .......................... 4.1 Canada ....................... 3.2 Brunei ........................ 2.2 Italy .......................... 1.8 New Zealand .................. 1.5 Other ........................ 32.5 Percentage of Outgoing Traffic 23.7% 10.3% ~® 9.20/o 7.7% ~4.1% 3.9% 3.4% 2.8% 2.2% 2.2% 1,7°/) 1.7% 1.6% 1.4% 1.3% 1.0% 0.7% 0.6% 0.5% 10.0% Total ........................ 324.5 © Tel~Geography, Iqc. 1999 National Traffic Balance Mi[]utes 1996 1997 1998 ...... ~,~g o,i~,g ............... 280.2 351.6 324.5 v,S,.~r~!oU..~, (g e,.,f,!~!t) ............. 76.2 .......... !~,0,~.:~ ....................... ~,,0~:.7. Total Volume 636.6 807.6 758.7 Note= Data are in mdhons of m~nutes of outgoing public switched telecommunications traffic. Data based on bdling point of traffic. © Tel’~G~ography, inc. 1999 201 TeleGeography 2000 © TeleGeographu, Inc. 1999 Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United Arab Emirates ........... 27.3 Germany ..................... 18.4 United States ................. 16.9 United Kingdom ............... 10.5 Turkey ....................... 10.0 Pakistan ..................... 10.0 Kuwait ........................ 9.9 Canada ....................... 7.0 Sweden ....................... 6.0 Saudi Arabia ................... 5.4 France ........................ 4.5 Japan ........................ 4.3 Italy...’ ....................... 3.6 Azerbaijan ..................... 3.3 Netherlands ................... 2.3 Russia ........................ 2.2 Austria ........................ 2.0 Qatar ......................... 1.9 Switzerland .................... 1.9 India .......................... 1.6 Other ........................ 27.9 Percentage of Outgoing Traffic 15.4% ....... ~~i-~ .~" lO.4% 9.6% 5.9% 5.7% 5.7% 5.6% 3.9% 3.4% 3.0% 2.6% 2.4% 2.1% 1.9% 1.3% 1.3% 1.2% ~.1% 0.9% 15.7% Total ........................ 177.0 i © Tei~Geography, Inc, 1999 National Traffic Balance Minutes 1996 1997 1998 0 ut,~,,o.,~ ............................................................ ~,.3,.: ,2,,,, ................. .!,6_,0,,,7_ ............. ........ ,s, ,,u, r, p, ],,u s,,,,(D,,,.e,.f.!.,c,,!,t)" ......................................... n.a. (30.5) 8..8. Total Volume n.a. 290.9 362.7 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. ©Tel"Geography,’ln,~.e! 1999 202 © TeleGeography, Inc. 1999 TeleGeography 2000 Ireland Largest Telecommunications Routes, FY 1998/99 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Minutes (millions) United Kingdom .............. 605.0 United States ................. 80.0 France ....................... 25.0 Germany ..................... 25.0 Netherlands .................. 21.0 Canada ...................... 12.0 Italy ......................... 10.0 Spain ......................... 9.0 Belgium ....................... 7.0 Australia ...................... 6.0 Other ........................ 85.0 Percentage of Outgoing Traffic 68.4% 9.0% 2.8% 2.8% 2.4% 1.4% 1.1% 1.0% 0.8% 0.7% 9.6% Total ........................ 885.0 ©TeleGeography, inc. 1999 National Traffic Balance _~,~,~,~g, 0,i n,g ................................................................. ,~80,.~, 695.0 ............. 8_,~.5:~, .......... ..... ,~U,[plus (Deficit) .......... n.a. n.a. n.a:~ ...... Total Volume n.a. n.a. n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Fiscal year ends 31 March. Data for FY 1996/97 exclude some cross-border traffic to Northern Ireland. © TeleGeography, Jnc. 1999 203 TeleGeography 2000 © Telegeography, Inc. 1999 f lsrael Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. Minutes (millions) United States ................ 185.0 United Kingdom ............... 55.0 Canada ...................... 40.0 France ....................... 35.0 Germany ..................... 35.0 Italy ......................... 25.0 Russia ....................... 22.0 Jordan ....................... 21.0 Other ....................... 243.0 Percentage of Outgoing Traffic 28.0% 8.3% 6.1% 5.3% 5.3% 3.8% 3.3% 3.2% 36.8% Total ........................ 661.0 ©TeleGeography, inc: 1899 National Traffic.Balance i Minutes 1996 1997 !~i~9 468.1 425.0 0 ~.~g~.~ ............ 319.7 459.0 ..~[~lu s (Deficit) ............. .!.~4~ .............. (34.0) Total Volume 787.8 884.0 1998 424.0 661.0 (237.0) 1,085.0 Note: Data are in mdlions of minutes of outgoing public switched telecommunicatzons traffic. Data based on billing point of traffic. © TeleGaogra ~hy,~ln e. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Italy Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Germany .................... 385.0 France ...................... 306.0 Switzerland .................. 226.0 United Kingdom .............. 220.0 United States ................ 171.0 Spain ....................... 112.0 Romania ..................... 76.0 Austria ....................... 75.0 Belgium ...................... 69.0 Netherlands .................. 55.0 Poland ....................... 55.0 Greece ....................... 51.0 Morocco ..................... 49.0 Albania ...................... 36.0 Tunisia ....................... 35.0 Croatia ....................... 35.0 Canada ...................... 30.0 Chile ......................... 25.0 Russia ....................... 24.0 Sweden ...................... 23.0 Other ....................... 582.0 Percentage of Outgoing Traffic .. ,,’~ ~,, ~~ ~. ,:"14.6% "~.-:.-,~-~°. 11.6% 8.6% 6.5% 4.2% 2.9% 2.8% 2.6% ~2.1% ~2.1% 1.9% 1.9% 1.4% 1.3% 1.3% 1.1% 0.9% 0.9% 0.9% 22.0% Total ...................... 2,640.0 © TeleGeography, Inc, 1999 National Traffic Balance Minutes 1996 1997 1998 _~ r p~ ~,(~ ~i~), ............................ 129.5 123.2 ..... ~:~ .......... Total Volume 4,377.5 4,827.0 n.a. Note: Data are =n mdlions of minutes of outgoing public switched telecommunications traffic. Data exclude some traffic to France, Slovema, and Switzerland. © TeleGeog raphy, tflc= 1999 205 TeleGeography 2000 © TeleGeography, Inc. 1999 Jamaica Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7, 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United States ................. 40.2 United Kingdom ................ 4.5 Canada ....................... 4.3 Cayman Islands ................ 1.5 Trinidad & Tobago .............. 1.3 Barbados ..................... 1.0 Bahamas ...................... 0.8 Germany ...................... 0.5 Turks & Caicos Islands .......... 0.3 Saint Lucia .................... 0.3 Antigua & Barbuda ............. 0.3 Cuba ......................... 0.3 Guyana ’, ....................... 0.2 India .......................... 0.2 Netherlands ................... 0.2 Japan ........................ 0.2 Saint Vincent & the Grenadines...0.2 Italy .......................... 0.2 Dominican Republic ............. 0.2 Panama ....................... 0.2 Other ......................... 3.3 Percentage of Outgoing Traffic 66.9% 7.5% 7.2% 2.5% 2.2% 1.6% 1.3% 0.8% 0.5% 0.5% 0.5% 0.5% 0.4% 0.3% 0.3% 0.3% 0.3% 1 0.3% 1 0.3% 1 0.3% 5.5% Total ......................... 60.1 ©TeloGeography, inc. 1999 National Traffic Balance Minutes 1996 1997 1998 I nc~Q~jo~q 288.4 n.a. o ............... ~:8 ....... .... 0,~,,u,.,t,,9 o~,.n,g ............................... 64.5 ¯ ,, n.a. 60.1 Surplus (Deficit) 223.9 n.a. 289.7 Total Volume 352.9 n.a. 409.9 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on bdhng point of tTaff~c. © TeleGeography, Inc. 1999 206 © TeleGeography, Inc. 1999 TeleGeo{~raphy 2000 Japan Largest Telecommunications Routes, FY 1998199 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United States ................ 395.0 China ....................... 253.7 Rep. of Korea ................ 169.4 Philippines ................... 159.4 Taiwan ....................... 90.5 Thailand ..................... 69.9 Brazil ........................ 65.0 Hong Kong ................... 61.4 United Kingdom ............... 51.8 Singapore .................... 48.8 Australia ..................... 40.8 Indonesia .................... 38.8 Canada ...................... 28.7 Malaysia ..................... 27.8 Germany ..................... 27.2 Russia ....................... 24.5 France ....................... 22.9 Vietnam ...................... 20.1 India ......................... 13.9 Seychelles ................... 13.1 Other ....................... 197.8 Percentage of Outgoing Traffic 21.7% -- ~-.,.. -~-.... 14.0% ~~ 9.3% 8.8% 5.0% 3.8% 3.6% 3.4% 2.9% 2.6% 2.2% 2.0% 1.6% 1.5% 1.5% 1.4% 1.3% 1.1% 0.8% 0.7% 10.9% Total ...................... 1,816.2 © TeleGeography, inc. 1999 National Traffic Balance Minutes FY 1996/97 FY 1997/98 FY 1998199 ..... c,~!,~,g~ 1,519.~ ..... 1,635.~ ............ ~! ;57~5:~ .... ~U~[plus (Deficit) ~, (191.5) ................... (~;7) ............. (3~,~.,.Q) ........... Total Volume 3,229.7 3,406.7 3,470.0 Note: Data are in millions of minutes of outgoing public switched telecommumcations traffic. Route data exclude some ISR traffic =ncluded in totals. © TeleGeography, |nc. 1999 207 TeleGeography 2000 © TeleGeograph~l, Inc. 1999 Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Saudi Arabia .................. 16.2 2. Israel ........................ 12.6 3. EGypt ........................ 12.1 4. Syria ......................... 9.6 5. United Arab Emirates ........... 8.9 6. Iraq .......................... 8.3 7. United States .................. 6.6 8. Lebanon ...................... 5.3 9. Palestinian Authority ............ 4.8 10. Kuwait ........................ 4.7 11, United Kingdom ................ 4,1 12. Qatar ......................... 2.5 13. Germany ....................... 2.2 14. Canada ....................... 2.0 15. Italy .......................... 1.5 16. France ........................ 1.4 17. Oman ......................... 1.3 18. Yemen ........................ 1.2 19. Bahrain ....................... 1.1 20. Turkey ........................ 1.0 Other ........................ 15.3 Total ........................ 122.6 Percentage of Outgoing Traffic 13.2% 10.3% 9.9% ~’~ ~"~i ..... 7.8% .... ~..... ...... ..’7.3% 4.3% 3.9% 3.8% 3.3% 2.0% 1.8% 1.6% 1,2% 1.2% 1.0% 1.0% 0.9% 0.8% 12.5% © TeleGeography, Inc. National Traffic Balance Minutes 1996 1997 1998 .... ~,~.~ g 74.6 91.9 122.6 S~,~plus (Deficit) .......... 58.5 ............ 53~.~ ........................ 5~:~ Total Volume 207.7 236.9 299.5 Note." Data are in millions of minutes of outgoing public switched telecommumcations traffic. Data based on billing point of traffic. © TeleGeegrap~y; Inc. 199~ © TeleGeography, Inc. 1999 TeleGeography 2000 Kazakhstan Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Russia ....................... 71.3 2. Uzbekistan .................... 9.1 3. KyrGyzstan .................... 6.3 4. Ukraine ....................... 5.4 5. Germany ...................... 4.5 6. Turkey ........................ 2.6 7. Belarus ....................... 1.9 8. Azerbaijan ..................... 1.7 9. United States .................. 1.5 10, United KinGdom ................ 1.4 11. Tajikistan ...................... 1.4 12. Turkmenistan .................. 1.2 13. Armenia ....................... 1.0 14. China ......................... 0.8 15. GeorGia ....................... 0.6 16. Korea, Rep ..................... 0.5 17. United Arab Emirates ........... 0.5 18. rvloldova ...................... 0.3 19. Canada ....................... 0.3 20. France ........................ 0.3 Other ......................... 6.5 Percentage of Outgoing Traffic 6o.o% 7.6% 5.3% 4.5% 3.8% 2.2% 1.6% 1.4% 1,3% 1.2% 1.2% 1.0% ,~ 0.8% 0.6% 0.5% 0.4% t 0.4% 0.3% 10.3% 1 0.2% 5.5% Total ........................ 118.9 ©TeleGeography, inc: t999 ¯ National TrafficBalance Minutes 1996 Total Volume 177.5 1997 1998 n.a. 137.5 114.7 118.9 n.a. 18.6 n.a. 256.4 Note: Data are in millions of minutes of outgoing public switched telecommunications traffm. Totals for 1996 include only traffic with other members of the Commonwealth of Independent States. © TeleG~ogra phy, lnc= 1999 209 TeleGeography 2000 © TeleGeographU, Inc. 1999 Kenya Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United Kingdom ................ 7.2 United States .................. 3.9 South Africa ................... 2.2 Germany ...................... 1.1 Italy .......................... 1.0 France ........................ 0.8 United Arab Emirates ........... 0.8 Netherlands ................... 0.8 Canada ....................... 0.7 Switzerland .................... 0.6 Ethiopia ....................... 0,5 Japan ........................ 0.5 Belgium’, ...................... 0.4 Zimbabwe ..................... 0.4 Rwanda ....................... 0.3 Pakistan ...................... 0.3 Sweden ....................... 0.3 Australia ...................... 0.2 Denmark ...................... 0.2 Nigeria ........................ 0.2 Other ......................... 6.8 Percentage of Outgoing Traffic 24.6% 13.2% 7.4% 3.8% 3.4% 2.7% 2.6% 2.6% 2.3% 2.1% 1.8% 13% 1.2% 1.1% 0.9% 0.8% 0.8% 0.7% 23.3% Total ......................... 29.2 ©TeleGeography, Inc. 1999 National Traffic Balance ¯ Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data exclude calls to Tanzania and Uganda. © "l:eleGaographyi, lnc. 1999 210 © TeleC, eograph~l, Inc. 1999 TeleGeography 2000 Korea, Rep. Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. United States ................ 230.0 2. ,Japan ....................... 172.3 3. China ....................... 144.4 4. Philippines ................... 26.1 Hong Kong ................... 24.9 Canada ...................... 24.5 Australia ..................... 21.8 Indonesia .................... 19.0 United Kingdorn ............... 17.6 Germany ..................... 17.1 Vietnam ...................... 16.5 Taiwan ....................... 13.7 P, akistan ..................... 13.3 Singapore .................... 13.2 Thailand ...................... 9.8 France ........................ 9.6 Russia ........................ 8.8 India .......................... 7.9 New Zealand .................. 6.9 Malaysia ...................... 5.9 Other ....................... 104.6 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Percentage of Outgoing Traffic 25.3% ~~- ::~ 18.o% 2.9% 2.7% 2.7% 2.4% 2.1% 1.9% 1.9% 1.8% 1.5% 1.5% 1.4% 1.1% 1.1% 1.o% 0.9% 0.8% 0.7% 11.5% Total ........................ 907.7 © TeleGeography, Inc~ 1999 National TrafficBalance Minutes 1996 1997 1998 Total Volume 1,439.9 1,667.0 1,627.1 Note: Data are in millions of minutes of outgoing public switched telecommunications traffzc. ©TeleGeography, inc. 1999 211 TeleGeography 2000 © TeleGeographg, Inc. 1999 ( Kuwait Largest Telecommunications Routes, 1997 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Egypt ........................ 27.8 17.3% 2. India ......................... 17.8 ~;~ .-~ ~-i ..4 11.1% 3. Saudi Arabia .................. 16.2 ~~~..---.,~ 10.1% 4. Syria ........................ 10.9 ~~ 6.8% 5. United Arab Emirates ........... 10.1 6.3% 6. United States .................. 8.8 5.5% 7, Iran .......................... 8,8 5.5% 8. Pakistan ...................... 8.5 5.3% 9. United Kingdom ................ 8.2 5.1% 10. Jordan ........................ 5.4 3.4% 11. Lebanon ...................... 4.6 2.9% 12. Bahrain ....................... 4.3 2.7% 13. Banglad’~sh .................... 2.4 1.5% 14. Philippines .................... 1.6 1.0% 15. Canada ....................... 1.6 1.0% 16. Qatar ......................... 1.5 0.9% 17. France ........................ 1.3 0.8% 18. Germany ...................... 1.3 0.8% 19. Sri Lanka ...................... 1.2 0.8% 20. Oman ......................... 1.2 0.7% Other ........................ 17.0 ¯ .~./ ,- 10.6% Total ........................ 160.5 ©TeleGeography, Inc. 1999 National Traffic Balance ¯ Minutes 1996 1997 1998 ..... !~o,~!p~g 131.2 n.a. ¯ ............ ~,,~., ....... Surplus (Deficit) ........... (9.5) n.a. ¯ ...................... ~:~: ........ Total Volume 271.9 n.a. n.a. Note: Data are in millions of m~nutes of outgoing pubhc switched telecommunications traffic. ¯ ©Tel~eGeography~ Inc. 1999 212 © TeleGeographyo Inc. 1999 TeleGeography 2000 Largest Telecommunications Routes, 1998 Latvia Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Russia ....................... 15.4 Lithuania ...................... 6.0 Estonia ........................ 5.1 Ukraine ....................... 4.1 6errnany ...................... 3.8 Sweden ....................... 2.8 Belarus ....................... 2.7 United Kingdom ................ 2.0 Finland ". ....................... 1.6 Denmark ...................... 1.4 Poland ........................ 1.2 United States .................. 0.9 Netherlands ................... 0.8 Norway ....................... 0.8 France ........................ 0.6 Italy .......................... 0.6 Belgium ....................... 0.5 Switzerland .................... 0.5 Austria ........................ 0.4 Kazakhstan .................... 0.4 Other ......................... 4.1 Percentage of Outgoing Traffic 27.7% 10.9% 9.2% 6.8% 5.0% k~ 4.0% 3,6% 2.B% 2.5% 2.1% 1.7% 1.4% 1.4% 1.1% 1.0% 0.9% 0.9% 0.7% 0.6% 7.4% Total ......................... 55.4 © TeleGeography, inc. 1999 National Traffic Balance Minutes 1996 1997 1998 ..... 0~.t.g o..~!q,g ............................................... 43.0 44.0 .~.u..~plus (Deficit) 26.:q ........................ ..3.,7~:..8,. ................. 3.~:...8~ Total Volume 112.0 125.8 142.5 Note: Data are in millions of minutes of outgoing public switched telecommumcat=ons traffic. Data based on billing point of traffic. ¯ © TeleEieography, inc. t999 213 TeleGeography 2000 © TeleGeographg, Inc, 1999 Luxembourg Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Minutes (millions) Germany ..................... 69.0 France ....................... 65.0 Belgium ...................... 64.0 Portugal ...................... 18.0 United Kingdom ............... 16.0 Italy ......................... 14.0 Netherlands .................. 10.0 Switzerland .................... 8.0 United States .................. 7.0 Spain ......................... 5.0 Austria ........................ 3,0 Denmark ...................... 3.0 Sweden’, ....................... 3.0 Greece ........................ 2.0 Poland ........................ 1.3 Other ......................... 5.5 Percentage of Outgoing Traffic 23.5% 22.1% 21.8% ~6.1% 5.4% 4.8% 3.4% 2.7% 2.4% 1.7% 1.0% 1.o% 0.7% 0.4% 1.9% Total ........................ 293.8 © TeleGeography, Inc. 1999 National Traffic Balance .......... Minutes 1996 1997 1998 i,.,n,,,,€...,o....,.m,...i,,n,, g 189,8 ............... .2.0, 8,~. .............. 242,6 ,,_ ...,O.,.,,u,,,.tg,.,o,,},n_g~ .................................................... .24,,8,.,:.,5, ................ 282.9 293.8 ,,,~,yrplus (DeficJ.t..) ....................................... .(....5.8..-.7) ................. (74.9) (51.2) Total Volume 438.3 490.9 536.4 Note: Data are =n millions of mznutes of outgoing public switched telecommunications traffic. .. © TeleGeography;, Inc. t999 214 © TeleGeography, Inc. 1999 TeleGeography 2000 Macau Largest Telecommunications Routes, 1998 Destination Percentage of Outgoing Traffic 1. 42.7% 2. 40.9% 3. 3.3% 4. 3.0% 2.0% 1.4% 1.2% 0.9% 0.9% 0.6% 0.5% 0.3% 0.3% 0.3% 0.2% 0.2% ¯ 0.1% o.1% o.1% o.1% 0.9% 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Hung Kong ................... 53.4 China ........................ 51.3 Portugal ....................... 4.1 Taiwan ........................ 3.8 United States .................. 2.5 Canada ....................... 1.8 Philippines .................... 1.5 Thailand ...................... 1.2 Australia ...................... 1.2 United Kingdom ................ 0.8 Singapore ..................... 0.6 Japan ........................ 0.4 Malaysia ...................... 0.3 Korea, Rep ..................... 0.3 France ........................ 0.3 Indonesia ..................... 0.2 New Zealand .................. 0.2 Germany ...................... 0.1 Brazil ......................... 0.1 Vietnam ....................... 0.1 Other ......................... 1.1 Total ........................ 125.2 © TeleGeography, Inc. 1999 National Traffic Balance Minutes 1996 1997 1998 ......... O,u,,t_g_qJ n g ..................................................... I_.I_2:.5.. ............... .!, !..,9.0 125,2 Total Volume 204.6 211.2 221.0 Note: Data are in millions of minutes of outgoing public switched telecommunications traff=c. Data based on billing point of traffic. © TeleGeography, tnc. 1999 215 TeleGeography 2000 © TeleGeogrophy, Inc. 1999 Macedonia Largest,Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Germany ...................... 5.5 14.9% 2, Bulgaria ....................... 3.2 8.5% 3. Switzerland .................... 3.1 8.3% 4. Greece ........................ 3.1 8.3% Italy .......................... 2.2 5.8% Slovenia ...................... 2.1 5.8% Croatia ........................ 2.1 5.8% United States .................. 1.6 ~~ 4.4% Austria ........................ 1.5 4.1% United Kingdom ................ 1.0 2.8% Australia ...................... 1.0 2.6% Albania ....................... 0.8 2.1% Bosnia-H,erzegovina ............ 0.7 2.0% France ........................ 0.7 1.8% Sweden ....................... 0.6 1.5% Netherlands ................... 0.6 1.5% Canada ....................... 0.5 1.2% Russia ........................ 0.4 1.2% Belgium ....................... 0.4 1.2% Czech Republic ................. 0.4 1.1% Other ......................... 5.6 15.0% 5o 6. 7. 8. 9. 10. 11, 12, 13. 14, 15, 16, 17. 18. 19, 20. Total ......................... 37.1 © TeleGeography, inc. 1999 National Traffic Balance Note: Data are in millions of minutes of outgoing public switched telecommumcat~ons traffic. Data for 1998 exclude an estimated 20 million minutes of traffic to Yugoslavia. ©TeleGeography~lnc. 1999 216 © TeleGeography, Inc. 1999 TeleGeography 2000 Malaysia Largest Telecommunications Routes, FY 1998/99 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Singapore ................... 350.0 51.1% 2. Indonesia .................... 46.0 6.7% 3. ,Japan ....................... 29.0 4.2% 4. Australia ..................... 25.0 3.6% 5. United States ................. 25.0 3.6% 6. United Kingdom ............... 24.0 3.5% 7. Thailand ..................... 16.0 2.3% 8. Hong Kong ................... 15.0 2.2% 9. India ........................ 14.0 2.o% 10. Philippines ................... 12.0 1.8% 11. Taiwan ...................... 11.0 1.6% 12. China ........................ 7.0 1.0% 13. Germany ...................... 6.0 0.9% 14. Bangladesh ................... 6.0 0.9% 15. Korea, Rep .................... 4.0 0.6% Other ........................ 95.0 13.9% Total ....................... 685.0 © TeleBeography, inc. 1999 ¯ National Traffic Balance Minutes ..... ,S,_u,[plus (Deficit) Total Volume 581.9 592.0 n.a. 570.5 588.5 685.0 11.4 3.5 n.a. 1,152.4 1,180.5 n.a. Note: Data are in milhons of minutes of outgoing public switched telecommumcations traffic. Data based on billing point of traffic. Fiscal year ends 31 March. Totals for FY 1996/97 and 1997/98 were for Telkom Malaysia only. ©TeleGeography,|nc. 1999 2.17 TeleGeography 2000 © TeleGeograph~l, Inc. 1999 f Malta Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United Kingdom ............... 12.0 Italy .......................... 5.6 Germany ...................... 3.4 Libya ......................... 1.6 France ........................ 1.5 United States .................. 1.3 Netherlands ................... 1.1 Australia ...................... 0.9 Russia ........................ 0.8 Switzerland .................... 0.7 Belgium ....................... 0.6 Austria ........................ 0.5 Canada ’, ...................... 0.5 Greece ........................ 0.4 Spain ......................... 0.4 Turkey ........................ 0.4 Sweden ....................... 0.4 Tunisia ........................ 0.4 Ireland ........................ 0.3 Egypt ......................... 0.3 Other ......................... 4.3 Percentage of Outgoing Traffic 32.1% 14.9% 9.1% 4.4% ~4.1% 3.4% 3.0% 2.5% 2.2% 2.0% 1.5% 1.3% 1.3% 1.1% 1,0% 1.0% 1.0% 0.9% 0.8% 0.8% 11.5% Total ......................... 37.3 © Telet3eog~aphy, inc.,1999 National Traffic Balance,, Minutes 1996 1997 1998 ¯ ,S,u,[P,!,US (g efi c it) ......................... 2.3 .............. 2~§ .............................. ~:~¯.~ ............ Total Volume 65.7 71.4 n.a. Note: Data are In millions of minutes of outgoing pubhc switched telecommunications traffic. ¯ © TeleGeography~ Inc. 1999 218 © TeleGeographg, Inc. 1999 TeleGeography 2000 Mauritius Largest Telecommunications Routes, FY 1998199 Destination Minutes (millions) Percentage of Outgoing Traffic 1. France ........................ 5.9 19.8% 2. United Kingdom ................ 4.7 -~~ ~,~ .,,,, .. 15.8% 3. R~union ....................... 3.0 10.1% 4. South Africa ................... 2.8 ~~ 9.3% 5. India .......................... 2.0 ~~ 6.6% 6. Italy .......................... 1.5 5.1% 7. Germany ...................... 0.9 3.1% 8. Australia ...................... 0.9 3.0% 9. Madagascar ................... 0.8 2.8% 10. Switzerland .................... 0.6 1.9% 11. Singapore ..................... 0.6 1.9% 12. United States .................. 0.5 1.8% 13. qhina ......................... 0.5 1.7% 14. Hong Kong .................... 0.5 1.7% 15. Belgium ....................... 0.4 1.4%o 16. Seychelles .................... 0.4 1.3% 17. Canada ....................... 0.3 1.1% 18. Taiwan ........................ 0.2 o.8% 19. Malaysia ...................... 0.2 0.7% 20. Kenya ........................ 0.2 0.6% Other ......................... 2.8 9.5% Total ......................... 29.7 © TeleGeography, Inc. 1989 National Traffic Balance Incoming 29.4 35.0 Surplus (Deficit) 7.7 10.4 Total Volume 51.1 59.6 icY 1998/~.......... 38.5 29.7 8.8 68.2 Note: Data are in millions of m~nutes of outgoing public switched telecommumcatmns traffic. Fiscal year ends 30 June. © TeleGeography, tnc. 1999 219 TeleGeography 2000 © TeleGeograpt~U, Inc. 1999 Mexico Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United States ............... 1,095.0 Canada ...................... 20.0 Spain ........................ 12.0 Italy .......................... 8.0 Germany ...................... 7.5 France ........................ 7.0 Guatemala ..................... 6.0 Argentina ..................... 6.0 Brazil ......................... 6.0 Colombia ...................... 6.0 Cuba ......................... 6.0 Chile .......................... 5.5 United Kingdom ................ 4.5 Costa Rica ..................... 4.0 El Salvador .................... 3.5 Peru .......................... 3.0 Honduras ..................... 3.0 Japan ........................ 3.0 Venezuela ..................... 3.0 Switzerland .................... 2.0 Other ........................ 97.5 Percentage of Outgoing Traffic 83.6% 1.5% 0.9% 0.6% I 0.6% 0.5% i 0.5% ! 0.5% 0.5% ! 0.5% 0.5% f 0.4% I 0.3% l 0.3% t 0.2% 1 0.2% ! 0.2% 1 0.2% ! 0.2% 1 0.2% 7.4% Total ...................... 1,310.0 © TeleSeography, Inc. 1999 National Traffic Balance Minutes ...... !,~.,~mJ~g~ Su_~(~!us (Deficit) Total Volume Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. 220 © TeleGeograp.h~y, In c. 1999 © TeleGeography, Inc. 1999 TeleGeoi~raphy 2000 Moldova Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Russia ....................... 22.5 40.2% 2. Ukraine ...................... 15.8 ~~~ 28.2% 3. Romania ...................... 6.1 10.9% 4. Belarus ....................... 1.8 3.3% 5. Germany ...................... 1.6 2.8% 6. Turkey ........................ 0.8 1.4% 7. Greece ........................ 0.7 1.3% 8. United States .................. 0.7 1.2% 9. Italy .......................... 0.7 1.2% 10. Israel ......................... 0.6 1.2% 11. BulGaria ....................... 0.6 1.1% 12. France ........................ 0.4 0.7% 13. Roland ........................ 0.4 0.6% 14. Hungary ....................... 0.3 0.5% 15. United Kingdom ................ 0.2 0.4% 16. Czech Republic ................. 0.2 1 0.4% 17. Kazakhstan .................... 0.2 0.3% 18. Canada ....................... 0.2 10.3% 19. Netherlands ................... 0.1 0.3% 20. Switzerland .................... 0.1 10.2% Other ......................... 2.0 3.5% Total ......................... 55.8 © TeleGeography, Inc. 1989 National Traffic Balance Minutes 1996 1997 1998 Total Volume 122.2 135.8 146.1 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on bdhng point of traffic. Totals for 1996 include only traffic wzth ot~er members of the Commonwealth of Independent States. © TeleGeographyo’lnc. 1999 221 TeleGeoi~raphy 2000 © TeleGeography, Inc. 1999 Monaco Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. France ...................... 157.0 67.1% 2. Italy ......................... 26.0 11.1% 3. United Kingdom ............... 12.0 5.1% 4. Switzerland .................... 7.0 3.0% 5. 6ermany ...................... 5.0 2.1% 6. United States .................. 5.0 2.1% 7. Belgium ....................... 2.0 0.9% 8. Netherlands ................... 2.0 0.9% Other ........................ 16.0 7.7% Total ........................ 234.0 ©TeleBeo#raphy, Inc: 1999 National Traffic Balance ..... Minutes 1996 1997 1998 .... S,,u[,plus (Deficit) n.a. n.a. (94.0) Total Volume n.a. n.a. 374.0 Note: Data are in millions of minutes of outgoing pubhc switched telecommunications traffic. Outgoing traffic includes an estimated 50 million minutes of calls reoriginated from other countries. © TeleGeography, Inc. 1999 222 © Tele~eogra~hy, I~. 1999 TeleGeography 2000 Morocco Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. France ....................... 70.0 2. Spain ........................ 15.0 3. italy .......................... 9.0 4. United Kingdom ................ 7.5 5. Belgium ....................... 7.0 6. Germany ...................... 6.5 7. United States .................. 6.5 8. Netherlands ................... 6.0 9. Saudi Arabia ................... 5.5 10. Canada ....................... 3.0 11. Switzerland .................... 3.0 12. Tunisia ........................ 2.5 13. A, Igeria ........................ 2.0 14. Egypt ......................... 1.5 15. United Arab Emirates ........... 1.5 Other ........................ 11.5 Total ........................ 158.0 Percentage of Outgoing Traffic 44.3% 9.5% 5.7% 4.7% 4.4% 4.1% 3.8% 3.5% 1.9% 1.9% 1.6% 1.3% 0.9% 0.9% 7.3% © TeleSeography, Inc. 1999 National Traffic,Balance Minutes Surplus (Deficit) Total Volume 1996 1997 1998 286.6 364.0 460.0 129.3 149.9 158.0 157.3 214.1 302.0 415.9 513.9 618.0 Note: Data are in millions of minutes of outgoing public switched telecommumcations traffic. ©TeleGeography, Inc. 1999 ¯ ......... TeleGeography 2000 © TeleGeogrophy, Inc. 1999 Namibia Largest Telecommunications Routes, 1998 Destination Percentage of Outgoing Traffic 1. 83.1% 2. 3.5% 3. 1.8% 4. 1.2% 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. Minutes (millions) South Africa .................. 51.4 Germany ...................... 2.1 United Kingdom ................ 1.1 United States .................. 0.8 Zimbabwe ..................... 0.6 Botswana ..................... 0.6 Angola ........................ 0.5 Spain ......................... 0.4 Zambia ........................ 0.4 France ........................ 0.3 Netherlands ................... 0.2 Portugal ....................... 0.2 Italy .. : ........................ 0.2 Switzerland .................... 0.2 China ......................... 0.1 Australia ...................... 0.1 Austria ........................ 0.1 Russia ........................ 0.1 Belgium ....................... 0.1 Other ......................... 2.3 1.0% 1.o% 0.7% o.6% o.6% 0.5% 0.3% o.3% 0.3% :0.3% 0.2% 0.2% 0.2% 0.2% 0.2% 3.8% Total ......................... 61.9 ¯ ©TeleGeography, Inc. 1999" ¯ National Traffic Balance Minutes 1996 Incom!~_g 41.7 ~__...S..~rplus (Deficit) Total Volume 93.1 1997 1998 42.3 45.3 49.7 61.9 92.0 107.2 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. ©TeleGeography, Inc. 1999 224 © TeleGeography, Inc. 1999 TeleGeography 2000 Netherlands Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Minutes (millions) Germany .................... 395.0 Belgium ..................... 300.0 United Kingdom .............. 225.0 France ...................... 145.0 United States ................ 120.0 italy ......................... 66.0 Spain ........................ 52.0 Switzerland ................... 44.0 Canada ...................... 35.0 Turkey ....................... 34.0 Sweden ...................... 28.0 Denmark ..................... 24.0 Austria ....................... 22.0 Poland ....................... 21.0 Morocco ..................... 15.0 Other ....................... 279.0 Percentage of Outgoing Traffic 21.9% ~- .:. 16.6% ~:./" ....... "’~ ... 12.5% 8.0% 6.6% 3.7% 2.9% 2.4% 1.9% 1.9% 1.6% 1.3% 1.2% 1.2% 0.8% Total ...................... 1,805.0 © TeleBeography, Inc: t989 National Traffic Balance Note: Data are m millions of minutes of outgoing public switched telecommunications traffic. ©TeleG~ eography, Inc. 1999 225 TeleGeography 2000 © TeleGeograph~l, Inc. 1999 ( Netherlands Antilles ¯ Largest Telecommunication Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Netherlands .................. 12.2 28.4% 2. United States ................. 11.2 26.1% 3. Aruba ......................... 4.7 10.9% 4. Dominican Republic ............. 3.0 7.0% 5. Venezuela ..................... 2.7 6.2% 6. Colombia ...................... 1.4 3.4% 7. Jamaica ...................... 0.7 1.7% 8. Canada ....................... 0.6 1.4% 9. Suriname ...................... 0.5 1.3% 10. Saint Kitts & Nevis .............. 0.5 1.1% 11. Trinidad & Tobago .............. 0.4 1.0% 12. United Kingdom ................ 0.3 0.8% 13. Dominicat ...................... 0.3 0,7% 14. Guyana ....................... 0.3 0.6% 15. Anguilla ....................... 0.3 O.6% 16. Haiti .......................... 0.2 0.5% 17, Barbados ..................... 0.2 0.5% 18, France ........................ 0,2 O.4% Other ......................... 3.2 7.5% Total ......................... 43.0 © TeleGeography, Inc, 1999 National Traffic ,Balance Minutes 1996 1997 1998 Inco~!~,9 ........... n.a. n.a. 113.8 ........ ~.g~.o~ !~g~ 40.1 n.a. 43.0 Total Volume n.a. n.a. 156.8 Note: Data are in milhons of minutes of outgoing public switched telecommunications traffic. -U.S. figures include traffic to the U.S. V~rgin Islands. ©TeleGeograp~lnc. 1999 226 © TeleGeography, Inc. 1999 TeleGeography 2000 New Zealand Largest Telecommunications Routes, FY 1998199 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Australia .................... 260.0 44.8% 2. United Kingdom ............... 70.0 12.1% 3. United States ................. 65.0 11.2% 4. Canada ...................... 30.0 5.2% 5. Philippines ................... 15.0 2.6% 6. Hong Kong ................... 10.0 E 1.7% 7. ,Japan ........................ 10.0 1.7% 8. Fiji ........................... 9.0 1.6% 9. Maiaysia ...................... 7.0 1.2% 10. SinGapore ..................... 7.0 1.2% 11. Germany ...................... 5.0 0.9% 12. Korea, Rep ..................... 5.0 0.9% 13. ,~outh Africa ................... 5.0 0.9% 14. Taiwan ........................ 5.0 0.9% 15. Western Samoa ................ 5.0 0.9% Other ........................ 72.0 ~~ 12.4% Total ........................ 580.0 © TeleGeography, Inc. 1999 National Traffic Balance Minutes FY 1996/97 FY 1997/98 FY 1998/99 ...... !,n~,€,,,£~g_ ................................................ 380.0 430.0 ........ .S,U.,,r,,p,,I,,,U,,,~, (D, ,e,,~!,€,.!,t,,),,, .............................. 27.0 Total Volume 733.0 837.0 n.a. Note: Data are in millions of minutes of outgoing public switched telecommumcat~ons traffic. Fiscal year ends 31 March. © TeleGeography, tnc. t999 TeleGeo~raphy 2000 © TeleGeography, Inc. 1999 Nicaragua Largest Telecommunications Routes, 1997 Destination Minutes (millions) Percentage of Outgoing Traffic 1. United States ................. 16.1 39.7% 2. Costa Rica .................... 9.2 22.8% 3, Guatemala .................... 3.3 8.2% 4. El Salvador .................... 3.0 7.5% 5. Honduras ..................... 2.8 7.0% 6. Mexico ....................... 1.0 2.6% 7. Panama ...................... 0.9 2.3% 8. Canada ....................... 0.7 1.7% 9. Spain ........................ 0.5 1.3% 10. Colombia ..................... 0.3 0.6% 11. Brazil ........................ 0.2 0.5% 12. Cuba ......................... 0.2 0.5% 13. Germany ....................... 0.2 0.4% 14. Argentina ..................... 0.2 0.4% 15. Italy .......................... 0.2 0.4% 16. Chile ......................... 0.1 0.3% 17. Netherlands ................... 0.1 1 0.3% 18. Peru ......................... 0.1 0.3% 19. Venezuela .................... 0.1 1 0.2% 20. France ....................... 0.1 I O.2% Other ......................... 1.1 2.8% Total ......................... 40.4 © l~eleGeography, Inc, 1999 National Traffic Balance .... Minutes 1996 1997 1998 In c0,,,,m.,!~ ........... 47.1 5~.~ ....................... _n,~.~,; .... ~.O,.,,u,~g£!,,n,& .............. 32.1 40.4 n.a. S,,~,[plu s (Deficit) 15.0 12.1 n.a. Total Volume 79.2 92.9 n.a. Note: Data are in mdhons of minutes of outgoing public swztched telecommunications traffic. Data based on billing point of traffic. ¯ ©TeleGeography, Inc. 1999 228 © TeleGeographyo Inc. 1999 TeleGeography 2000 Norway Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Minutes (millions) Sweden ..................... 140.0 United Kingdom ............... 70.0 Denmark ..................... 65.0 United States ................. 45.0 Germany ..................... 35.0 France ....................... 21.0 Netherlands .................. 17.0 Finland ....................... 15.0 Spain ........................ 11.0 Italy ......................... 10.0 Other ....................... 111.0 Percentage of Outgoing Traffic 25.9% 13.0% 12.0% 8.3% 6.5% 3.9% ~@ 2.8% 1.9% 20.6% Total ........................ 540.0 ¯ ¯ © TeleGeography, Inc. 1999 National Traffic Balance Minutes 1996 ...... ],n,,~?~i_~,g ......... 422.3 ~urplus (D ef~.€it_) ................................ (~1.2) Total Volume 865.8 1997 1998 515.0 n.a. 481.0 540.0 34.0 n.a. 996.0 n.a. Note: Data are in millions of minutes of outgoing public switched te[ecommunmat~ons traff=c. Data based on billing point of ~’affic. © Tel~G~ography, |nc. 1899 229 TeleGeography 2000 © TeleGeography, Inc. 1999 Oman Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. India ......................... 25.2 2. United Arab Emirates ........... 25.0 3. Pakistan ...................... 5.5 4. United Kingdom ................ 5.5 Egypt ......................... 3.7 Saudi Arabia ................... 2.7 Bahrain ....................... 2.2 United States .................. 2.1 Jordan ........................ 1.6 Kuwait ........................ 1.3 Qatar ......................... 1.3 Sri Lanka ...................... 1.3 Philippines .................... 1.3 Canada ....................... 1.1 Tanzania ...................... 1.0 Sudan ........................ 0.8 France ........................ 0.7 Germany ...................... 0.7 Netherlands ................... 0.6 Italy .......................... 0.5 Other ......................... 9.0 6. 7, 8. 9, 11, 12, 13, 14. 15. 16. 18. 19. 20. Percentage of Outgoing Traffic 27.1% 26.9% 5.9% 5.9% 3.9% 2.9% 2,4% 2.2% 1.7% 1.4% 1.4% 1.4% 1.3% 1.1% 1.1% 0.8% 0.8% 0.8% 0.6% 0.5% 9.7% Total ......................... 92.9 "© Telel]eography, lnc, 1999 National Traffic Balance Minutes 1996 1997 1998 Surplus (Deficit) (4.6) (3.9) (18.7) Total Volume 120.6 144.7 167.1 Note= Data are in mflhons of m~nutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Excludes some cross-border calls to the United Arab Emzrates. "" ’ ¯ ©TeleSeography~lnc. 230 © TeleGeography, Inc. 1999 TeleGeography 2000 Pakistan Largest Telecommunications Routes, FY 1998199 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United Kingdom ............... 16.8 United States ................. 12.7 United Arab Emirates ........... 11.6 Saudi Arabia ................... 7.8 Canada ....................... 6.8 Germany ...................... 2.9 Italy .......................... 2.6 Iran .......................... 2.3 India .......................... 2.1 Japan ........................ 2.0 Singapore ..................... 2.0 Kuwait ........................ 1.9 F~ance ........................ 1.8 Hong Kong .................... 1.3 China ......................... 1.3 Netherlands ................... 1.1 Turkey ........................ 1.0 Bangladesh .................... 1.0 Oman ......................... 1.0 Australia ...................... 0.8 Other ......................... 6,9 Percentage of Outgoing Traffic o ¯ ,~\~19.2~/oo ".~:~:-. :. 14.5% 9.0% 3.3% 3.0% 2.6% 2.3% 2.3% 2.2% 2.0% 1.5% 1.5% 1.3% 1.2% 1.1% 1.1% 1.0% 7.9% Total ......................... 87.5 © Tel’~Geography; inca 1999 National Traffic Balance Minutes FY 1996/97 FY 1997/98 FY 1998199 ..... !...n.....c.,~O m! n g ............................................... ~.,8.~:..4. ..................... 55...7.~ ............ ,.,,,64...~:~ ..... .... ..O....,uo~t~g.ooi_q.g ............................................. ,,7..7..0 ........................... .8..4.:I 87.5 ....... Surplus (Deficit) 411.4 .~3..~ ......... 552.9 Total Volume 565.4 641.9 727.9 Note; Data are zn milhons of minutes of outgoing public switched telecommunications traffic. Fzscal year ends 30 June. Data exclude some cross-border traffm to Indm. © TeleGeography, tnc. 1999 231 TeleGeography 2000 © TeleGeography, Inc. 1999 ( Palestinian Authority Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Jordan ....................... 11.8 2. United States .................. 2.9 3. Egypt ......................... 1.9 4. Saudi Arabia ................... 1.5 5. United Arab Emirates ........... 1.1 6. Germany ...................... 0.7 7. United Kingdom ................ 0,6 8. Italy .......................... 0.5 9. Canada ....................... 0.4 10. Ukraine ....................... 0.3 11. Syria ......................... 0.3 12. Turkey ........................ 0.3 13. France .’ ....................... 0.3 14. Lebanon ...................... 0.3 15. Russia ........................ 0.2 16. Qatar ......................... 0.2 17. Morocco ...................... 0.2 18. Norway ....................... 0.2 19. Tunisia ........................ 0.2 20. Kuwait ........................ 0.2 Other ......................... 3.6 Total ......................... 27.6 Percentage of Outgoing Traffic 42.7% ~~ 10.6% 6.8% 5.3% 4.o% 2.5% 2.0% 1.9% 1.4% 1.3% 1.2% 1.1% 1.0% 0.9% 0.8% 0.7% 0.7% 0.7% 0.6% 0.6% 13.1% © TeleGe~ogr~phy, inc. t999 National Traffic Balance ¯ ¯ ............ Minutes Surplus (Deficit) Total Volume 1996 1997 1998 n.a. n.a. 16.6 n.a. n.a. 27.6 n.a. n.a. (11.0) n.a. n.a. 44.3 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data exclude traffic with Israel. ©~eleGeographyilnn. 1999 © TeleGeograph~l, Inc. 1999 TeleGeography 2000 Panama Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United States ................. 20.1 Colombia ...................... 5.4 Costa Rica ..................... 3.8 Mexico ....................... 2.3 Venezuela ..................... 1.3 Guatemala ..................... 1.2 Ecuador ....................... 1.0 Dominican Republic ............. 1.0 El Salvador .................... 0.9 Spain ......................... 0.9 Nicaragua ..................... 0.9 Canada ....................... 0.9 Brazil ......................... 0.8 Peru .......................... 0.7 Honduras ..................... 0.7 Cuba ......................... 0.6 United Kingdom ................ 0.6 Argentina ..................... 0.6 Chile .......................... 0.5 China ......................... 0.3 Other ......................... 5.3 Percentage of Outgoing Traffic 40.3% 10.8% 4.5% 2.7% 2.5% 2.1% 2.0% 1.9% 1.9% 1.8% 1.8% 1.6% 1.4% 1.4% 1.2% 1.2% 1.1% 1.0% 0.7% ~~ lO.7% Total ......................... 50.0 © TeleGeography, Inc, 1999 National Traffic Balance Minutes 1996 1997 1998 .... ~,u,tg,o i_~ g, 41.2 41.4 50.0 Total Volume 138.9 136.5 147.6 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. © TeleGeography, |nc. t~999 233 TeleGeography 2000 © TeleGeographu, Inc. 1999 Paraguay Largest Telecommunications, Routes, 1998 Destination Minutes (millions) 1. Argentina .................... 12.0 2. Brazil ......................... 8.6 3. United States .................. 3.8 4. Uruguay ....................... 1.5 5. Chile .......................... 1.3 6. Germany ...................... 0.6 7. Bolivia ........................ 0.5 8. Spain ......................... 0.5 9. Lebanon ...................... 0.4 10. Taiwan ........................ 0.4 11. Korea, Rep ..................... 0.4 12. Peru .......................... 0.4 13. France .’ ....................... 0.3 14. China ......................... 0.3 15. Italy .......................... 0.3 16. Mexico ....................... 0.3 17. Japan ........................ 0.2 18. Canada ....................... 0.2 19. Switzerland .................... 0.2 20. United Kingdom ................ 0.2 Other ......................... 5.5 Percentage of Outgoing Traffic 31.8% 22.8% lO.1% 3.9% 3.4% 1.5% 1.4% 1.3% 1.1% 1.1% 1.0% 1.0% 0.8% 0.7% 0.7% 0.7% 0.6% 0.6% 0.5% 0.4% 14.5% Total ......................... 37.8 .... © TeleGeography~ inc. 1999 National Traffic Balance Note: Data are in millions of minutes of outgoing public switched telecommunications traffzc. ©TeleGe~graphy~ Inc. 1999 234 © TeleGeography, Inc. 1999 TeleGeography 2000 Peru Largest Telecommunications Routes, F’Y 1998/99 Destination Minutes (millions) Percentage of Outgoing Traffic 1. United States ................. 30.3 33.6% 2. Chile .......................... 7.0 7.7% 3. Argentina ..................... 6.9 7.6% 4. Spain ......................... 5.5 6.1% 5. Colombia ...................... 3.8 4.2% 6. Italy .......................... 3.3 3.7% 7. Venezuela ..................... 2.9 3.2% 8. Bolivia ........................ 2.8 3.1% 9. Mexico ....................... 2.8 3.0% 10. Brazil ......................... 2.7 3.O% 11. Canada ....................... 2.5 2.8% 12. Japan ........................ 2.5 2.7% 13. Ecuador ....................... 2.0 2.2% 14. Germany ...................... 1.6 1.8% 15. United Kingdom ................ 1.3 1.4% 16. France ........................ 1.1 1.2% 17. Switzerland .................... 0.8 0.9% 18. Panama ....................... 0.7 0.8% 19. Costa Rica ..................... 0.6 0.6% 20. Korea, Rep ..................... 0.4 0.5% Other ......................... 8.8 9.8% Total ......................... 90.4 © TeleGeography, Inc. t 999 National Yraffic Balance : Minutes FY 1996/97 FY 1997/98 FY 1998/99 ...... !,n~o~!~9 ........................................ 22~:~5~ ........................ ~.~ ......................... ~7,~.~ .......... Total Volume 293.2 336.3 363.0 Note: Data are in millions of minutes of outgoing pubhc switched telecommunications traffic. Data based on billing point of traffic. Fiscal year ends 31 March. © TeleG~eogra phy, Inc. 1999 235 TeleOeography 2000 © TeleGeography, Inc. 1999 Philippines Largest Telecommunications Routes, F’Y 1998199 Destination Minutes (millions) Percentage of Outgoing Traffic 1. United States ................. 85.0 28.8% 2. Japan ....................... 45.0 15.3% 3. Hong Kong ................... 38.0 12,9% 4. Canada ...................... 20.0 6.8% 5. "]’aiwan ...................... 18.0 6.1% 6. Singapore .................... 15.0 5.1% 7. Australia ..................... 13.0 4.4% 8. Korea, Rep .................... 7.0 2.4% 9. Malaysia ..................... 7.0 2.4% 10. Saudi Arabia .................. 7.0 2.4% Other ........................ 40.0 13.6% Total .. ;,. .................... 295.0 © Tela6eegraphy, lnc, 1999 National Traffic, Balance ...... Minutes FY 1996/97 FY 1997/98 FY 1998/99 ..... !on_F~g ~oo 767:~,0~ .................... 9~0.:Q ........... n.a. ...... 0~,t,~p,,,i~o. 240.0 295.0 295.0 .... o,S,~[~!u s (Deficit) 527.0 635.0 n.a. Total Volume 1,007.0 1,225.0 n.a. Note= Data are in millions of mznutes of outgozng public swztched telecommunications traffic. Data based on billing point of traffzc. Fiscal year ends 31 March. ©TeleGeography, Inc~ 1999 236 © TeleGeographg, Inc. 1999 TeleGeo{~raphy 2000 Poland Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic ................... ~- o 1. Germany .233.8 38.8% 2. United Kingdom ............... 35.9 6.0% 3. Italy ......................... 35.6 5.9% 4. France ....................... 33.0 5.5% 5. United States ................. 26.3 4.4% 6. Austria ....................... 20.3 3.4% 7. Netherlands .................. 20.1 3.3% 8. Ukraine ...................... 18.8 3.1% 9. Sweden ...................... 16.8 2.8% 10. Czech Republic ................ 14.9 2.5% 11. Russia ....................... 14.5 2.4% 12. Belgium ...................... 13.5 2.2% 13. V, ietna m ...................... 12.1 2.0% 14. Denmark ..................... 10.3 1.7% 15. Switzerland .................... 8.2 1.4% 16. Belarus ....................... 7.5 1.3% 17. Spain ......................... 7.3 1.2% 18. Canada ....................... 7.0 1.2% 19. Hungary ....................... 5.8 1.0% 20. Norway ....................... 5.3 0.9% Other ........................ 55.5 9.2% Total ........................ 602,4 ©Te[eBeogr~phy, inc. 1999 National Traffic Balance Minutes 1996 1997 1998 Total Volume 1,162.7 1,329.2 1,746.6 Note: Data are in milhons of minutes of outgoing public swztched telecommumcations traffic. Data based on billing point of traffic. 237 TeleGeography 2000 © TeleGeography, Inc. 1999 f Portugal Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7, 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) France ....................... 90.1 Spain ........................ 81.2 United Kingdom ............... 49.2 Germany ..................... 46.9 Switzerland ................... 21.9 Brazil ........................ 21.3 United States ................. 18.2 Italy ......................... 16.2 Netherlands .................. 16.1 Belgium ...................... 12.7 Angola ........................ 9.6 Canada ....................... 7.3 Cape Vet, de .................... 5.5 Luxembourg ................... 4.7 Sweden ....................... 4.6 Mozambique ................... 3.4 South Africa ................... 3.1 Denmark ...................... 2.9 Venezuela ..................... 2.7 Guinea-Bissau ................. 2.7 Other ........................ 40.0 Percentage of Outgoing Traffic 19.6% 17.6% lo.7 /o 10.2% 4.8% 4.6% 4.0% 3.5% 3.5% 2.8% ~2.1% 1.6% 1.2% 1.0% 1.0% 0.7% 0.7% 0.6% 0.6% o.6% 8.7% Total ........................ 460.2 ©TeleGeography, Inc, 1999 National Traffic Balance ..... : Minutes 1996 1997 1998 I n c o mi~g 571.4 628.8 711.5 _. ~.U.m!.U~o.~(.De_~! ~[t) 23!:4 235.5 251.3 Total Volume 911.4 1,022.1 1,171.7 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Data may exclude some cross-border traffic to Spain. 238 ©Te!eGeography, Inc. 1999 © TeleGeography, Inc. ! 999 TeleGeography 2000 Qatar Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. India ......................... 22.0 19.6% 2, United Arab Ernirates ........... 17.9 ~J~l~~ 15.9% 3. Egypt ......................... 8.7 7.7% 4. Saudi Arabia ................... 7.5 6.7% 5. United Kingdom ................ 6.6 5.8% 6. Pakistan ...................... 6.1 5.5% 7. Bahrain ....................... 6.1 ~~ 5.4% 8. United States .................. 3.3 3.0% 9. Jordan ........................ 3.0 2.7% 10. Kuwait ........................ 2.4 2.1% 11. Iran .......................... 1.7 1.5% 12. France ........................ 1.5 1.3% 13. Qman ......................... 1.3 1.2% 14. Syria ......................... 1.0 0.9% 15. Canada ....................... 0.8 0.7% 16. Germany ...................... 0.8 0.7% 17. Morocco ...................... 0.6 0.5% 18. Japan ........................ 0.4 0.4% 19. Tunisia ........................ 0.4 0.3% 20. Turkey ........................ 0.2 0.2% Other ........................ 20.1 17.9% Total ........................ 112.5 © Telei~eography, in~c. 1999 National Traffic Balance Note: Data are in millions of minutes of outgoing public switched telecommunications traffm. © TeleGeography, inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 Romania Largest Telecommunications Routes, 1997 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Germany ..................... 20.5 18.5% 2. Italy ......................... 15.6 14.1% 3. Hungaw ...................... 7.7 6,9% 4. France ....................... 7.4 6.7% 5. Greece ....................... 5.9 5.3% 6. United Kingdom ................ 5.8 5.2% 7. Turkey ........................ 5.6 5.1% 8. United States .................. 4.9 ~~ 4.4% 9. Austria ....................... 4.8 4.3% 10. Netherlands ................... 2.8 2.5% 11. Switzerland ................... 2.3 2.1% 12. Belgium ...................... 2.2 2.0% 13. Israel .’ ....................... 2.1 1.9% 14. Canada ....................... 1.6 1.4% 15. Yugoslavia .................... 1.5 1.4% 16. Sweden ...................... 1.3 1.2% 17. Spain ........................ 1.2 1.1% 18. Bulgaria ...................... 1.1 1.0% 19. Russia ........................ 0.9 0.6% 20. Ukraine ....................... 0.8 0.7% Other ........................ 14.8 13.4% Total ........................ 110.8 © TeleGeography, Inc. 1999 National Traffic Balance Minutes 1996 Incoming. 237.5 ,, S,,,~_rp,.!~U~, ,,(,. D.~..f~! .,€ !t) 146.0 Total Volume 329.0 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. 1997 1998 278.6 n.a. 110.8 n.a. 167.8 n.a. 389.4 n.a. ©TeleGeography~ Inc. t999 © TeleGeography, Inc. 1999 TeleGeo{~raphy 2000 Russia Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Ukraine ..................... 290.8 2. Belarus ..................... 105.3 3. Kazakhstan ................... 76.9 4. Armenia ...................... 52.3 Germany ..................... 51.3 Georgia ...................... 40.8 Azerbaijan .................... 38.6 Moldova ..................... 33.6 Uzbekistan ................... 32.7 Latvia ........................ 20.5 Kyrgyzstan ................... 14.2 Lithuania ..................... 14.1 Tprkey ....................... 12.7 Italy ......................... 12.5 Finiand ....................... 12.3 United States ................. 12.2 United Kingdom ............... 11.9 France ....................... 10.8 Estonia ....................... 10.6 China ......................... 8.8 Other ....................... 175.3 6. 7. 8. 9. 11. 12. 13. 14. 17. 20. Percentage of Outgoing Traffic 28.0% ~~~ 10.1% ~~ 7.4% ~~ 5.o0/o 3.9% 3.70//o 3.2% 3.1% 2.0% 1.2% 1.20/o 1.2% 1.2% 1.1% 1.0% 1.0% 0.8% Total ...................... 1,038.3 © TeldGeography, inc. t999 National Traffic Balance Minutes 1996 1997 1998 ..... !~.?,~!~g.. 1.,037.6 1,035.6 !~0o~.~ .... ~Utg~!ng ............................. 851.3 ..... ~6~.6 ...................... 1~ ~,.~8.~ .... ~,~p~!~.§.~( D~ e~f!~ it) ............ 186.3 66.0 Total Volume 1,888.9 2,005.2 2,068.1 Note: Data are in milhons of minutes of outgoing public switched telecommunicatmns traffzc. Data are for Rostelecom only. Other operators sent an estimated 200 million minutes of additmnal traffic in 1998. © TeleGeogra phy, tnc. 1899 241 TeleGeography 2000 © TeleGeographU, Inc. 1999 Saudi Arabia Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Egypt ....................... 164.0 2. india ........................ 114.0 3. Pakistan ..................... 85.0 4. United Arab Emirates ........... 43.0 5. Syria ........................ 42.0 6. Bahrain ...................... 38.0 7. United States ................. 37.0 8. United Kingdom ............... 30.0 9. Yemen ....................... 28.0 10. Philippines ................... 28.0 11. Kuwait ....................... 27.0 12. Canada ...................... 22.0 13. Jordan ~,. ..................... 22.0 14. Bangladesh ................... 22.0 15. Sudan ....................... 22.0 Other ....................... 208.6 Total ........................ 932.6 Percentage of Outgoing Traffic 17.6% "~.~. ~-,.-.~12.2% ~9.1% 4.6% 4.5% ~4.1% 4.0% 3.2% 3.0% 3.0% 2.9% 2.4% 2.4% 2.4% 2.4% 22.4% ©TeleGeography, Inc. t999~ National Traffic Balance ~, Minutes 1996 1997 1998 In co~i~_n~g ...................... ~:~a.~ ...................... ~:~ ...................... ~,~’~ ..... ...... S~plus (Deficit) n.a. n.a. n.a. Total Volume n.a. n.a. n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. ¯ ©TeleGeography~ in c. 1999 242 © TeleGeography, Inc. 1999 TeleGeography 2000 Singapore Largest Telecommunications Routes, FY 1998/99 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Malaysia .................... 437.0 35.4% 2. t-long Kong ................... 90.0 7.3% 3. Indonesia .................... 85.0 6.9% 4. United States ................. 65.0 5.3% 5. Australia ..................... 60.0 4.9% 6. China ........................ 60.0 4.9% 7. Japan ........................ 50.0 4.0% 8. Thailand ..................... 40.0 3.2% 9. Philippines ................... 35.0 2.6% 10. India ......................... 30.0 2.4% 11. United Kingdom ............... 30.0 2.4% Other ....................... 253.0 20.5% Total ...................... 1,235.0 © Tele6eography, Inc. 1998 NationalTraffic Balance Minutes FY 1996/97 FY 1997/98 FY 1998/99 Incomi.,,,n,,g ,,n,;a,: ..................... £.,..a,,.. ..... ~0~u,,t,,,g~9 [n,g ....................................... 942.0 ,,,,,,1,,,,.,,!61~,0,, Surplus (Deficit) ..... ,n;,~,; Total Volume n.a. n.a. n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffzc. Data based on billing point of traffzc. Rscal year ends 31 March. © Tele~eography, ]~c. 1999 243 TeleGeography 2000 © TeleGeographg, Inc. 1999 ( Slovak Republic Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Czech Republic ................ 62.8 Germany ..................... 17.6 Austria ....................... 12.6 Hungary ....................... 7.7 United Kingdom ................ 6.3 Italy .......................... 5.6 Poland ........................ 4,5 Ukraine ....................... 3.7 United States .................. 3.6 France ........................ 3.0 Switzerland .................... 2.8 Russia ........................ 2.5 NetherlaQds ................... 2.3 Belgium ....................... 1.4 Yugoslavia ..................... 1.1 Croatia ........................ 1.1 Spain ......................... 1.0 Canada ....................... 0.9 Greece ........................ 0.9 Sweden ....................... 0.7 Other ......................... 9.4 Total ........................ 151.8 Percentage of Outgoing Traffic 41.4% 11.6% 8.3% 5.1% 3.7% 3.0% 2.4% 2.4% 2.0% 1.8% 1.7% 1.5% o.9% 0.8% 0.7% 0.6% 0.6% 0.6% 0.5% 6.2% ~TeleGqography, Inc. 1999 National Traffic~Balance Minutes 1996 1997 1998 ~p~.~ g~ 159.0 174.4 ..... ,,,0.,~,~ ............... 134.1 144.7 151:~ Total Volume 293.1 319.1 338.1 Note: Data are in millions of minutes of outgozng public switched telecommumcatzons traffic. Data based on billing poznt of traffic. 244 © TeleGeography,’,lnc. t999 © TeleGeography, Inc. 1999 TeleGeography 2000 Slovenia Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Croatia ....................... 31.3 24.1% 2, Germany ..................... 15.7 12.1% 3. Austria ....................... 14.6 11.3% 4. italy ......................... 14.4 11.1% 5. Yugoslavia .................... 13.2 10.2% 6. Bosnia-Herzegovina ............ 6.1 4.7% 7. Macedonia .................... 3.4 2.6% 8. Switzerland .................... 3.0 2.3% 9. United Kingdom ................ 2.9 2.2% 10. France ........................ 2.8 2.1% 11. Hungary ....................... 2.3 1.8% 12. United States .................. 2.1 1.6% 13. Russia ........................ 1.7 1.3% 14. Netherlands ................... 1.6 1.2% 15. Czech Republic ................. 1.4 1.1% 16. Sweden ....................... 1.2 0.9% 17. Belgium ....................... 1.1 0.9% 18. Poland ........................ 0.8 O.6% 19. Canada ....................... 0.8 0.6% 20. Chile .......................... 0.7 0.5% Other ......................... 8.5 6.6% Total ........................ 129.6 © Tele6eographyo Inc. 1989 National Traffic Balance Minutes Surplus (Deficit) Total Volume 1996 1997 1998 113.9 118.9 137.0 105.3 113.5 129.6 8.6 5.4 7.4 219.2 232.4 266.6 Note: Data are m milhons of minutes of outgoing public switched telecommunications traffic. © TeleGeogra phy, Inc. 1999 245 TeleGeography 2000 © TeleCeographg, Inc. 1999 South Africa Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Minutes (millions) United Kingdom ............... 85.0 United States ................. 38.0 Namibia ...................... 36.0 Zimbabwe .................... 32.0 Botswana .................... 16.5 Mozambique .................. 16.5 6ermany ..................... 15.0 Swaziland .................... 13.0 Australia ..................... 11.0 Lesotho ...................... 10.0 Canada ....................... 8.0 France ........................ 8.0 Netherlands ................... 8.0 Italy .......................... 7.0 Zambia ........................ 6.0 Other ........................ 95.0 Percentage of Outgoing Traffic 21.0% 9.4% ~~ 8.9% 7.9% ~4.1% ~4.1% 3.2% 2.7% 2.5% 2.O% 2.0% 2.0% 1.7% 1.5% 23.5% Total ........................ 405,0 © Tel~Geography, Inc. 1999 National Traffic Balance Minutes Surplus (Deficit) Total Volume 1996 1997 1998 n.a. 343.2 n,a. 353.0 368.8 405.0 n.a. (25.6) n,a. n.a. 712.0 n.a. Note: Data are in millions of minutes of outgoing public switched telecornmumcatzons traffic. 246 © TeleGeography, Inc. 1999 TeleGeography 2000 Largest Telecommunications Routes, 1998 Spain Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) United Kingdom .............. 270.0 France ..................... 265.0 (3erman¥ .................... 265.0 Italy ........................ 115.0 Portugal ..................... 77.0 United States ................. 76.0 Switzerland .................. 48.0 Belgium ..................... 47.0 Netherlands .................. 47.0 Morocco .................... 30.0 Argentina .................... 23.0 Andorra ..................... 22.0 Sweden ..................... 19.0 Chile ........................ 17.0 Mexico ...................... 17.0 Brazil ....................... 16.0 Colombia .................... 14.0 Denmark ..................... 13.0 Cuba ........................ 12.0 Austria ...................... 11.0 Other ....................... 201,0 Percentage of Outgoing Traffic 16.8% 16.5% 16.5% 7.2% 3.0% 2.9% 2.9% 1.9% 1.4% 1.4% 1.2% 1.t% 1.1% 1.0% 0.9% 0.8% 0.7% 0.7% ". ¯ ¯ .: ¯ .~12"5% Total ...................... 1,605.0 © Telel3eography, Inc. 1899 National Traffic Balance Minutes 1996 1997 1998 ~,~p] ~§J~ it,) ....................... n.a. n.a. n.a. Total Volume n.a. n.a. n.a. Note: Data are in mdhons of minutes of outgoing pubhc swztched telecommunications traffic. Data based on billing point of traffzc. © TeleGeography, tnc. 1999 247 TeleGeography 2000 © TeleGeographg, Inc. 1999 t Sri Lanka Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. India .......................... 5.7 14.5% 2. United Kingdom ................ 4.0 ~~ ,-.-~-- ~.~ 10.1% 3. Singapore ..................... 2.9 7.4% 4, Japan ........................ 2.2 5.7% 5, United States .................. 2.1 5.4% 6. Australia ...................... 1.8 4.5% 7, Germany ...................... 1.7 ~~ 4.2% 8, Hong Kong .................... 1.5 3.8% 9, United Arab Emirates ........... 1.3 3.3% 10, Canada ....................... 1.2 3.2% 11. Saudi Arabia ................... 1.1 2.9% 12. Korea, Rep ..................... 1.1 2.8% 13, Italy ...,~. ..................... 1.0 2.5% 14. Kuwait ........................ 0.8 2.0% 15. Maldives ...................... 0.8 2.0% 16. France ........................ 0.7 1.8% 17. Thailand ...................... 0.7 1.7% 18. Malaysia ...................... 0.6 1.5% 19. Switzerland .................... 0.6 1.5% 20. Pakistan ...................... 0.6 1.4% Other ......................... 7.0 17.8% Total ......................... 39.3 © TeleGeography, Inc. 1999 ¯ National Traffic Balance Minutes 1996 1997 1998 ~_,,,0, v,u,,~,g,,,,o,, !,,n,.~ 29.3 33.2 39.3 ¯ ~S~[plus (Deficit) .............................. 66.7 91.1 Total Volume 125.3 157.5 186.1 Note= Data are in millions of minutes of outgoing pubhc switched telecommunications traff=c. © TeleGeogrgphy,,In c. 1999 © TeleGeograph~l, Inc. 1999 TeleGeography 2000 Sudan Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Saudi Arabia ................... 6.6 Egypt ......................... 1.9 United Arab Emirates ........... 1.9 United Kingdom ................ 1.2 Qata r ......................... 0.6 United States .................. 0.5 China ......................... 0.5 Jordan ........................ 0.3 Canada ....................... 0.3 Germany ...................... 0.3 Malaysia ...................... 0.2 Netherlands ................... 0.2 Italy .......................... 0.2 France ........................ 0.2 India .......................... 0.2 Oman ......................... 0.2 Kenya ........................ 0.2 Switzerland .................... 0.2 Yemen ........................ 0.1 Libya ......................... 0.1 Other ......................... 2.4 Percentage of Outgoing Traffic 35.9% 10.1% 10.1% 6.3% 3.5% 2.9% 2.8% 1.8% 1.7% 1.6% 1.3% 1.2% 1.2% 1.2% 1.2% 0.9% 0.9% 0.9% 0.8% 0.7% 13.2% Total ......................... 18.4 ©TeleGeography, Jnc. 1999 National Traffic Balance Minutes 1996 1997 1998 ¯ ol,~ing ......... 30.4 43.2 .............. 88:0 ......... ~,v._~9~!,~g ....................................................... 12.9 14.8 18.4 S U.~P, I~us~(D~f~i~it) 17.5 28.~4 ........... 69:6, Total Volume 43.3 58.0 106.4 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic Data exclude some cross-border traffic to Chad. © Tele(~eegraphy, inc. 1999 249 TeleGeography 2000 © TeleGeographg, Inc. 1999 ( Sweden Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Minutes (millions) United Kingdom .............. 185.0 Finland ...................... 160.0 Norway" ..................... 150.0 United States ................ 125.0 Denmark .................... 120.0 Germany .................... 110.0 Poland ....................... 65.0 France ....................... 55.0 Netherlands .................. 40.0 Switzerland ................... 30.0 Other ....................... 165.0 Percentage of Outgoing Traffic 15.4% 13.3% ~.~ 10.4% ~~/- io.o% 4.6% 3.3% 2.5% 13.7% Total ’ ...1205.0 © TeleGeography, Inc, 1999 National Traffic Balance Minutes 1996 1997 1998 .... !~E~og~!~g n.a. n.a. n.a. ..... O...u_.t~.g..q!..n_~]. ........................................................ !.:976 ...9. .................. !.,...I.. 4~:.0’ ............. _...I. ,~_0....5.~0" ..... ~,U l,u_,s,,(_,~, ,e,fi~.!t,,,,) .............................................. .n,,.,a,,,., ................... ,n,.,,,a,,.: ...... n.a. Total Volume n.a. n.a. n.a. Note: Data are in milhons of m~nutes of outgoing public switched telecommumcations traffic. Data based on billing point of traffic. 1999 250 © TeleGeography, Inc. 1999 TeleGeography 2000 Switzerland Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Germany .................... 550.0 2. France ...................... 380.0 3. Italy ........................ 320.0 4. United Kingdom .............. 125.0 5. Austria ...................... tt0.0 6. United States ................ 105.0 7. Portugal ...................... 75.0 8. Spain ........................ 75.0 9. Netherlands .................. 60.0 t0. Yugoslavia .................... 55.0 Other ....................... 570.0 Percentage of Outgoing Traffic 22.7% ~... -. ...... ,’.’~15.7% ¯ -"- ’ \.. I:! 13.2% @~ 3.1% 2.5% 2.3% 23.5% Total ...................... 2,425.0 ¯ ©TeleGeography, Inc. 1999" National ,Traffic Balance Minutes 1996 1997 1998 ~.g~P !n~g ... 1,935.5 2,164.0 2,425.0 Total Volume 3,498.3 3,887.0 n.a. Note: Data are in millions of minutes of outgoing pubhc switched telecommunications traffzc. © TeleGeogra phy,, ]nc. 1999 251 TeleGeography 2000 © TeleGeography, Inc. 1999 Syria Largest TelecommunicationsRoutes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Lebanon ..................... 26.0 25.2% 2. SaudiArabia .................. 16.4 :~’_. .~ .~--:~ .,. -~. 15.9% 3. Jordan ........................ 7.2 7.0% 4. United Arab Emirates ........... 6.5 6.3% 5. Kuwait ........................ 5.3 5.1% 6. United States .................. 4.3 4.2% 7. Egypt ......................... 3.2 3.1% 8. Canada ....................... 2.6 2.5% 9. France ........................ 2.5 2.5% 10. Germany ...................... 2.3 2.2% 11. United Kingdom ................ 2.2 2.2% 12. Turkey ........................ 2.0 1.9% 13. Italy .. : ....................... 1.4 1.3% 14, Iraq .......................... 1.4 1.3% 15. Russia ........................ 1.2 1.2% 16, Sudan ........................ 0.8 0.8% 17. Greece ........................ 0.7 0.7% 18. Qatar ......................... 0.7 0.7% 19. Sweden ....................... 0.7 0.7% 20. Yemen ........................ 0.6 0.5% Other ........................ 14.9 .... 14.5% Total ........................ 103.0 National Traffic Balance ,© Tele~ieography, inc. t999 Minutes 1996 1997 1998 In c o m i~9 156.0 ............ ,~73~;~ .................... ~n~ .~ ...... ,,~0~ t,~g~!,n~ ....................... 78.9 ......... 89,,~ ...................... 1~.~ .......... Su[plus (Deficit) 77.1 83.9 n.a. Total Volume 234.9 262.5 n.a. Note: Data are in milhons of minutes of outgoing public switched telecommumcations traffic. Data based on billing point of traffic. ©TeleG~ography,~ln c. !999 252 © TeleGeography, Inc. 1999 TeleGeography 2000 Taiwan Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) China ....................... 243.1 United States ................ 143.3 Japan ........................ 72.4 Hong Kong ................... 46.5 Philippines ................... 38.5 Singapore .................... 23.7 Canada ...................... 23.1 Thailand ..................... 19.5 Indonesia .................... 17.5 Australia ..................... 16.8 Malaysia ..................... 11.9 United Kingdom ................ 9.5 Korea, Rep ..................... 9.3 Germany ...................... 9.3 Vietnam ....................... 6.3 France ........................ 6.1 New Zealand .................. 5.2 Russia ........................ 3.7 Macau ........................ 3.3 Netherlands ................... 3.1 Other ....................... 150.0 Percentage of Outgoing Traffic 28.2% : ~.-., ... ~.-~i ~6.6% 8.4% 5.4% 4.5% 2.8% 2.7% 2.3% 2.0% 1.9% 1.4% 1.1% 1.1% 1.1% 0.7% 0.7% 0.6% 0.4% 0.4% 0.4% "~ 17.4% Total ........................ 862.0 © T~ele6eography, Inc, 1999 National Traffic Balance Minutes FY 1996/97 1997 1998 Surplus (Deficit) 62.8 53.2 (80.2) Total Volume 1,410.8 1,631.2 1,643.9 Note: Data are in milhons of minutes of outgoing public switched telecommunications traff=c. Data based on billing point of traffic. Data based on calendar year reporting for 1997 and 1998 and fiscal year repo~ng for P-Y 1996/97. ©~eleGeography, Jnc. t999 253 TeleGeography 2000 © TeleGeographg, Inc. 1999 ( Thailand Largest Telecommunications Routes, 1998~ Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Japan ........................ 44.0 United States ................. 41).2 Singapore .................... 33.8 Hong Kong ................... 21.8 United Kingdom ............... 16.4 Taiwan ....................... 16.2 Australia ..................... 15.0 Germany ..................... 13.3 China ........................ 12.8 India .......................... 7.2 France ........................ 6.6 Korea, Rep ..................... 6.1 Philippines .................... 5.9 Indonesia ..................... 5.2 Switzerland .................... 4.3 Canada ....................... 3.9 Sweden ....................... 3.9 Vietnam ....................... 3.4 New Zealand .................. 2.1 Pakistan ...................... 1.8 Other ........................ 32.5 Percentage of Outgoing Traffic 14.6% 13.6% ~.. --. : ::- ’ -~11.4% 7.4% 5.5% 5.5% 5.1% 4.5% 4.3% 2.4% 2.2% 2.1% 2.0% 1.8% 1.5% 1.3% 1.3% 1.1% 0.7% 0.6% "~~ .~-~-. ’ -11.o% Total ........................ 296.4 ©~TeleBeography, Inc. 1999 NationalTraffic Balance ¯ Minutes 1996 1997 I n~,,0om~,~ ........ ~7~.~ ....................... ~.u..,r~!.M.s.....(~,~t!,~!~ ..................... 128.8 130.1 Total Volume 623.6 1998 358.6 296.4 62.2 655.0 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. Data exclude some cross-border traffic with Laos, Malaysia, and Myanmar. ©Te[eGeography, Inc. t999 254 © TeleCeograph~l, Inc. 1999 TeleGeography 2000 Trinidad & Tobago Largest Telecommunications Routes, FY 1998/99 Destination Minutes (millions) Percentage of Outgoing Traffic 1. United States ................. 30.5 47.4% 2. Canada ....................... 5.8 9.0% 3. United Kingdom ................ 4.1 6.3% 4. Barbados ..................... 3.4 5.3% Grenada ...................... 2.7 4.1% Guyana ....................... 2.0 3.1% Jamaica ...................... 2.0 3.1% Saint Vincent & the Grenadines...1.9 3.0% Venezuela ..................... 1.7 2.6% Saint Lucia .................... 1.5 2.3% Antigua & Barbuda ............. 0.9 1.4% Dominica ...................... 0.5 0.7% Saint Kitts & Nevis .............. 0.4 0.7% Germany ...................... 0.4 0.7% Netherlands Antilles ............ 0.4 0.6% British Virgin Islands ............ 0.4 0.6% India .......................... 0.3 1 0.4% France ........................ 0.3 0.4% Suriname ...................... 0.2 0.4% Mexico ....................... 0.2 ! 0.3% Other ......................... 4.8 7.4% 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Total ......................... 64.4 ¯ © TeleGeography, Inc. t999 National Traffic Balance Minutes FY 1996/97 FY 1997/98 9o98/99 ,,v~ .o, in g 60.8 62.0 .............. ,.6~.~o~, .~.~plus (Deficit) ............................ ~]~!o .................. 6,~. 0~ ......................... 7~ .~ Total Volume 178.7 192.0 206.0 Note= Data are in milhons of m~nutes of outgoing public sw=tched telecommumcat~ons traffic. Data based on billing point of traffic. Fiscal year ends 31 March. ©TeleGeography, tnc. t999 255 TeleGeography 2000 © TeleCeographu, Inc. 1999 ( Turkey Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Germany .................... 186.4 28.9% 2. United Kingdom ............... 48.5 7.5% 3. France ....................... 34.3 5.3% 4. Netherlands .................. 28.4 4.4% 5. United States ................. 24.6 3.8% 6. Russia ....................... 21.9 3.4% 7. Italy ......................... 20.1 3.1% 8. Romania ..................... 17.1 2.7% 9. Bulgaria ...................... 16.1 2.5% 10. Switzerland ................... 15.4 2.4% 11. Austria ....................... 15.2 2.4% 12. Belgium ...................... 12.7 2.0% 13. Ukraine ’. ..................... 10.1 1.6% 14. Azerbaijan .................... 10.1 1.6% 15. Greece ........................ 9.3 1.4% 16. Iran .......................... 6.6 1.0% 17. Saudi Arabia ................... 5.6 0.9% 18. Israel ......................... 5.6 0.9% 19. Sweden ....................... 4.9 0.8% 20. Denmark ...................... 4.9 0.8% Other ....................... 146.3 22.7% Total ........................ 644,1 ©Tele6eography, Inc. 1999 National Traffic Balance Minutes 1996 1997 1998 _,,,!,,,nc,~.m, !,n,,,9!. ........................ 755.0 ..... ,8,3.,,6.~ .............. ,9, 5,.5.,.~ ........... ..... 9~,~,t~g 473.4 557.5 .... 64~.,!~ .......... ~,U,[plus (Deficit) 281.6 278.5 311.7 Total Volume 1,228.4 1,393.5 1,600.0 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data exclude some cross-border traffic to Irsn. ©TeieGeography; Inc. t999 256 © TeleGeographM, Inc. 1999 TeleGeography 2000 Ukraine Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Russia ...................... 280.3 60.2% 2. Belarus ...................... 26.1 5.6% 3. Moldova ..................... 18.6 4.0% 4. Germany ..................... 17.0 3.7% Poland ....................... 11.1 2.4% Armenia ....................... 6.5 1.4% United States .................. 5.7 1.2% Czech Republic ................. 5.3 1.1% Kazakhstan .................... 5.1 1.1% United Kingdom ................ 4.6 1.0% Latvia ......................... 4.5 1.0% Azerbaijan ..................... 4.3 0.9% FLungary ....................... 4.2 0.9% Georgia ....................... 4.0 0.9% Italy .......................... 4.0 0.9% Uzbekistan .................... 4.0 0.9% Turkey ........................ 3.7 0.8% Israel ......................... 3.5 0.8% Lithuania ...................... 3.3 0.7% Bulgaria ....................... 2.7 1 0.6% Other ........................ 47.3 10.2% 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Total ........................ 465.9 © TeleGeography, inc. 1999 National Traffic Balance Minutes 1996 1997 1998 ...... !~.~!~ 9 175.0 n.a. n.a. .... ~,tg~!~,~ ............................ ~0:8~ ..... 486.8 465.9 ....~plus (Deficit) (165.8) Total Volume 515.8 n.a. n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Totals for 1996 include only traffic with other members of the Commonwealth of Independent States. © Tel~eographyo tnc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 United Arab Emirates Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. india ........................ 205.0 2. Pakistan ..................... 88.0 3. Egypt ........................ 50.0 4. Saudi Arabia .................. 45.0 5. United Kingdom ............... 45.0 6. United States ................. 34.0 7. Syria ........................ 32.0 8. Oman ........................ 31.0 9. Iran ......................... 26.0 10. Jordan ....................... 21.0 11. Philippines ................... 21.0 12. Bangladesh ................... 20.0 13. Lebanon ...................... 20.0 14. Qatar ........................ 18.0 15. Bahrain ...................... 15.0 Other ....................... 203.8 Percentage of Outgoing Traffic 23.4% 10.1% 5.7% ~5.1% ~5.1% 3.5% 3.0% 2.4% 2.4% 2.3% ~2.1% 1.7% 23.3% Total ........................ 874.8 ©TeleGeography, 1nc.3999 ,National Traffic Balance Minutes Incoming~ ~Y[~YS (D efi c it) Total Volume 1996 1997 1998 n.a. n.a. Il.a. 589.3 738.0 874.8 n.a, n.a. n.a. n.a. n.a. n.a. Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. ..... "ole~o----le-e graphy,~lnc. 1999 258 © TeleGeography, Inc. 1999 TeleGeography 2000 United Kingdom ,Outgoing Largest Telecommunications Routes, FY 1998/99 Destination Minutes (millions) 1. United States ............... 1,260.0 2. Ireland ...................... 815.0 3. Germany .................... 725.0 4. France ...................... 650.0 5. Italy ........................ 360.0 6. Spain ....................... 360.0 7. Canada ..................... 310.0 8. Netherlands ................. 305.0 9. Australia .................... 240.0 10. Belgium ..................... 165.0 11. Switzerland .................. 150.0 12. Greece ...................... 130.0 13. Sweden ..................... 120.0 14. I-lonG Kong .................. 115.0 15. Turkey ...................... 105.0 16. South Africa ................. 100.0 17. india ......................... 95.0 18. Denmark ..................... 95.0 19. Pakistan ..................... 90.0 20. Poland ....................... 85.0 Other ...................... 1,950.0 Percentage of Outgoing Traffic ~~9.9% 8.8% 7.9% 4.4% 4.4% 3.8% 3.7% 2.9% 2.0% 1.8% 1.6% 1.5% 1.4% 1.3% 1.2% 1.2% 1.2% 1.1% 1.0% .7% Total ...................... 82.25.0 ©Te!eBeograpfly, Inc. t899 NationalTraffic Balance Minutes FY 1996/97 ........ !~P~!~g .......... 4,360.0 .... ~,~g0i~g .......................... 4,569.2 Surplus (Deficit) (209.2) Total Volume 8,929.2 FY 1997/98 FY 1998199 n.a. (1,825.0) n.a. 14,625.0 Note: Data are in mdhons of m~nutes of outgoing public switched telecommumcations traffic. Data based on billing point of traffic. Fiscal year ends 31 March. Data include approximately one billion minutes of traffic refiled via the U.K., thus overstating U.K.-originat- ed volumes. FY 1996/97 totals are for BT and C&W Communications only. © Teie~Geogra phy, ~nc. 1899 TeleGeography 2000 © TeleGeogrophy, Inc. 1999 ( United Kingdom--Incoming Largest Telecommunications Routes, FY 1998/99 Destination Minutes (millions) Percentage of Outgoing Traffic 1 United States .............. 1,530.0 23.9% 2 Ireland ..................... 605.0 9.5% 3 France ..................... 485.0 7.6% 4 Germany .................... 400.0 6.3% 5 Spain ....................... 270.0 4.2% 6 Australia .................... 240.0 3.8% 7 Canada ..................... 230.0 3.6% 8 Nethedands ................. 225.0 3.5% 9 Italy ........................ 220.0 3.4% 10 Sweden .................... 185.0 2.9% 11 Belgium .................... 140.0 2.2% 12 Switzerland ................. 125.0 2.0% 13 Greece ..................... 100.0 1.6% 14 Denmark ..................... 90.0 1.4% 15 Norway ...................... 70.0 1.1% 16 South Africa .................. 85.0 1.3% 17 Hong Kong ................... 70.0 1.1% 18 Japan ....................... 50.0 0.8% 19 Portugal ..................... 50.0 0.8% 20 Turkey ....................... 50.0 0.8% Other ..................... 1,180.0 .:o_ ; ................. 18.4% Total ...................... 6,400.0 Inc. ~999 260 © TeleGeographg, Inc. 1999 TeleGeography 2000 U.K. Top 100 Correspondents Outgoing Incoming Outgoing Incoming Country Q3 96197 Q3 97/98 Q3 96/97 03 97/98 Country Q3 96/97 Q3 97/98 Q3 96/97 Q3 97/98 Albania n.a. 1.7 n.a. 0.2 Libya 1.6 1.5 1.2 1.4 Algeria 4.8 3.5 2.0 2.1 Lithuania 1.2 1.8 0.2 0.3 Argentina 2.2 2.7 1.2 1.1 Luxembourg 6.0 6.5 3.8 3.9 Australia 55.9 65.3 59.3 67.6 Macedonia n.a. 1.6 n.a. 0.4 Austria 19.8 25.7 8.4 7.2 Malaysia 6.3 5.8 6.2 3.7 Bahrain 2.3 2.2 3.1 1.6 Mall n.a. 2.6 n.a. 0.7 Bangladesh 5.6 5.8 0.9 2.3 Malta 3.6 4.0 2.8 3.1 Barbados 2.2 2.3 1.1 1.0 Mauritius 2.8 3.2 1.1 2.1 Belgium 43.8 37.2 32.5 33.3 Mexico 2.7 2.8 1.3 1.3 Brazil 5.8 6.7 4.1 4.1 Monaco 1.7 3.0 1.4 1.9 Bulgaria 2.2 3.8 0.9 0.9 Morocco 9.3 7.3 1.5 2.7 Cameroon n.a. 1.5 n.a. 0.3 Netherlands 61.9 63.2 49.0 55.7 Canada 42.6 48.8 38.9 53.8 New Zealand 12.1 15.0 12.8 14.2 Chile 2.5 3.6 1.0 0.9 Nigeria 11.9 16.4 3.7 4.9 China 8.4 6.7 3.0 3.6 Norway 17.6 21.6 16.9 21.1 Colombia 2.8 7.0 1.0 1.2 Oman 2.1 1.9 1.5 1.5 Congo, Dem. Rep. 2.1 1.3 0.5 0.7 Pakistan 22.5 23.5 7.1 4.0 Croatia 3.6 3.3 1.5 1.7 Peru n.a. 1.5 n.a. 0.4 Cuba n.a. 1.7 n.a. 1.5 Philippines 4.0 5.3 1.6 1.3 Cyprus ’, 9.3 10.9 7.8 7.6 Poland 20.2 28.0 8.0 9.4 Czech Republic 8.9 11.3 4.6 4.5 Portugal 19.4 19.8 11.0 13.1 Denmark 22.1 23.5 15.2 16.6 Qatar 2.2 1.6 1.6 1.7 Ecuador 1.1 1.7 0.2 0.2 Romania 2.9 5.8 1.7 1.8 Egypt 7.2 9.1 2.6 2.9 Russia 19.8 14.7 8.5 3.5 Ethiopia 2.2 1.4 0.4 0.3 Rwanda n.a. 1.5 n.a. 0.0 Finland 9.2 11.1 7.4 5.4 Saudi Arabia 15.3 7.7 12.5 6.9 France 147.4 135.0 101.4 106.0 Senegal 2.9 1.3 0.2 0.2 Gambia n.a. 1.7 n.a. 0.3 Singapore 13.0 8.8 6.1 5.2 Germany 151.3 177.9 98.6 81.6 Slovak Republic 3.9 4.9 15.0 2.0 Ghana 5.7 9.2 2.4 2.7 Slovenia 1.3 1.6 0.8 0.9 Gibraltar 2.1 2.4 1.5 1.7 South Africa 29.6 25.5 18.9 21.0 Greece 29.2 31.0 21.7 25.0 Spain 73.4 83.1 52.0 59.0 Hong Kong 25.1 22.4 12.9 7.6 Sri Lanka 4.3 6.5 1.0 1.2 Hungary 8.2 13.2 4.1 4.0 Sweden 26.8 37.2 41.2 38.3 Iceland 1.6 1.7 1.2 1.4 Switzerland 35.6 35.7 27.6 23.7 India 25.6 33.0 11.8 9.4 Syria 2.1 4.3 0.5 0.6 Indonesia 4.2 2.0 3.0 2.0 Taiwan 3.4 3.7 2.3 2.3 Iran 4.0 5.2 2.2 3.1 Tanzania 1.3 2.2 0.5 0.6 Ireland 190.3 197.4 128.5 149.2 Thailand 4.3 5.2 3.8 3.6 Israel 10.1 10.7 4.3 6.4 Trinidad & Tobago 3.7 2.0 1.6 1.2 Italy 81.4 90.7 46.7 43.8 Tunisia 3.9 3.9 0.7 0.9 Jamaica 6.7 6.5 1.3 1.3 Turkey 24.0 30.0 10.1 12.6 Japan 19.4 20.4 12.8 11.5 UAE 9.5 11.1 11.5 9.4 Jordan 3.4 3.9 1.0 1.5 Uganda n.a. 1.8 n.a. 0.6 Kazakhstan n.a. 1.8 n.a. 0.6 Ukraine 3.2 8.6 1.1 1.3 Kenya 3.7 4.4 1.6 1.8 United States 264.0 323.0 332.8 344.0 Korea, Rep. 4.8 2.5 3.7 2.7 Vietnam 7.5 5.0 0.9 4.3 Kuwait 4.1 4.4 2.0 2.1 Yemen 2.3 1.7 0.3 1.3 Latvia 2.0 4.0 0.8 0.5 Yugoslavia 2.8 4.9 1.3 1.3 Lebanon 4.4 4.4 0.6 1.6 Zimbabwe 2.4 1.9 2.2 2.6 Note: Data are millions of minutes of public switched traffic for the largest U.K. international carriers during the third quarters (October-December) of FY 1997/98 and FY 1998/99. Route data may include some calls refiled via the U.K., thus overstating actual U.K.-ong=nated traffic. Source: Dffice of Telecommunications (OR’EL) © TeleGeography, Inc. 1999 261 TeleGeography 2000 © TeleGeograph~j, Inc. 1999 United States--Outgoing Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1, Canada .................... 3,988.0 2. rvlexico .................... 3,117.6 3. United Kingdom ............. 1,531,4 4. Germany ................... 1,194.2 5. Japan ....................... 833,3 6. India ........................ 772.9 7. Hung Kong .................. 635.6 8. Brazil ....................... 592.3 9. France ...................... 580.2 10. Philippines ................... 573.1 11. Italy ........................ 532.2 12. Australia .................... 459.7 13. Dominican Republic ........... 452.9 14. China ....................... 434.5 15. Korea, Rep ................... 397.3 16. Taiwan ...................... 327.9 17. Jamaica ..................... 282.1 18. Netherlands ................. 273.2 19. Venezuela ................... 231.1 20. Spain ....................... 230.9 Other ...................... 7,508.8 Percentage of Outgoing Traffic 16.0% 12.5% 6.1% 4.8% 3.3% 3.1% 2,5% 2.4% 2.3% 2.3% ~2.1% 1.8% 1.8% 1.7% 1.6% 1.3% 1.1% 1.1% 0.9% 0.9% 30.1% Total ..................... 24,949.3 ,., ........... ..© TeleG,eography, ln~c. 1999 National Traffic Balance ¯ Minutes Inco~.io~g ......................... Surplus (Deficit) Total Volume 1996 1997 1998 (10,924.2) (13,788.4) (14,350.5) 27,314.0 32,215.0 35,548.1 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffzc. Data include include one to two billion minutes of traffic refiled via the U.S., thus overstating traffic originating from the U.S, Carriers and traffic from pmnta beyond the United States, Puerto Rico, and the U.S. Virgin Islands are excluded. ©TeleGeography, lQc~ t999 262 © TeleGeographg, Inc. 1999 TeleGeography 2000 United States--Incoming Largest Telecommunications Routes, 1998 Destination Minutes (millions) Canada ................... 3,400.8 United Kingdom ............ 1,261.4 Mexico .................... 1,095.4 Germany .................... 394.7 Japan ...................... 338.4 France ..................... 289.1 Australia .................... 242.1 Korea, Rep .................. 231.5 Sweden .................... 218.8 Israel ....................... 180.3 Italy ........................ 170.7 Brazil ....................... 189.6 Dominican Republic .......... 154.2 Taiwan ..................... 152.9 Netherlands ................. 122.2 Switzerland ................. 102.1 Chile ........................ 81.1 China ....................... 79.4 Ireland ...................... 77.7 Spain ....................... 78.0 Other ..................... 1,784.5 Percentage of Outgoing Traffic 32.1% 11.9% 10.3% 3.7% 3.2% 2.7% 2.3% 2.2% 2.0% 1.7% 1.6% 1.6% ~, 1.5% 1.4% 1.2% 1.0% 0.8% 0.7% 0.7% 0.7% 16.6% Total ..................... 10,598.8 ¯ ©TeleGeography, Inc. 1999 263 TeleGeography 2000 © TeleGeograph~l, Inc. 1999 Top 100 Correspondents Outgoing Minutes Incoming Minutes Outgoing Minutes Incoming Minutes Country 1997 1998 1997 1998 Country 1997 1998 1997 1998 Antigua & Barbuda 52.6 Argentina 228.1 Aruba 16,8 Australia 383,5 Austria 58.8 Bahamas 66,4 Bangladesh 59.0 Barbados 38.4 Belgium 120.7 Belize 14.5 Bermuda 38.7 Bolivia 34.5 Brazil 497.3 Bulgaria 12.2 Canada 3,926.6 Cayman Islands 21.9 Chile 115.5 China 388.1 Colombia 260.5 Congo, Dem. Rep. ,8.0 Costa Rica 75.3 C6te d’lvoire 14.8 Croatia 22.5 Cuba 118.9 Czech Republic 26.9 Denmark 75.3 Dominica 16.4 Dominican Republic 392.1 Ecuador 156.7 Egypt 111.2 El Salvador 154.6 66.9 6.7 7.1 Jamaica 262.6 282.1 50.2 51.9 220.8 38.8 46.1 Japan 830.5 833.3 339.4 336.4 25.6 6.6 5.7 Jordan 57.1 51.0 5.4 6.8 459.7 123.5 242.1 Kenya 24.2 24.3 3.5 4.1 74.3 24.9 26.1 Korea, Rep. 421.9 397.3 203.5 231.5 81.6 46.5 49.7 Kuwait 48.3 46.3 8.8 8.8 51.9 4.5 4.7 Lebanon 38.6 56.6 4.4 11.2 46.2 12.8 12.6 Luxembourg 10.9 15.4 6.2 6.9 116.1 49.7 49.7 Malaysia 83.4 101.5 26.4 25.2 15.3 4.2 5.0 Mexico 2,757.6 3,117.6 948.5 1,095.4 42.7 29.9 30.0 Morocco 13.8 57.2 5.8 6.4 49.3 3.6 5.8 Netherlands 222.3 273.2 105.8 122.2 592.3 159.6 169.6 Netherlands Antilles 42.3 61.7 8.1 7.2 14.5 1.3 1.6 New Zealand 107.8 84.8 40.6 64.8 3,988.0 3,173.3 3,400.8 Nicaragua 50.3 62.7 6.1 8.2 26.2 12.1 14.2 Nigeria 101.8 158.9 8.7 17.4 125.8 75.1 81.1 Norway 66.5 89.0 52.5 42.2 434.5 61.7 79.4 Pakistan 162.9 210.9 10.5 15.6 218.0 68.7 51.2 Pana ma 68,2 68.5 19.3 18.9 19.2 2.0 0.2 Paraguay 10,4 15.4 3.3 3,7 73.2 26.4 26.6 Peru 164.9 182.9 24.5 31.4 16.3 4.4 3,4 Philippines 417.2 573.1 32.2 50.8 23.9 6.0 5.1 Poland 170.9 189.6 27.3 27.8 143.4 1.2 1.2 Portugal 54,8 71.6 15.1 23.1 30.3 9.2 9,9 Romania 28.9 74.9 5.2 31.5 75.0 27,9 26.5 Russia 102.9 154.9 32.9 22.0 29.1 1.9 2,1 Saint Lucia 14,7 19.5 3.2 3.5 452.9 95.6 154.2 Saint Vincent 16,7 14.9 1.7 2.0 112,2 17.8 13,6 Saudi Arabia 111.4 108.1 32.1 36.9 111.5 10.7 11,7 Senegal 18,9 19.5 1.7 2.0 198.1 12.8 16.0 Sierra Leone 11.3 41.6 1.2 1.2 21.4 1.6 1.8 Singapore 197.8 168.8 63.5 65.8 30.6 17.8 14.1 South Africa 110,6 129.8 31.2 36.5 580.2 216.6 289.1 Spain 190.8 230.9 88.1 76.0 1,194.2 324.6 394.7 Sri Lanka 16.2 20.0 2.0 3.1 44.0 5.2 8.0 Sweden 140.0 149.1 100.6 216.6 115.7 35.0 42.8 Switzerland 225.8 196.4 108.1 102.1 63.5 2.4 2.3 Syria 15.2 15.5 2.9 4.1 144.9 15.5 31.2 Taiwan 379.1 327.9 174.3 152.9 48.6 12.5 7.8 Thailand 117.7 120.7 26.2 32.2 99.7 5.1 7.1 Trinidad & Tobago 94.4 108.3 26.2 31.5 149.1 11.5 13.4 Turkey 62.0 105.4 28.3 42.7 635.6 82.4 68.2 Ukraine 44.4 48.5 6.9 6.3 37.7 15.2 13.4 United Arab Emirates 52.3 59.7 28.4 32.2 772.9 49.7 58.5 United Kingdom 1,532.6 1,531.4 950.2 1,261.4 99.8 28.1 28.6 Uruguay 26.2 33.6 7.5 8.2 63.7 16.0 17.7 Venezuela 219.8 231.1 63.9 60.2 163.8 49.0 77.7 Vietnam 157.0 190.4 4.0 4.5 223.2 161.7 180.3 Yemen 23.9 28.5 1.2 1.3 532.2 121.9 170.7 Yugoslavia 15.7 33.9 5.5 8.4 Ethiopia 18.0 Finland 29.4 France 500.0 Germany 994.3 Ghana 50.0 Greece 97.4 Grenada 16.9 Guatemala 126.1 Guyana 59.7 Haiti 92.2 Honduras 111.1 Hong Kong 673.2 Hungary 41.0 India 579.4 Indonesia 119.2 Iran 48.0 Ireland 134.8 Israel 216.5 Italy 475.6 Note: All data are millions of minutes of public switched and International Simple Resale (ISR) traffzc. Because data are based on the billing point of the traffic, route data may not exactly correspond with traffic volumes as measured I}y the ortgmat~n9 point of traffic (see Methodology on page 2"/1 I. Carriers and traffic from outside the ~J.S. states, Puerto Rico, and the U.S. Virgin Islands (e.g., Guam) are excluded. © TeleGeography, Inc. 1999 264 © TeleGeography, Inc. 1999 TeleGeography 2000 Uruguay Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Argentina .................... 40.6 Brazil ........................ 10.5 United States .................. 7.5 Spain ......................... 3.2 Chile .......................... 2.0 Paraguay ...................... 1.7 Italy .......................... 1.5 France ........................ 0.9 Germany ...................... 0.8 Mexico ....................... 0.8 Canada ....................... 0.7 Venezuela ..................... 0.7 United Kingdom ................ 0.6 Peru .......................... 0.5 Australia ...................... 0.4 Colombia ...................... 0.4 Switzerland .................... 0.4 Bolivia ........................ 0.3 Netherlands ................... 0.3 Sweden ....................... 0.2 Other ......................... 4.3 Percentage of Outgoing Traffic 51.9% 13.4% 9.6% ~4.1% 2.6% 2.2% 1.9% 1.1% 1.o% 1.o% 0.9% 0.9% 0.8% 0.6% 0.5% 1 0.5% 0.5% 0.4% 10.4% 10.3% 5.5% Total ......................... 78.3 © Telei~eography, Inc, 1989, National Traffic Balance Minutes 1996 ........ !.n..,€....,o,_m,,,,.~n,,~9 80.1 ¯ ,Q,u,,tg ~in,.g ............. 54.5 Total Volume 134.6 1997 1998 93.7 97.0 68.4 78.3 25.3 18.7 162.1 175.3 Note: Data are in millions of minutes of outgoing pubhc switched telecommunications traffic. ©TeleSeography, Inc. 1999 265 TeleGeography 2000 © TeleGeograph~t, Inc. 1999 Uzbekistan Largest Telecommunications Routes, 1998 Destination Minutes (millions) Percentage of Outgoing Traffic 1. Russia ....................... 32.3 35.2% 2. Kazakhstan ................... 18.5 ~__~ ~" ..~20.1% 3. Kyrgyzstan .................... 6.7 7.3% 4. Tajikistan ...................... 4.8 5.2%° 5. Ukraine ....................... 4.0 4.4% 6. Turkmenistan .................. 3.0 3.2% 7. Germany ...................... 1.6 1.7% 8. United States .................. 1.5 1.6% 9. Belarus ....................... 1.2 1.3% 10. China ......................... 1.1 1.2% 11. Korea, Rep ..................... 0.9 0.9% 12. United Kingdom ................ 0.8 0.9% 13. Armenia’, ...................... 0.7 0.8% 14. Azerbaijan ..................... 0.7 0.8% 15. France ........................ 0.5 0.5% 16. Georgia ....................... 0.4 0.5% 17. Japan ........................ 0.4 0.4% 18. Italy .......................... 0.4 0.4% 18. Switzerland .................... 0.3 10.3% 20. Moldova ...................... 0.2 0.3% Other ........................ 11.9 13.0% Total ......................... 91.7 266 ,+ National Traffic Balance ¯ +++ Minutes 1996 1997 1998 ¯ +S+u.¯,rp~,U,,S" (D+e,,fici~) ..... n.a. .......... ~:a,,., .............. (~7.+0,), Total Volume n.a. n.a. 166.5 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Totals for 1996 include only traffic with other members of the Commonwealth of Independent States. © TeleGeog~aphy, Jnc. 1999 © TeleGeography, Inc. 1999 TeleGeography 2000 Venezuela Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. United States ................. 74.5 2. Colombia ..................... 15.8 3. Canada ...................... 12.9 4. Spain ......................... 9.4 italy .......................... 7.6 United Kingdom ................ 6.2 Peru .......................... 4.1 Mexico ....................... 3.4 Trinidad & Tobago .............. 3.0 Argentina ..................... 3.0 Ecuador ....................... 2.6 Brazil ......................... 2.5 P, ortugal ....................... 2.5 France ........................ 2.4 Dominican Republic ............. 2.1 Chile .......................... 1.8 Germany ...................... 1.4 Cuba ......................... 1.3 Panama ....................... 1.1 Sweden ....................... 0.2 Other ......................... 6.9 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18, 19. 20. Percentage of Outgoing Traffic 45.3% 9.6% 7.8% 5.7% 4.6% 2.5% 2.0% 1.8% 1.8% 1.6% 1.5% 1.5% 1.5% 1.2% 1.1% 0.9% 0.8% 0.7% i 0.1% 4.2% Total ........................ 164.6 © TeleGeography, inc,,1899" National Traffic Balance Minutes 1996 1997 1998 ..... !n..£,O~!.n_9. ..... .,2.,.2,8.8 .... ,o...y.t.goin.g .............. 139.0 159.2 16,.4.;5 ~...u~p,,.!.y.,s..,,,(,,,.D,,e. ,f,!...,c,,!,,t,..)..’ ...................................... ,,,,,8,9,:,.,8~ ....................... !,2,.7:7, ............ I~:~ .............. Total Volume 367.8 446.1 462.6 Note: Data are in millions of minutes of outgoing public switched telecommunications traffm. Data based on bdhng point of traffic. © TeleBeog raphy, f~c. 1999 267 TeleGeography 2000 f Vietnam Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. Australia ..................... 11.0 2. Japan ........................ 6.0 3. Taiwan ........................ 5.0 4. United States .................. 4.5 5. United Kingdom ................ 3.5 6. Korea, Rep ..................... 3.0 7. China ......................... 2.5 8. Hong Kong .................... 2.0 9. Thailand ...................... 2.0 10. Singapore ..................... 1.5 Other ........................ 15.0 © TeleGeography, Inc. 1999 Percentage of Outgoing Traffic "-:/ --!.. ~. ¯ ~, 19.6% 10.7% 8.9% 8.0% 6.3% 5.4% 4.5% 3.6% 3.6% 2.7% 26.8% Total ...’ ...................... 56.0 © Tel~Geography,,Inc, 1999 National Traffic Balance Minutes 1996 1997 1998 ...... ~r.plus (Deficit} .............................. n.a. ~2~4:4~ ............... ~7,,,8:,0.. ........ Total Volume n.a. 344.8 390.0 Note: Data are in millions of minutes of outgoing public switched telecommunicationstraffzc. © TeleGeography, Inc. 1999 268 © TeleGeography, Inc. 1999 TeleGeography 2000 Yugoslavia Largest Telecommunications Routes, 1998 Destination 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Minutes (millions) Germany ..................... 41.1 Croatia ....................... 22.6 Austria ....................... 19.7 Switzerland ................... 18.6 Italy ......................... 13.0 Bosnia-Herzegovina ........... 10.5 Macedonia .................... 8.9 France ........................ 7.7 United States .................. 7.0 Canada ....................... 6.5 Greece ........................ 6.0 United Kingdom ................ 5.2 Russia ........................ 4.4 Sweden ....................... 4.4 Netherlands ................... 3.2 Romania ...................... 2.3 Bulgaria ....................... 2.1 Australia ...................... 2.0 Belgium ....................... 1.6 Denmark ...................... 1.2 Other ........................ 31.8 Percentage of Outgoing Traffic 18.7% ~- .~--~ -~, lO.3% 9.0% 8.5% ~~ 5.9% 4.8 ~/o 4.0% 3.5% 3.2% 3.0% 2.4% 2.0% 2.0% 1.5% 1.1% 0.9% 0.9% 0.7% 0.5% 14.5% Tota I ........................ 219.5 © TeleGeograp,hy, Inc. 1999 National Traffic Balance Minutes 1996 1997 1998 .... !~£~m!~ng ............ 325.7 332.0 423.3 ....... ~,,,u,~g~,,,n,,~g, ........... 237.2 ....... ,,2,,.17:0,,~ 219.5 S,,ur.p,!us (.D,,,efic~!~),.~i~,,,,.,.il ........................... 8#:,5, ....... 115.0 .......... ,2,0~,:,8~ Total Volume 562.9 549.0 642.9 Note: Data are in millions of minutes of outgoing public switched telecommunications traffic. Data based on billing point of traffic. © TeleGeography, ]nc.’1999 269 TeleGeography 2000 © TeleGeographU, Inc. 1999 ( Zimbabwe Largest Telecommunications Routes, 1998 Destination Minutes (millions) 1. South Africa .................. 22.5 2. United Kingdom ................ 4.1 3. United States .................. 3.1 4. Botswana ..................... 2.7 5. Germany ...................... 0.8 6. india .......................... 0.7 7. Australia ...................... 0.7 8. Kenya ........................ 0.7 9. France ........................ 0.6 10. Netherlands ................... 0.5 11. Namibia ....................... 0.4 12. Switzerland .................... 0.4 13. Canada ’, ...................... 0.4 14. Belgium ....................... 0.4 15. Italy .......................... 0.3 16. Japan ........................ 0.3 17. Mauritius ...................... 0.3 18. Denmark ...................... 0.3 19. Sweden ....................... 0.3 20. Tanzania ...................... 0.3 Other ........................ 13.0 Total ......................... 52.8 Percentage of Outgoing Traffic 42.6% 7.9% 5.8% 5.0% 1.5% 1.3% 1.2% 1.2% 1.1% 1.0% 0.8% 0.8% 0,8% 0.7% 0.7% 0.7% 0.6% 0.6% 0.5% 0.5% ",+. _ 24.7% © TeleGeo~raphy~ Inc, t999 National Traffic Balance- Minutes 1996 1997 1998 _!_u,._t..9o_.i.n..l. 48.9 ,!:!:_ .................. ~o.,.! ...... ,,S,._u,,rplus (Deficit) n.a. _ ...... .n.:a_: ......................... o..]. Total Volume n.a. n.a. 106.0 Note: Data are in milhons of minutes of outgoing public switched telecommunications traffic. ¯ ©TeleGeography, no. 1999 270 © TeleGeography, Inc. 1999 TeleGeography 2000 Methodology The traffic statistics in TeleGeography 2000 were compiled pri- marily from an independent survey of telecommunications ser- vice providers by TeleGeography, Inc. (TGI). For some countries and carriers, traffic data have been estimated based upon annual reports, government publications, and industry inter- views. See the footnotes to each table for further information. The Direction of Traffic 1999 publication, jointly compiled by TGI and the International Telecommunication Union (ITU), was also consulted. To enable comparisons of countries’ international traffic statis- tics, TGI has endeavored to apply a consistent methodology. When reviewing the traffic statistics in TeleGeogruphy 2000, however, readers should keep in mind the following issues: Public Switched Network vs. Private Line Traffic Traffic volumes in TeleGeogruphy 2000 are generally reported in minutes. In most cases, the statistics refer to paid minutes of traffic on public switched circuits and thus include voice as well as fax traffic. Traffic carried by International Simple Resale (ISR) carriers is also included. ISR carriers resell the capacity of international private lines (IPLs) for switched services by interconnecting their IPLs to the public switched network at one or both ends. Traffic carried by "pure" resellers of international switched voice services is included in this report. These resellers do not own or lease their own international transmission facilities. Instead, they resell the traffic of other carriers; thus, pure resale traffic is counted as part of the minutes for the facilities-based carrier whose services are resold. Internet Traffic Statistics do not include traffic carried on the public Internet. Traffic volumes also exclude minutes sent over private lines using Internet Protocol (IP). For an overview of Voice over IP traffic volumes, see "VolP Traffic and Settlements." Cross-Border Traffic Neighboring countries may not classify local cross border traffic in the same way. That is, one country may treat some cross- border traffic as domestic, while its neighbor counts all such traffic as international. Billing Point vs. Originating Point of Traffic Unless otherwise stated in the notes to a table, the outbound minutes reported for countries in TeleGeography 2000 refer to outbound traffic originated in the reporting country even if it is billed in another country. In the past, most international calls were billed at the point of origination. The number of billed minutes thus coincided with the volume of outgoing traffic. Billed minutes also included col- lect or reverse charge calls because the calls were set up by an operator in the originating country. However, the recent use of calling card and call-back services has shifted the billing point for many international calls. For example, calls from Italy to the United States (or a third country, such as Argentina) may now be set up and billed in the U.S. Some countries, including the U.S., report international traffic data based solely on the location where the traffic is billed. Consequently, "outbound" traffic data for these countries can include traffic actually originating in another. Thus, incoming minutes reported for one country may not match the outgoing traffic on the same route by the correspondent country. Some double counting may also occur. For example, a call from Thailand to the U.S. which is billed to a U.S. calling card is reported by the U.S. carrier as outbound U.S. traffic; the same call also is reported as outbound minutes by Thailand. Accordingly, in countries where calling card and call-back ser- vices are widely used, a year-to-year comparison of national traffic also requires examining the statistics of countries, such as the U.S. and the U.K., where the calls are hubbed. Transit Traffic Unless otherwise stated, TeleGeography 2000 excludes refile traffic from the totals of countries acting as transit hubs. Notable exceptions include the U.K. and U.S. statistics, which do include some traffic reoriginated from other countries. Fixed vs. Mobile Traffic Traffic volumes include international calls originated and termi- nated on both fixed and mobile networks. Rounding Rounding may cause the figures on total national incoming and outgoing traffic to appear inconsistent with other national data. Revised Data Some differences exist between the historical statistics (1997 or eariier} reported in TeleGeography 2000 and data stated in prior TGI reports or Direction of Traffic. The variations reflect corrections and/or revised data subsequently provided to TGI. 271 TeleGeography 2000 © TeleGeograph~l, Inc. 1999 272 Global Reference: Blue Pages TeleGeography 2000 © TeleGeography, Inc. 1999 National Telecommunications Indicators (A-L) GDP 1998 Population Main Lines Lines Per Cellular Users International Internet Hosts Country (US$ billions) 1998 (millions) 1998 (thous.) 100 people 1998 (thous.) Carriers 1999 1999 (thous.) Albania (a, c) 3.0 3.1 116 3.7 6 1 n.a. Algeria (c) 48.4 30.1 1,600 5.3 19 1 n.a. Andorra n.a. 0.1 33 44.1 14 1 n.a. Angola 6.6 12.1 72 0.6 10 1 n.a. Argentina 344.4 36.1 7,323 20.3 2,822 2 114 Armenia 1.9 3.5 556 15.7 7 1 n.a. Aruba n.a. 0.1 33 35.3 8 1 n.a. Australia (b) 364.2 18.7 9,580 51.2 5,392 28 1,037 Austria (c) 212.1 8.1 3,999 49.1 2,030 17 229 Azerbaijan (a, c) 3.9 7.7 680 8.9 65 1 n.a. Bahamas n.a. 0.3 104 35.2 8 1 n.a. Bahrain (a) 5.3 0.6 158 24.6 92 1 n.a. Bangladesh 42.8 124.8 n.a. n.a. n.a. 1 n.a. Belarus (a, c) n.a. 10.3 2,490 24.1 12 1 n.a. Belgium (a) 247.1 10.1 5,073 50.0 1,748 18 302 Bolivia 8.6 8.0 535 6.7 218 1 n.a. Bosnia-Herzegovina (c) n.a. 3.7 332 9.0 25 3 n.a. Brazil (a) 778.3 165.9 19,987 12.0 7,761 2 302 Bulgaria 12.3 8.3 2,742 32.9 127 1 n.a. Canada (a) 598.8 30.3 19,241 63.5 5,320 49 2,346 Chile 78.0 14.8 2,977 20.0 956 10 54 China 960.9 1255.7 87,421 7.0 23,863 1 43 Costa Rica (a) 91.1 3.8 660 17.2 109 1 n.a. CSte d’lvoire 11.0 14.3 170 1.2 91 1 n.a. Cyprus (a, c) n.a. 0.7 405 58.5 116 1 n.a. Czech Republic (a) 56.4 10.3 3,741 36.4 965 1 108 Denmark (a) 15.5 5.3 3,496 66.0 1,931 18 317 Dominican Republic 56.4 8.2 764 9.3 256 3 n.a. Egypt 174.3 66.0 3,972 6.0 91 1 n.a. El Salvador 15.5 5.9 360 6.0 106 10 n.a. Estonia (a) 5.5 1.5 499 34.3 247 1 26 Fi n I a n d n.a. 5.2 2,855 55.4 2,933 8 634 France 1,432.9 59.7 34,000 57.0 11,210 50 778 Germany (a) 2,142.0 82.1 46,500 56.6 13,925 40 1,676 Ghana 7.5 19.2 106 0.6 44 1 n.a. Greece 120.3 10.6 5,536 52.2 2,057 1 70 Hong Kong (a, b) 166.6 6.7 3,729 55.8 3,174 80 208 Hungary (a) 47.8 10.2 3,423 33.6 1,073 1 116 India (a, b, c) 383.4 982.2 21,594 2.2 1,195 1 32 Indonesia (a) 96.3 206.3 5,572 2.7 1,066 2 20 Iran 100.3 65.8 7,355 11.2 390 1 n.a. Ireland (b) 80.9 3.7 1,600 43.5 946 25 52 Israel (a) 100.0 6.0 2,800 46.8 2,500 3 235 Italy (c) 1,171.0 57.7 25,986 45.1 20,489 15 534 Jamaica (a) 6.3 2.5 419 16.5 74 2 n.a. Japan (b) 3,783.1 126.2 60,381 47.8 47,308 50 2,373 Jordan (a) 7.4 6.0 403 6.7 70 1 n.a. Kazakhstan 21.0 17.1 1,818 10.6 30 3 n.a. Kenya 11.6 29.0 272 1.0 7 1 n.a. Korea, Rep. 297.9 46.4 20,089 43.3 13,982 24 318 Kuwait n.a. 1.8 412 22.7 250 1 n.a. Latvia (a) 6.4 2.5 741 30.2 167 1 n.a. Luxembourg 16.5 0.4 293 69.2 131 3 n.a. Note: Figures reflect year end totals except International Carriers (Sept. 1999) and Internet Hosts ISept. 1999) Source: TeleGeography, Inc., International Telecommunication Union, World Bank, and Telcordia Technologies © TeleGeography, Inc. 1999 274 © TeleGeography, Inc. 1999 TeleGeography 2000 International Telephone Traffic (A-L) Outgoing 1997 1998 40.8 49.1 n.a. 121.3 42.2 47.4 21.9 27.3 223.4 358.7 48.8 56.6 20.3 21.1 1,510.0 1,690.0 995.5 1,160.0 n.a. 42.9 62.7 83.5 106.6 124.4 46.9 41.8 148.6 176.1 1,340.0 1,480.0 22.7 31.6 66.2 94.9 459.1 545.8 76.0 96.0 4,286.3 4,805.0 242.0 259.4 1,631.8 1,711.5 66.9 82.7 40.1 57.3 154.4 182.0 306.1 317.4 607.5 710.0 142.0 157.5 119.3 127.3 34.3 43.1 66.3 75.1 371.1 410.8 3,545.0 4,115.0 4,813.0 5,495.0 21.9 28.9 593.7 681.3 1,718.0 1,879.8 287.1 296.3 420.5 436.2 351.6 324.5 160.7 177.0 695.0 885.0 459.0 661.0 2,351.9 2,640.0 n.a. 60.1 1,771.7 1,895.0 91.9 122.6 114.7 118.9 29.0 29.2 885.0 907.7 160.5 n.a. 44.0 55.4 282.9 293.8 Incoming MiTT (millions) MiTr (millions) Traffic Balance % Change 1997 1998 % Change 1997 1998 20.2% n.a. 52.2 n.a. n.a. 3.1 n.a. n.a. n.a. n.a. n.a. n.a. 12.3% 30.1 32.2 7.0% -12.1 ÷15.2 24.7% 18.8 22.3 18.3% -3.1 -5.0 60.6% 444.2 n.a. n.a. 220.8 n.a. 16.0% n.a. 94.3 n.a. n.a. 37.7 3.9% 25.8 31.0 20.3% 5.5 9.9 11.9% 1,250.0 n.a. n.a. -260.0 n.a. 16.5% 957.7 n.a. n.a. -37.8 n.a. n.a. n.a. 46.0 n.a. n.a. 3.2 1.3% n.a. 90.0 n.a. n.a. 26.5 16.7% 85.4 102.1 19.6% -21.2 -22.3 -11.0% 187.0 196.2 4.9% 140.1 154.4 18.5% 185.2 193.5 4.5% 36.6 17.3 9.0% 1,420.0 n.a. n.a. 80.0 n.a. 39.2% 69.3 76.4 10.3% 46.6 44.8 43.3% n.a. 159.2 n.a. n.a. 64.3 18.9% 761.3 806.9 6.0% 302.2 261.1 26.4% 135.0 201.0 48.9% 59.0 105.0 12.1% 4,635.1 n.a. n.a. 348.8 n.a. 7.2% n.a. n.a. n.a. n.a. n.a. 4.9% n.a. n.a. n.a. n.a. n.a. 23.6% 111.6 112.9 1.1% 44.7 30.2 42.9% 50.6 46.6 -7.9% t0.5 -10.7 17.9% 115.2 120.6 4.7 % -39.2 -61.4 3.7% 355.0 406.9 14.6% 48.9 89.5 16.9% 682.0 n.a. n.a. 74.5 n.a. 10.9% 476.9 730.5 53.2% 334.9 573.0 6.7% 451.2 475.3 5.3% 332.0 348.0 25.6% 168.2 149.2 -11.3% 133.9 106.1 13.3% 67.0 79.2 18.2% 0.7 4.1 10.7% n.a. n.a. n.a. n.a. n.a. t6.1% 3,609.0 n.a. n.a. 64.0 n.a. 14.2% 5,618.0 n.a. n.a. 805.0 n.a. 31.9% n.a. 100.8 n.a. n.a. 72.0 14.8% 634.6 710.1 11.9% 40.9 28.8 9.4% 2,100.3 1,833.0 -12.7% 382.3 -46.8 3.2% 324.6 374.5 15.4% 37.5 78.2 3.7% 1,256.6 1,498.8 19.3% 836.1 1,062.6 -7.7% 456.0 434.2 -4.8% 104.4 109.7 10.1% 130.2 185.7 42.7% -30.5 8.8 27.3% n.a. n.a. n.a. n.a. n.a. 44.0% 425.0 424.0 -0.2% -34.0 -237.0 12.2% 2,475.1 n.a. n.a. 123.2 n.a. n.a. n.a. 349.8 n.a. n.a. 289.7 7.0% 1,635.0 1,575.0 -3.7% -136.7 -320.0 33.4% 145.0 176.9 22.0% 53.1 54.4 3.7% n.a. 137.5 n.a. n.a. 18.6 0.6% 69.5 72.5 4.3% 40.5 43.3 2.6% 782.0 719.4 -8.0% -103.0 -t88.3 n.a. n.a. n.a. n.a. n.a. n.a. 25.8% 81.8 87.2 6.5% 37.8 31.8 3.9% 208.0 242.6 16.6% -74.9 -51.2 Country Albania (a, c) Atgeria (c) Andorra Angola Arqentina Armenia Aruba Australia (b) Austria (c) Azerbaijan (a, c) Bahamas Bahrain (a) Bangladesh Belarus (a, c) Belgium (a) Bolivia Bosnia-Herzegovina (c) Brazil (a) Bulgaria Canada (a) Chile China Costa Rica (a) COte d’lvoire Cyprus (a, c) Czech Republic (a) Denmark (a) Dominican Republic Egypt El Salvador Estonia (a) Finland France 6ermany (a) 6hana Greece Hong Kong (a, b) Hungary(a) India (a, b, c) Indonesia (a) Iran Ireland (b) Israel (a) Italy (c) Jamaica (a) Japan (b) Jordan (a) Kazakhstan Kenya Korea, Rep. Kuwait Latvia (a) Luxembourg Notes: Data are in millions of minutes of public switched traffic. a. International minutes based on billing point of traffic. b. International traffic foryear ending 31 March. Australia and Pakistan ends 30 June. c. Traffic data exclude some carriers or routes. (See country table for details.) d. 1997 and 1998 traffic data not directly comparable. (See country table for details.) © TeleGeography, Inc. 1999 275 TeleGeography 2000 © TeleGeography, Inc. 1999 National Telecommunications Indicators (M-Z) GDP 1998 Population Main Lines Lines Per Cellular Users International Internet Hosts Country (US$ billions) 1998 (millions) 1998 (thous.) 100 people 1998 (thous.) Carriers 1999 1999 (thous.) Macau (a) n.a. 0.4 174 40.9 65 1 n.a. Macedonia (c, d) 2.2 2.0 408 20.4 31 1 n.a. Malaysia (a, b, d) 71.3 22.2 4,384 19.8 2,200 5 59 Malta 3.7 0.4 192 49.9 23 1 n.a. Mauritius (b) 4.2 1.1 245 21.4 60 1 n.a. Mexico (a) 393.2 95.8 9,927 10.4 3,302 16 200 Moldova (a) 1.6 4.4 657 15.0 9 1 n.a. Monaco n.a. n.a. n.a. n.a. n.a. 2 n.a. Morocco 35.1 27.9 1,515 5.4 116 1 n.a. Namibia 3.1 1.7 101 6.1 20 1 n.a. Netherlands 382.5 15.7 9,337 59.3 3,351 30 817 Netherlands Antilles n.a. 0.2 n.a. n.a. n.a. 1 n.a. New Zealand (b) 54.1 3.9 1,868 47.9 790 19 241 Nicaragua (a) 2.1 4.4 128 2.9 20 1 n.a. Norway (a) 145.9 4.4 2,935 66.0 2,107 14 391 Oman (a) 15.0 2.4 220 9.2 103 1 n.a. Pakistan (b, c) 63.9 142.3 2,557 1.8 110 1 n.a. Palestinian Authority n.a. n.a. n.a. n.a. n.a. 1 n.a. Panama (a) 9.2 2.8 104 n.a. n.a. I n.a. Paraguay 8.8 5.2 289 5.5 215 1 n.a. Peru (a, b) 64.1 24.8 1,655 6.7 743 18 n.a. Philippines (a, b) 65.1 72.9 2,700 3.7 1,595 12 n.a. Poland (a) 148.9 38.7 8,812 22.8 1,928 1 155 Por[ugal (a) 106.7 10.0 4,117 41.4 3,075 1 65 Qatar n.a. 0.6 151 26.0 66 1 n.a. Romania 38.2 22.6 3,777 16.8 643 1 17 Russia (c) 276.6 147.7 26,875 18.2 800 35 197 Saudi Arabia 125.8 20.2 2,878 14.3 627 1 n.a. Singapore (a, b) 85.4 3.2 1,778 56.2 1,095 1 109 SIovak Republic (a) 20.4 5.4 1,539 28.6 465 1 20 SIovenia 18.2 2.0 722 36.3 196 1 22 South Africa 116.7 44.3 5,075 11.5 2,500 1 207 Spain (a) 551.9 39.4 16,289 41.4 7,051 16 382 Sri Lanka 15.1 18.5 524 2.8 174 1 n.a. Sudan (c) n.a. 28.3 162 0.6 9 1 n.a. Sweden (a) 225.0 8.9 5,965 67.4 4,109 16 615 Switzerland 264.4 7.1 4,803 67.5 1,672 40 315 Syria (a) 16.3 15.3 1,530 10.0 n.a. 1 n.a. Taiwan (a) n.a. 21.9 11,500 52.5 4,727 1 842 Thailand (a, c) 111.3 60.3 4,827 8.0 1,957 2 34 Trinidad & Tobago (a, b) 5.8 1.3 266 20.7 24 1 n.a. Turkey (c) 198.8 66.7 16,960 25.4 3,506 1 79 Ukraine 49.7 50.9 9,410 18.5 143 2 32 United Arab Emirates 47.2 2.4 915 38.9 493 1 n.a. United Kin~Jdom (a, b) 1,357.4 58.9 32,072 54.4 14,874 215 2,073 United States (a) 8,210.6 270.5 172,452 63.8 69,209 679 44,230 Uruguay 20.2 3.3 824 25.0 196 1 14 Uzbekistan 14.2 23.8 1,490 6.2 25 1 n.a. Venezuela (a) 105.8 23.2 2,712 11.6 2,015 1 n.a. Vietnam 25.9 77.6 2,000 2.6 187 1 n.a. Yugoslavia (a) n.a. 10.6 2,319 21.8 240 2 n.a. Zimbabwe 6.5 12.7 212 1.7 15 1 n.a. Note: Figures reflect year end totals except International Carriers (Sept. 1999) and Internet Hosts (Sept. 1999) Source: TeleGeography, Inc., International Telecommunication Union, World Bank, and Telcordia Technologies © TeleGeography, Inc. 1999 276 © TeleGeography, Inc. 1999 TeleGeography 2000 International Telephone Traffic (M-Z) 1997 119.0 51.7 588.5 34.4 24.6 1,213.6 55.6 149.9 49.7 1,615.0 407.0 40.4 481.0 74.3 84.1 41.4 26.1 79.4 295.0 529.0 393.3 99.9 110.8 969.6 801.3 1,161,0 144.7 113.5 368.8 1,319.0 33.2 14.8 1,140.0 2,164.0 89.3 789.0 278.4 62.0 557.5 486.8 738.0 6,800.0 23,001.7 68.4 63.1 159.2 50~2 217.0 n.a. Outgoing MiTF (millions) Incoming MiT]" (millions) Traffic Balance 1998 % Change 1997 1998 % Change 1997 1998 125.2 5.2% 92.2 95~8 3.9% -26.8 -29.4 37.1 n.a. 85.0 91.7 n.a. 33.3 54.6 685.0 n.a. 592.0 n.a. n.a. 3.5 n.a. 37.3 8.4% 37.0 n.a. n.a. 2.6 n.a. 29.7 20.8% 35.0 38.5 10.0% 10.4 8.8 1,310.0 7.9% 2,819.3 3,060.0 8.5% 1,605.7 1,750.0 55.8 0.4% 80.2 90.3 12.6% 24.6 34.4 234.0 n.a. n.a. 140.0 n.a. n.a. -94.0 158.0 5.4% 364.0 460.0 26.4% 214.1 302.0 61.9 24.5% 42.3 45.3 7.1% -7.4 -16.6 1,805.0 11.8% n.a. n.a. n.a. n.a. n.a. 43.0 n.a. n.a. 113.8 n.a. n.a. 70.9 580.0 42.5% 430.0 n.a. n.a. 23.0 n.a. n.a. n.a. 52.5 n.a. n.a. 12.1 n.a. 540.0 12.3% 515.0 n.a. n.a. 34.0 n.a. 92.9 25.1% 70.4 74.2 5.4% -3.9 -18.7 87.5 4.0% 557.8 640.4 14.8% 473.7 552.9 27.6 n.a. n.a. 16.6 n.a. n.a. -11.0 50.0 20.8% 95.1 97.6 2.7% 53.7 47.6 37.8 44.8% n.a. 57.4 n.a. n.a. 19.6 90.4 13.8 % 256.9 272.6 6.1% 177.5 182.3 295.0 0.0% 930.0 n.a. n.a. 635.0 n.a. 602.4 13.9% 800.2 1,144.2 43.0% 271.2 541.8 460.2 17.0% 628.8 711.5 13.1% 235.5 251.3 112.5 12.6% 59.5 70.0 17.8% -40.4 -42.5 n.a. n.a. 278.6 n.a. n.a. 167.8 n.a. 1,038.3 7.1% 1,035.6 1,029.8 -0.6% 66.0 -8.5 932.6 16.4% n.a. n.a. n.a. n.a. n.a. 1,235.0 6.4% n.a. n.a. n.a. n.a. n.a. 151.8 4.9% 174.4 186.4 6.9% 29.7 34.6 129.6 14.2% 118.9 137.0 15.2% 5.4 7.4 405.0 9.8% 343.2 n.a. n.a. -25.6 n.a. 1,605.0 21.7% n.a. n.a. n.a. n.a. n.a. 39.3 18.4% 124.3 146.8 18.1% 91.1 107.5 18.4 24.5% 43.2 88.0 103.7% 28.4 69.6 1,205.0 5.7% n.a. n.a. n.a. n.a. n.a. 2,425.0 12.1% 1,723.0 n.a. n.a. -441.0 n.a. 103.0 15.3% 173.2 n.a. n.a. 83.9 n.a. 862.0 9.3% 842.2 781.8 -7.2% 53.2 -80.2 296.4 6.5% 408.5 358.6 -12.2% 130.1 62.2 64.4 3.9% 130.0 141.5 8.9% 68.0 77.1 644.1 15.5% 836.0 955.9 14.3% 278.5 311.7 465.9 -4.3% n.a. n.a. n.a. n.a. n.a. 874.8 18.5% n.a. n.a. n.a. n.a. n.a. 8,225.0 21.0% n.a. 6,400.0 n.a. n.a. -1,825.0 24,949.3 8.5% 9,213.3 10,598.8 15.0% -13,788.4 -14,350.5 78.3 14.5% 93.7 97.0 3.5% 25.3 18.7 91.7 45.4% n.a. 74.7 n.a. n.a. -17.0 164.5 3.3% 286.9 298.1 3.9% 127.7 133.6 56.0 11.6% 294.6 334.0 13.4% 244.4 278.0 219.5 1.2% 332.0 423.3 27.5% 115.0 203.8 52.8 n.a. n.a. 53.2 n.a. n.a. 0.4 Country Macau (a) Macedonia (c, d) Malaysia (a, b, d) Malta Mauritius (b) Mexico (a) Moldova (a) Monaco Morocco Namibia Netherlands Netherlands Antilles New Zealand (b) Nicaragua (a) Norway (a) Oman (a) Pakistan (b, c) Palestinian Authority Panama (a) Para~juay Peru (a, b) Philippines (a, b) Poland (a) Portugal (a) Qatar Romania Russia (c) Saudi Arabia Singapore (a, b) Slovak Republic (a) Slovenia South Africa Spain(a) Sri Lanka Sudan(c) Sweden (a) Switzerland Syria (a) Taiwan (a) Thailand (a, c) Trinidad & Tobago (a, b) Turkey (c) Ukraine United Arab Emirates United Kingdom (a, b) United States (a) Uruguay Uzbekistan Venezuela (a) Vietnam Yugoslavia (a) Zimbabwe Notes: a. International Mi’l-r based on billing point of traffic. b. International traffic for year ending 31 March (30 June for Australia, Pakistan) c. Traffic data exclude some carriers or routes. (See country table for details.) d. 1997 and 1998 traffic data are not directly comparable. (See country table for details.) © TeleGeography, Inc. 1999 277 © TeleGeography, Inc. 1999 TeleGeography 2000 Internati 1 Canada Guam Northern Marianas United States Caribbean 20 Egypt 212 Morocco 213 Algeria 216 Tunisia 218 Libya 220 Gambia 221 Senegal 222 Mauritania 223 Mall 224 Guinea 225 Ivory Coast 226 Burkina Faso 22"/ Niger 228 Togo 229 Benin 230 Mauritius 231 Liberia 232 Sierra Leone 23:3 Ghana 234 Nigeria 235 Chad 236 Central African Republic 23? Cameroon 238 Cape Verde Islands 239 SaoTome and Principe 240 Equatorial Guinea 241 Gabon 242 Congo (Brazzaville) 243 Congo (Kinshasa) 244 Angola 245 Guinea-Bissau 246 Diego Garcia 247 Ascension Island 246 Seychelles 249 Sudan 250 Rwanda 251 Ethiopia 252 Somalia 253 Djibouti 254 Kenya 255 Tanzania 256 Uganda 257 Burundi 258 Mozambique 260 Zambia 261 Madagascar 262 Reunion Island 263 Zimbabwe 264 Namibia 265 Malawi onal Dialing 266 Lesotho 267 Botswana 268 Swaziland 269 Comoros & Mayotte 27 South Africa 290 St. Helena 291 Eritrea 297 Aruba 298 Farce Islands 299 Greenland 30 Greece 31 Netherlands 32 Belgium 33 France 34 Spain 350 Gibraltar 351 Portugal; Azores 352 Luxembourg 353 Ireland 354 Iceland 355 Albania 356 Malta 357 Cyprus 368 Finland 359 Bulgaria 36 Hungary 370 Lithuania 371 Latvia 372 Estonia 373 Moldova 374 Armenia 375 Belarus 376 Andorra 377 Monaco 378 San Marino 379 Vatican City 380 Ukraine 381 Yugoslavia 385 Croatia 386 Slovenia 387 Bosnia-Hercegovina 389 Macedonia 39 Italy 46 Romania 41 Switzerland 41-75 Liechtenstein 420 Czech Republic 421 Slovak Republic 43 Austria 44 United Kingdom 45 Denmark 46 Sweden 47 Norway 46 Poland 49 Germany Codes, by 500 Falkland Islands 501 Belize 502 Guatemala 503 El Salvador 504 Honduras 505 Nicaragua 506 Costa Rica 507 Panama 508 St. Pierre & Miquelon 509 Haiti 51 Peru 52 Mexico 53 Cuba 54 Argentina 55 Brazil 50 Chile 57 Colombia 58 Venezuela 550 Guadeloupe 501 Bolivia 592 Guyana 593 Ecuador 594 French Guiana 595 Paraguay 596 Martinique 597 Suriname 598 Uruguay 599 Netherlands Antilles 60 Malaysia 61 Australia 62 Indonesia 63 Philippines 64 New Zealand 65 Singapore 66 Thailand 672 Australian Territories 673 Brunei 674 Nauru 675 Papua New Guinea 676 Tonga Islands 677 Solomon Islands 678 Vanuatu 6"/9 Fiji 680 Palau 681 Wallis & Futuna 682 Cook Islands 683 Niue 684 American Samoa 685 Western Samoa 686 Kiribati 687 New Caledonia 688 Tuvalu 689 French Polynesia 690 Tokelau 691 Micronesia Number 692 Marshall Islands 7 Kazakhstan Russia 800 International Freephone 81 Japan 82 South Korea 84 Vietnam 850 North Korea 852 Hong Kong 853 Macau 855 Cambodia 856 Laos 86 China 870 Inmarsat Special 871 Inmarsat East Atlantic 872 Inmarsat Pacific 87:3 Inmarsat Indian 874 Inmarsat West Atlantic 880 Bangladesh 881x Global Mobile Satellite Systems 886 Taiwan 90 Turkey 91 India 92 Pakistan 9:3 Afghanistan 94 Sri Lanka 95 Myanmar 960 Maldives 961 Lebanon 962 Jordan 963 Syria 964 Iraq 965 Kuwait 966 Saudi Arabia 967 Yemen 968 Oman 971 United Arab Emirates 972 Israel 973 Bahrain 974 Qatar 975 Bhutan 976 Mongolia 977 Nepal 98 Iran 992 Tajikistan 993 Turkmenistan 994 Azerbaijan 995 Georgia 996 Kyrgyzstan 998 Uzbekistan 279 TeleGeography 2000 © TeleGeography, Inc. 1999 International Dialing Codes, by Country Afghanistan ................. 93 Albania .................... 355 Tirana ................. 42 Algoria ................... 213 Algiers ................. 2 American Samoa ............ 684 Andorra ................... 376 Angola ................... 244 Luanda ................. 2 Anguilla ................. 1-254 Antigua & Barbuda ........ 1-268 Argentina ................... 54 Buenos Aires ............. 1 Armenia ................... 374 Yerevan ................. 2 Aruba .................... 297 Ascension Island .......... 247 Anstralia ................... 61 Melbourne ............... 3 Sydney .................. 2 Australian Territories ........ 672 Austria ..................... 43 Vienna .................. 1 Azerbaijan ................. 994 Baku .................. 12 Bahamas ................. 1-242 Bahrain .................... 973 Bangladesh ................ 880 Dhaka .................. 2 Barbados ................ 1-246 Belarus .................... 375 Minsk ................ 172 Belgium .................... 32 Brussels ................ 2 Belize .................... 501 Belmopan ............... 8 Benin .................... 229 Bermuda ................. 1-441 Bhntaa .................... 976 Bolivia .................... 6gl La Paz .................. 2 Bnsnia .................... 387 Sarajevo ................ 71 Botswana ................. 267 Brazil ...................... 56 Brasilia ................ 61 Rio de Janeiro ........... 21 S~o Paulo ............... 11 British Virgin Islands ...... 1-284 Brunei ..................... 673 Bandar Seri Begawan ....... 2 Bulgaria ................... 359 Sofia ................... 2 Burkina Faso ............... 226 Burundi ................... 257 Cambodia .................. 855 Cameroon ................. 237 Canada ...................... 1 Montreal ........... 514/450 Ottawa ................ 613 Toronto ................ 416 Vancouver .............. 604 Cape Verde ............... 238 Cayman Islands ........... 1-345 Central African Republic .... ,236 Bangui ................. 61 Chad ..................... 235 Chile ....................... 56 Santiago ................ 2 China, People’s Republic of .... 86 Beijing ................. 10 Guangzhou .............. 20 Shanghai ............... 21 Columhia ................... -~7 Bogota ................. 1 Cocos Islands; Norfolk & Christmas Islands .......... 672 Comoros .................. 269 Congo .................... 242 Brazzaviile ......... 81/82/83 Congo .................... 243 Kinshasa ............... 12 Costa Rica ................. 506 Craatia .................... 386 Zagreb .................. 1 Cnba ....................... 63 Havana ................. 7 Cyprus ..................... 367 Nicosia ................. 2 Czech Republic ............. 420 Prague .................. 2 Denmark .................... 45 Diego Garcia .............. 246 Djibouti ................... 253 Dominica ................ 1-767 Dominican Republic ....... 1-809 Ecuador ................... 593 Quito ................... 2 Egypt ....................... 20 Cairo ................... 2 El Salvador ................. 503 Equatorial Guinea .......... 240 Eritrea ..................... 291 Estonia .................... 372 Tallinn .................. 2 £thiepia ................... 261 Addis Ababa ............. 1 Falkland Islands ............ 500 Faroe Islands ............... 298 Fiji ...................... 679 Finland .................... 358 Helsinki ................. 9 France ..................... 33 Paris ................... 1 Marseille ................ 4 French Antilles ............. 598 French Guiana .............. 594 French Polynesia ........... 689 Gabon ..................... 241 Gambia .................... 220 Georgia .................... 995 Tbilisi ................. 32 Germany .................... 49 Berlin .................. 30 Bonn ................. 228 Frankfurt ................ 69 Munich ................. 89 Ghana ..................... 233 Accra .................. 21 Gibraltar ................... 350 Global Mobile Satellite Systems G iobalstar ........ 8818/8819 ICO Global Comm..8810/8811 Iridium .......... 8816/8817 Greece ..................... 30 Athens ................. 1 Greenland ................. 299 Grenada ................. 1-473 Guadeloupe ................ 598 Guam .................... 1-671 Guatemala ................. 502 Guinea .................... 224 Guinea-Bissau .............. 245 Guyana .................... 592 Georgetown ............. 2 Haiti ...................... 509 Honduras .................. 504 Hong Kong ................. 852 Hungary .................... 36 Budapest ................ 1 Iceland .................... 354 India ....................... 91 Mumbai ................ 22 Calcutta ................ 33 New Delhi .............. 11 Indonesia .................. 6~ Jakarta ................ 21 Inmarsat Special ................ 870 East Atlantic ............ 871 Pacific ................ 872 indian ................. 873 West Atlantic ........... 874 International Freephone ...... 800 Iran ....................... 98 Tehran ................. 21 Iraq ...................... 964 Baghdad ................ 1 Ireland .................... 353 Bublin .................. 1 Israel ..................... 972 Jerusalem ............... 2 Tel Aviv ................. 3 Italy ....................... 39 Rome .................. 06 Milan .................. 02 Ivory Coast ................. 225 Jamaica ................. 1-876 Japan ...................... 81 Osaka ................... 6 Tokyo ................... 3 Jordan .................... 962 Amman ................. 6 Kazakhstan .................. 7 Almaty ............... 3272 Kenya ..................... 254 Nairobi ................. 2 Kiribati .................... 686 Kuwait .................... 965 Kyrgyzstan ................. 996 Bishkek .............. 3312 Lads ...................... 856 Latvia ..................... 371 Riga ................... 2 Lebanon ................... 961 Beirut .................. 1 Lesotho .................... 266 Liberia .................... 231 Libya ...................... 218 Tripoli ................. 21 Liechtenstein ............. 41-75 Lithuania .................. 370 Vilnius .................. 2 280 © TeleGeography, Inc. 1999 TeleGeography 2000 Luxembourg ................ 352 Macau .................... 853 Macedonia ................. 389 Skopje ................. 91 Madagascar ............... 3-61 Malawi ................... 3-65 Malaysia ................... 60 Kuala Lumpur ............ 3 Maldives .................. 960 Mali ..................... 3.23 Malta ..................... 356 Marshall Islands ............ 692 Martinique ................. 596 Mauritania ................ ~.2 Mauritius ................. 3.30 Mayotte ................... 269 Mexico ..................... 52 Guadalajara ............. 36 Mexico City .............. 5 Monterrey .............. 83 Mierano$ia ................ 691 Moldnva ................... :37:3 Chisinau ................ 41 Monaco ................... 377 Mongolia .................. 976 Ulaanbaatar ............. 1 Mant~errat ............... 1-664 Morocco ................... 212 Casablanca ............... 2 Rabat .................. 7 Mozambique .............. 3-58 Maputo ................. 1 Myaumar ................... 95 Namihia .................. 264 Windhoek .............. 61 Nanru ..................... 674 Nepal ..................... 977 Kathmandu ............... 1 Netherlands ................. 31 Amsterdam ............. 20 Netherlands Antilles ........ 599 New Caledonia ............. 687 New Zealand ................ 64 Auckland ................ 9 Wellington .............. 4 Nicaragua ................. 505 Managua ................ 2 Niger ..................... 3.27 Nigeria ................... 234 Lagos ................... 1 Niue ...................... 683 North Korea ................ 850 Pyongyang ............... 2 Northern Marianas .......... 670 Saipan ................ 322 Norway ..................... 47 Oslo ................... 22 Oman ..................... 968 Pakistan .................... 92 Islamabad .............. 51 Palau ..................... 680 Panama ................... 507 Papua New Guinea .......... 675 Paraguay .................. 595 Asuncion ............... 21 Peru ....................... 51 Lima .................. 14 Philippines .................. 63 Manila ................. 2 Poland ..................... 48 Warsaw ............... 22 Portugal ................... 351 Lisbon .................. 2 Puerto Rico ............... 1-787 Qatar ...................... 974 Reunion Island ............ 3-62 Romania .................... 40 Bucharest ............... 1 Russia ....................... 7 Moscow .............. 095 St. Petersburg ........... 812 Rwanda .................. 258 St. Helena ................. 3-90 St. Kitts & Nevis ........... 1-869 St. Lucia ................. 1-758 St. Pierre & Miquelon ........ 508 St. Vincent & the Grenadines .............. 1-809 San Marino ................ 378 S~o Tome and Principe ..... 2219 Saudi Arabia ............... 966 Riyadh .................. 1 Senegal .................. 3.21 Seychelles ................. 248 Sierra Leone .............. 3-32 Freetown ............... 22 Singapore ................... 65 Slovak Republic ............ 421 Bratislava ............... 7 Slovenia ................... 386 Ljubljana ............... 61 Solomon Islands ............ 677 Somalia ................... 252 Mogadishu .............. 1 South Africa ................. 27 Johannesburg ............ 11 Pretoria ................ 12 South Korea ................. 92 Seoul ................... 2 Spain ...................... 34 Madrid ................ 91 Barcelona ............... 93 Sri Lanka ................... 94 Colombo ................ 1 Sudan ..................... 249 Khartoum .............. 11 Suriname .................. 597 Swaziland ................. 258 Sweden .................... 46 Stockholm ............... 8 Switzerland ................. 41 Berne .................. 31 Zurich ................... 1 Syria ...................... 963 Damascus .............. 11 Tahiti ...................... 689 Taiwan .................... 886 Taipei .................. 2 Tajikistan .................. 992 Dushanbe ............ 3772 Tanzania ................... 255 Bar Es Salaam ........... 51 Thailand .................... 66 Bangkok ................ 2 Togo ...................... 228 Tokelau .................... 690 Tonga ..................... 676 Trinidad & Tobago ......... 1-858 Tunisia .................... 216 Tunis ................... 1 Turkey ...................... 90 Ankara ................ 312 Istanbul ............... 212 Turkmenistan ............... 993 Ashkhabad ............. 12 Turks & Caicos ........... 1-649 Tuvalu ..................... 688 Uganda .................... 256 Kampala ............... 41 Ukraine .................... 388 Kiev ................... 44 United Arab Emirates ........ 971 Abu Dhabi ............... 2 Dubai ................... 4 Hnited Kingdom .............. 44 Cardiff ............... 1222 Glasgow .............. 141 London ............ 171/181 Manchester ............ 161 tlnited States ................. 1 Chicago ............ 312/773 Houston ............ 713/281 Los Angeles ......... 213/323 Miami ............. 305/786 New York .......... 212/917 Washington ............ 202 U.S. Virgin Islands ......... 1-340 Uruguay ................... 598 Montevideo .............. 2 Uzbekistan ................. 998 Tashkent ............. 3712 Vanuatu ................... 678 Vatican City ................ 379 Venezuela .................. 58 Caracas ................. 2 Vietnam .................... 84 Wallis & Futuna ............ 581 Western Samoa ............. 585 Yemen ..................... 967 Sanaa .................. 1 ¥ngoslavia ................. :381 Belgrade ................ 11 Zambia .................... 268 Lusaka .................. 1 Zanzibar (Tanzania) .......... 255 Zimbabwe ................. 263 Harare .................. 4 281 TeleGeol~raphy 2000 World Dialing Codes © TeleGeographg, Inc. 1999 Oslo GERMANY 122) :. UKRAINE MOLDOVA COLOMBIA ¯ Bmsilia -~.’COAST’ 282 © TeleGeography, Inc. 1999 TeleGeography 2000 © TeleGeography, Inc. 1999 RUSSIA KAZAKHSTAN 7 ~. MONGOLIA umanoaatar NEPAL BHUTAN ~aba ~0~L~ .............. AUSTRALIA Svdnev (2), Canberra, ,: ::::i:.:: ;;:~ i :: 283 TeleGeography 2000 © TeleGeography, Inc. 1999 North American Area Nabama Birmingham ............... 205 Huntsville ................. 256 Montgomery ............... 334 Alaska ........................ 907 Alberta Calgary ................. .403 Edmonton ................. 780 Anguilla ...................... 264 Antigua & Barlinda ............ 268 Arizona Buckeye .................. 623 Phoenix ................. .602 Tempe .................. .480 Tucson ................... 520 Arkansas Jonesboro ............... .870 Little Rock ................ 501 Bahamas ...................... 242 Barbados ..................... 246 Bermuda ...................... 441 British Columbia Victoria .................. .250 Vancouver ................. 604 British Virgin Islands ........... 284 California Anaheim .................. 714 Bakersfield ................ 805 Burbank .................. 626 Concord ................. 925 Fresno ................... 209 Irvine .................... 949 Long Beach ................ 562 Los Angeles ........... 213/323 Monterey ................. 831 Oakland .............. .344/510 Palm Springs .............. 760 Redding .................. 530 Riverside ................. 909 Sacramento .............. .916 Pale Alto ................ .650 San Diego ................. 619 San Femando ............. .818 San Francisco ......... .415/628 San Jose ............. .669/408 Santa Ana ............ .657/714 Santa Monica ......... .424/310 Santa Rosa ................ 707 Visalia ................... 559 Cnyman Islands ................ 345 Colorado Colorado Springs ........... 719 Denver .............. 303/720 Ft. Collins ................. 970 Connecticut Bridgeport ................ 203 Hartford ................. .860 Qulawaro .................... 302 District of Columbia Washington ............... 202 Bominina .................... 267 Bomininnn Repnblia ........... .809 Florida Ft. Lauderdale .............. 954 Ft. Myers ................. 941 Gainesville ................ 352 Jacksonville ............... 904 Miami ................ 305/786 Orlando .............. 321/407 Tallahassee ............... 850 Tampa ............... .813/727 W. Palm Beach ............. 561 Georgia Athens ................... 706 Atlanta .............. A04/678 Madetta .............. 770/678 Savannah ................. 912 Grenada ...................... 473 Guam ........................ 671 Hawaii ....................... 808 Idaho ....................... 208 Illinois Aurora ................ 630/331 Cairo .................... 618 Chicago ........... 312/773/872 Evanston .............. 847/224 Oak Brook ............ .708/464 Peoria .................... 309 Rockford .................. 815 Springfield ................ 217 Indiana Evansville ................. 812 Gary ..................... 219 Indianapolis .............. 317 Lafayette ................. 765 Iowa Council Bluffs .............. 712 Des Moines ............... 515 Dubuque .................. 319 Jamaica ...................... 876 Kansas Kansas City ............... 913 Topeka ................... 785 Wichita .................. 316 Kentucky Dade Park ................. 270 Lexington ................. 606 Louisville ................. 502 Louisiana Baton Rouge ............... 225 New Odeans .............. 504 Shreveport ............... .318 Maine ........................ 207 Manitoba ..................... 204 Maryland Baltimore ............ .410/443 Rockville ............. .301/240 Massachusetts Boston .................. .617 Cambridge ................ 781 Lowell ................... 978 New Bedford .............. 508 Springfield ................ 413 Michigan Ann Arbor ................. 734 Detroit ................... 313 Flint ................. .586/810 Grand Rapids .............. 616 Lansing ................... 517 Pontiac .................. 248 Sault Ste. Made ............ 906 Minnesota Duluth .................. 218 Minneapolis .............. .812 Rochester ................. 507 St. Cloud ................. 320 St. Paul .................. .651 Biloxi .................... 228 Jackson .................. 601 Codes, by Jurisdiction Missouri Jefferson City ............. .573 Kansas City ............... 816 St. Joseph ............... .660 St Louis ................. 314 Springfield ........ ...... .417 Montserrnt ................... .664 Montana ..................... .406 Nebraska North Platte .............. .308 Omaha .................. .402 Nevada Las Vegas ................ 702 Reno ................... .775 New Brunswick ............... .500 New Hampshire ................ 003 New Jersey Elizabeth ................. .908 Jersey City ................ 201 Newark ................. .973 New Brunswick ........... 232 Trenton .................. .609 New Mexico .................. 505 New York Albany .................. .518 Queens .............. .347/718 Buffalo .................. .716 Long Island ............... .516 Ithaca ................... .607 Manhattan ........ 212/646/917 Syracuse ................. .315 White Plains ............... 914 Newfoundland ................. 709 North Carolina Asheville ................ .828 Chadotte ................. 704 Fayetteville ............... .910 6reensboro ............... .336 Raleigh .................. .919 Rocky Mountain ........... 252 North Qukota ................. .701 Northern Marianas ............ .670 NW Territories/Yukon .......... .867 Nova Scotia & Prince Edward Island ........... .902 Ohio Canton .................. .330 Cincinnati ................. 513 Cleveland ................ 216 Columbus ................. 614 Dayton .................. .937 Madetta ................. .740 0bedin .................. .440 Toledo ...... ........... .419 Enid .................... .580 Oklahoma City ............ .405 Tulsa .................... .918 Ontario Hamilton .................. 906 London ................... 519 North Bay ................ .705 Ottawa ................... 613 Thunder Bay ............... 807 Toronto .............. .416/647 Oregon Eugene .................. .541 Portland ............... 503/971 Allentown ............ 484/610 Altoona .................. 814 Harrisburg ................ 717 New Castle ............... 724 Philadelphia .......... 215/267 Pittsburgh ................ .412 Pnerto Rico .................... 787 Quebec Montreal ................. 514 Quebec ................... 418 Sherbrooke ................ 819 Trois Rivi~res ............. .450 Rhode Island ................. .401 SL Kitts & Nevis ................ 869 St. Lucia ...................... 788 St. Vincent/Grenadines ......... .809 Saskatchewan ................. 306 South Carolina Charleston ............... .843 Columbia ................. 803 Greenville ................. 864 South Dakota ................. .605 Tennessee Columbia ................. 931 Memphis ................. 901 Nashville ................. 615 Knoxville ................. .423 Texas Amarillo ................. .806 Austin .................... 512 Brownsville ............... 956 Dallas ............... .469/214 Deer Park ............ .832/281 Del Rio ................... 830 El Paso ................... 915 Fort Worth ............... .817 Galveston ................. 409 Houston .......... .713/281/832 Irving ................. 469/972 San Antonio .............. .210 Tyler .................... .903 Waco .................... 254 Wichita Falls ............. .940 Trinidad & Tobago ............. .868 Turks & Coicos ................. 649 U.S. Virgin Islands ............. .340 Utah Salt Lake City .............. 801 Provo ................... .435 Vermont ....................... 802 Virginia Alexandria ................ 703 Richmond ................. 804 Roanoke .................. 540 Norfolk ................... 757 Washington Bellevue ................. .425 Olympia ................. .360 Seattle ................... 206 Spokane .................. 509 Tacoma ................... 253 West Virginia ................. .304 Wisconsin Green Bay ................. 920 Madison ................. .608 Milwaukee ................ 414 Eau Claire ................. 715 Wyoming ...................... 307 Note: Two codes separated by a slash (e.g., in Houston, Texas)indicate an overlay; multiple codes are used for the same geographic area. ..............284 © TeleGeography, Inc. 1999 TeleGeography 2000 North American Area Codes, by Z01 New Jersey 407 Florida 618 Illinois 202 District of Columbia 408 California 619 California ZO3 Connecticut 409 Texas 623 Arizona 204 Manitoba 410 Maryland 626 California 205 Alabama 411 Directory Assistance 628 California 206 Washington 412 Pennsylvania 630 Illinois 207 Maine 413 Massachusetts 646 New York 208 Idaho 414 Wisconsin 647 Ontario 209 California 415 California 649 Turks & Caicos 210 Texas 416 Ontario 650 California 212 NewYork 417 Missouri 651 Minnesota 213 California 418 Quebec 657 California 214 Texas 419 Ohio 660 Missouri 215 Pennsylvania 423 Tennessee 661 California 216 Ohio 424 California 664 Montserrat 217 Illinois 425 Washington 669 California 218 Minnesota 435 Utah 670 Nor~hernMarianas 219 Indiana 440 Ohio 671 Guam 224 Illinois 441 Bermuda 678 Georgia 225 Louisiana 443 Maryland 701 North Dakota 228 Mississippi 450 Quebec 702 Nevada 240 Maryland 464 Illinois 703 Virginia 242 Bahamas 469 Texas 764 North Carolina 246 Barbados 473 Grenada 705 Ontario 240 Michigan 460 Arizona 706 Georgia 250 British Columbia 484 Pennsylvania 707 California 262 North Carolina 500 Personal Comm.Serv.(PCS) 708 Illinois 253 Washington 501 Arkansas 705 Newfoundland 264 Texas 502 Kentucky 710 U.S. Government Emergency 256 Alabama 503 Oregon 712 Iowa 264 Anguilla 504 Louisiana 713 Texas 267 Pennsylvania 505 New Mexico 714 California 208 Antigua & Barbuda 506 New Brunswick 715 Wisconsin 270 Kentucky 507 Minnesota 716 NewYork 281 Texas 508 Massachusetts 717 Pennsylvania 264 British Virgin Islands 509 Washington 718 New York 301 Maryland 510 California 719 Colorado 302 Delaware 512 Texas 720 Colorado 303 Colorado 513 Ohio 724 Pennsylvania 304 West Virginia 514 Quebec 727 Florida 305 Florida 515 Iowa 732 New Jersey 306 Saskatchewan 516 NewYork 734 Michigan :307 Wyoming 517 Michigan 740 Ohio 308 Nebraska 518 NewYork 757 Virginia 309 Illinois 519 Ontario 750 St. Lucia 310 California 520 Arizona 760 California 312 Illinois 530 California 705 Indiana 313 Michigan 540 Virginia 767 Dominica 314 Missouri 541 Oregon 770 Georgia 315 NewYork 559 California 773 Illinois 316 Kansas 501 Florida 775 Nevada 317 Indiana 562 California 780 Alberta 318 Louisiana 570 Pennsylvania 781 Massachusetts 319 Iowa 573 Missouri 784 St VincentJGrenadines 320 Minnesota 580 Oklahoma 705 Kansas 321 Florida 586 Michigan 705 Florida 323 California 501 Mississippi 787 Puerto Rico 330 Ohio 602 Arizona 800 Toll-free services 331 Illinois 603 New Hampshire 801 Utah 334 Alabama 604 British Columbia 802 Vermont 336 North Carolina 505 South Dakota 803 South Carolina 340 U.S. Virgin Islands 606 Kentucky 804 Virginia 344 California 507 New York 805 California 345 Cayman Islands 608 Wisconsin 806 Texas 347 NewYork 609 New Jersey 807 Ontario 352 Florida 610 Pennsylvania 808 Hawaii 360 Washington 611 Repair Service 809 Dominican Republic 401 Rhode Island 612 Minnesota 809 St. Vincent & Grenadines 402 Nebraska 613 Ontario 810 Michigan 403 Alberta 614 Ohio 812 Indiana 404 Georgia 615 Tennessee 813 Florida 405 Oklahoma 616 Michigan 814 Pennsylvania 406 Montana 617 Massachusetts 815 Illinois Number 816 Missouri 817 Texas 818 California 819 Quebec 828 North Carolina 830 Texas 831 California 832 Texas 843 South Carolina 847 Illinois 850 Florida 858 California 860 Connecticut 864 South Carolina 867 NW Territories 868 Trinidad &Tobago 869 St. Kitts/Nevis 870 Arkansas 847 Illinois 876 Jamaica 877 Toll-free services 080 Toll-free services 881 Toll-free services 882 Toll-free services 888 Toll-free services 900 Information Services 081 Tennessee 902 Nova Scotia & PEI 903 Texas 904 Florida 905 Ontario 906 Michigan 507 Alaska 508 New Jersey 509 California 910 North Carolina 911 Emergency Services 912 Georgia 913 Kansas 914 New York 915 Texas 916 California 917 New York 918 Oklahoma 919 North Carolina 920 Wisconsin 925 California 931 Tennessee 935 California 937 Ohio 940 Texas 941 Florida 949 California 954 Florida 956 Texas 970 Colorado 971 Oregon 972 Texas 973 New Jersey 978 Massachusetts 285 TeleGeography 2000 © TeleGeograPhyo Inc. 1999 North American Area Codes BRITISH COLUMBIA 250 Vafi~uver ," WASHINGTON 360 ,; 509 OREGON Edmonton ¯ 780 %;’~:~.~ ................................ ALBERTA SASKATCHEWAN Regina ¯ Winnipeg ¯ NORTH DAKOTA MONTANA 218 MINNESOTA iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii IDAHO ji SOUTH DAKOTA !i!iiiiiiiiiiiiijiiiiiiiiiiii!iiiiiiiiiiii!~07 CALIFORNIA;,’ ’/ .......... "’7 3~7 "’"’.. NEBRASKA ’i ..... !...~V; NEW MEXICO 970 ( Den~,er 1 ,"913~8161 ’,, 720 ; ! 785 719 I KANSAS .......... 316 ...... .... ¯ ! 412 ; OKLAHOMA ! 91~ ~," 903 ......... 9f5 fEXAS 409 B30 512 286 © TeleGeographyo Inc. 1999 TeleGeography 2000 ONTARIO 705 MAINE For map licenses, send email to: info@telegeogra phy.com © TeleGeography, Inc. 1999 287 TeleGeography 2000 © TeleGeography, Inc. 1999 A Primer on Bits Measuring Bytes Bit by Bit Below are the standard metric prefixes used in the Sl (SystOme International) conventions for scientific measurement. With units of time (e.g., gigabits per second) or things that come in powers of 10, they retain their usual meanings of multiplication by powers of 1,000 = 103. When used with bytes (e.g., gigabytes of data storage) or other things that naturally come in powers of 2, they usually denote multiplication by pow- ers of 1,024 = 210. Base 10 Base 2 1 Kilobit!s = 1,0001 = 103 = 1,000 1 Kilobyte = 1,0241 = 21° = 1,024 1 Megabit/s = 1,0002 = 106 = 1,000,000 1 Megabyte = 1,0242 = 220 = 1,048,576 1 Gigabit/s = 1,0003 = 109 = 1,000,000,000 1 Gigabyte = 1,0243 = 230 = 1,073,741,824 1 Terabit/s = 1,0004 = 1012 = 1,000,000,000,000 1 Terabyte = 1,0244 = 24o = 1,099,511,627,775 1 Petabit/s = 1,0005= 1015 = 1,000,000,000,000,000 1 Petabyte = 1,0245= 25o = 1,125,899,906,842,624 1 Exabit/s = 1,0006 = 1016 = 1,000,000,000,000,000,000 1 Exabyte = 1,0246 = 260 = 1,152,921,504,606,846,976 1 Zettabit/s = 1,0007 = 1021 = 1,000,000,000,000,000,000,000 1 Zettabyte = 1,0247 = 27o = 1,180,591,620,717,411,303,424 1 Yottabit/s = 1,0006 = 1024 = 1,000,000,000,000,000,000,000,000 1 Yottabyte = 1,0248 = 26° = 1,208,925,819,614,629,174,706,176 © TeleGeography, Inc. 1999 Measuring Telecommunications BandwidthmDS-O to 0C-192 Carrier Technology Data Rate(Mbps) Description 64 Kbps Circuits* DS-O 0.064 Base rate in the Digital Signal (DS) level hierarchy 1 T-1 (DS-1) 1.544 Primary level of the American T-carrier multiplexing 24 system; capaciW is the same as a DS 1 carrier Four times the capacity of T-1 28 times the capacity of T-1 168 times the capacity of T-1 T-2 (DS-2) 6.312 T-3 (DS-3) 44.736 T-4 (DS-4) 274.176 E-1 2.048 E-2 8.448 E-3 34.368 E-4 139.264 E-5 565.148 OC-1/STS-1 51.840 0 C-3/STM 1 155.520 OC-12/STM-4 622.080 0C-24 1,244.160 0 C-48/STM 16 2,488.320 0C-192/STM-64 9,953.280 Primary level of the European E-carrier multiplexing system Carries four multiplexed E-1 signals Carries four E-2 signals Carries four E-3 signals Carries four E-4 signals Basic signaling rate of SONET hierarchy Exactly three times the capacity of 0C-1"* 12 times the capacity of 0C-1 24 times the capacity of 0C-I 48 times the capacity of 0C-1 192 times the capacity of 0C-1 96 672 4,032 30 120 480 1,920 7,680 810 2,430 9,720 19,440 38,880 155,520 Key "OC" "STM" "STS" T-carrier system in U.S., Canada, and Japan with 1.544 M bps as the primary level (24 voice channels x 64 Kbps per channel). Digital Signal that travels on the T-carrier or E-carrier. Used in countries other than U.S., Canada, and Japan. The hierarchy was established by the CEPT (Conference Europ~enne des Pastes et T~l~communications) with 2.048 Mbps as the primary level ([30 voice channels 2 channels for overhead] x 64 Kbps per channel). Optical Carrier interface designed to work with STS-n (Synchronous Transport Signal) signaling rate in a SONET (Synchronous Optical Network). Synchronous Transport Module refers to a large carrier (base signal 155.52 Mbps) in a SONET. Synchronous Transport Signal is the electrical counterpart to the Optical Carrier (OC). Voice traffic on digital networks is commonly encoded at 64 Kbps, but the actual number of simultaneous conversations possible over a given carrier may vary depending on the encoding scheme used. ** In the "E" and "T" hierarchies, each higher level is set to be "almost but not exactly" a multiple of the bit rate for the previous order (plesiochronous). To elimi- nate problems associated with plesiochronous multiplexing, SONET, a synchronous hierarchy, was defined in the United States in 1986. As a result, the "OC" and "STM" carriers are exact bit-rate multiples of their primary levels, 0C-1 and STM-1, respectively. Source: TeleGeography, Inc.; Alcatel; Newton’s Telecom Dictionary © TeleGeography, Inc. 1999 288 Over 1,700 international telephone companies are now in business. How many can you name? New International Carriers 1999 The market for international telephone services is global. And so is our directory. Because we know that it’s hard to stay ahead of the competi- tion if you don’t know who it is. TeleGeography’s unique three volume directory has the market intelli- gence you need to stay on top of the market. The books provide cru- cial information on over 1,700 international carriers in the world’s top 150 economies. Each carrier listing contains: ¯ Address, telephone, fax, web site ¯ Names of key contacts and executives ¯ Status of international service authorization ¯ Details on shareholders and affiliates ¯ Stock symbols and exchange listings Available in paper directory or on-line database (exportable) format. To order, contact us at: TeleGeography, Inc. tel: +1 202 467 0017 fax: +1 202 467 0851 http://www.telegeography.com info@telegeography.com About TeleGeography, Inc. telegeography \t~l’a-j~-0g’ra-f~ n (I 990) abbrv, of telecommunications geography [fr. Gk tale, far off, at a distance and L. communicatus, pp. of communicare to impart fr. Gk geo (earth) graphein, (to write)] I. a new branch of geography that maps the pattern of tele- phone traffic and other electronic communication flows; 2. places created by or perceived solely via telecommunications (e.g., a computer network address); 3. the telecommunications artifacts (radio antennae, terminals, signs) on a site; 4. the balance of telecommunications power in one country or region vis-a-vis another (cf. geopolitics, archaic). The old geography of countries and coast lines is giving way to a new geography marked by tele- phone codes, satellite "foot prints" and lnternet addresses. Electronic networks have made the world smaller. But they also have created countless new places where people work and play. This expanding electronic terrain--call it telegeography~demands a new cartography. That is the raison d’etre of TeleGeography, Inc. (TGI). Founded by Washington, D.C. lawyer, Gregory C. Staple, TGI is the world’s leading publisher of reports on international telecommunica- tions flows. It also is an authoritative source of data on the ownership of information-communica- tion companies worldwide. In addition, the company provides customized mapping and consulting services. TGI’s flagship report on international traffc--TeleGeography--has been published annu- ally since 1989. Other TGI publications include: Direction of Traffic 1999, a historical traffic atlas compiled with the ITU; New International Carriers, a three volume directory of competing international telephone companies; The International Bandwidth Report, a guidebook for purchasing international capacity (available Spring 1999); Global Communications Cable and Satellite Map, a poster-sized map of telecom infrastructure; and Global Communications Traffic Map, a wall map of international traffic flows. TGI’s directories are used by thousands of leading communication companies, governments, and financial institutions in over 100 countries. To learn more, please visit us at http:I/www.telegeography.com. Tel.+ I TELEGEOGRAPHY, INC. 1730 Rhode Island Avenue, NW ¯ Washington, DC 20036 USA (202) 467-0017 ¯ Fax +I (202) 467-0851 ¯ E-mail: info@telegeography.com http :I/www.telegeography. corn One Hu fln unications TeleGeography, Inc About TeleGeography, Inc. telegeography \t~l’a-j~-~g’ra-f~ n ( 1990] abbrv, of telecommunications geography [fr. Gk tele, far off, at a distance and L. comrnunicatus, pp. of communicare to impart fr. Gk geo (earth) graphein, (to write)] 1. a new branch of geography that maps the pattern of tele- phone traffic and other electronic communication flows; 2. places created by or perceived solely via telecommunications (e.g., a computer network address); 3. the telecommunications artifacts (radio antennae, terminals, signs) on a site; 4. the balance of telecommunications power in one country or region vis-a-vis another (cf. geopolitics, archaic). The old geography of countries and coast lines is giving way to a new geography marked by tele- phone codes, satellite "foot prints" and lnternet addresses. Electronic networks have made the world smaller. But they also have created countless new places where people work and play. This expanding electronic terrain--call it telegeography--demands a new cartography. That is the raison d’etre of TeleGeography, Inc. (TGI). Founded by Washington, D.C. lawyer, Gregory C. Staple, TGI is the world’s leading publisher of reports on international telecommunica- tions flows. It also is an authoritative source of data on the ownership of information-communica- tion companies worldwide. In addition, the company provides customized mapping and consulting services. TGI’s flagship report on international tra~c--TeleGeography--has been published annu- ally since 1989. Other TGI publications include: Direction of Traffic 1999, a historical traffic atlas compiled with the ITU; New International Carriers, a three volume directory of competing international telephone companies; The International Bandwidth Report, a guidebook for purchasing international capacity (available Spring 2000); Global Communications Cable and Satellite Map, a poster-sized map of telecom infrastructure; and Global Communications Traffic Map, a wall map of international traffic flows. TGI’s directories are used by thousands of leading communication companies, governments, and financial institutions in over 100 countries. To learn more, please visit us at http://www.telegeography.com. TELEGEOGRAPHY, INC. 1730 Rhode Island Avenue, NW ¯ Washington, DC 20036 USA (202) 467-0017 ¯ Fax 1 (202) 467-0851 ¯ E-mail: info@telegeography.com http://www.telegeography.com