acm-header
Sign In

Communications of the ACM

Communications of the ACM

Moving Toward a Non--U.S.-Centric International Internet


The international structure of the Internet has been very U.S.-centric. It has been cheaper for ISPs in other countries to send traffic to the U.S. than to transmit it directly to regional centers. As a consequence, the Internet looked as though the U.S. ruled the roost. This structure is beginning to change—dramatically. Over the past year and a half, the European parts of the Internet have become much more interconnected, with high-capacity bandwidth finally linking most areas. And Asia began to see the first stages of interconnectivity expand rapidly to form an intra-Asian "mesh" network, effectively joining its major national centers. While this mesh previously offered connections in the KBps range and was often filled with communications traffic, it is now beginning to deliver tens of MB of capacity. By the end of 1999, Europe and Asia should both be able to enjoy an infrastructure that is no longer as U.S.-centric as it was just a year or two ago.

Before 1998, especially during the early development of international links between the U.S. and other national Internet connections, the ties went from the U.S. to Sweden, to France, to the U.K., or to Japan, but almost never involved a regional level of connectivity. It was cheaper to connect from Stockholm to the U.S. and then to London, though the physical route to London was much shorter. In Europe and Asia, there were few intraregional connections [2]. The connections between Scandinavian countries and the U.K. or to continental Europe often flowed through the U.S., because it was cheaper for international carriers and ISPs to lease data circuits across the Atlantic than to lease circuits between parts of Europe.

Europe was the first to change. Significant large new high-capacity bandwidth networks were built in 1998, beginning with the Ulysses project deployed by MCI WorldCom between London, Amsterdam, and Paris [3]. This project resulted in substantially increased bandwidth between major cities. By early 1999, MCI WorldCom's Ulysses network linked cities in France, Germany, and the U.K. through several 155Mbps circuits. Only a year earlier, many of these cities had connections to major European centers on a much smaller scale, offering much less communications capacity than their links to the U.S. And other carriers and carrier alliances announced plans to deploy similar high-speed European networks, including Deutsche Telekom-France Telecom, British Telecom-AT&T, Cable & Wireless, and Qwest. Earlier this year, Qwest began service on a new European backbone network offering far more capacity than Ulysses; today, it has 2.5Gbps capacity in place between most of the largest European business centers.

To give some idea of the scope of this change, we can compare bandwidth from several European "gateway" cities—those in which international connections are usually made through an international gateway—to the U.S. and to other European centers (see Table 1). For example, while bandwidth to the U.S. more than doubled during 1998, the bandwidth between some of the largest European gateway cities grew nearly twice as rapidly. This dramatic growth completely reversed a pattern of dependence on links to the U.S. that is still common in Asia (see Table 2).

During 1999, the growth of bandwidth between major European cities may be even more dramatic than it was during 1998. Some cities that were not well connected to the European Internet before 1998 saw their links expand rapidly, including Milan, Italy, which was previously a peripheral European center in terms of Internet infrastructure. In 1998, it gained nearly 68Mbps of capacity to the U.S. Internet and about 30Mbps in connections to European centers. The expansion of MCI WorldCom's network will add about 34Mbps of European capacity to the links already in place between Milan and other European centers. Meanwhile, some cities in Germany, particularly Frankfurt, the country's principal financial center, are benefiting from the expansion of MCI WorldCom's European infrastructure. The city is gaining 155Mbps links to other European centers, such as Paris and Amsterdam, and to other German locations thanks to this expansion. It will gain additional connectivity from the other trans-European networks that will be deployed by British Telecom-AT&T, Cable & Wireless, and France Telecom-Deutsche Telecom-Global One.

Back to Top

Corporate Connections

What global economic forces are driving this growth? The need for connectivity between large, multinational corporations has motivated many communications carriers to aggressively acquire local ISPs and deploy sizable trans-European networks. Thus, the new European Internet infrastructure meets the demands of corporate customers building Intranets and extranets to connect their business partners to their networks and create a global corporate mesh network. Each carrier helps create this new network for all to use. Through it, employees in large corporations can access information in any part of the corporation and collaborate easily with far-flung subsidiaries. They can even distribute their work on a 24-hour basis, so the sun never sets on projects with close deadlines.

Cross-border mergers, such as Ford Motor's recent acquisition of Volvo and Daimler's acquisition of Chrysler, drive the need for global connectivity even more. Ford, a major player in the Automotive Industry Exchange (known as ANX), the industry's high-capacity, high-quality Internet backbone (provided by certified ISPs), has put tens of thousands of its employees on the Internet and is using the Net as a way to gain competitive advantage through reduced costs for designing cars and manufacturing parts. Engineers and managers in Dearborn, Mich., want to "see" what is happening in Volvo as easily as they follow activities in Ford plants in the U.S.


A critical business issue is whether U.S. dominance in e-commerce might prompt a reversal in the decentralization process.


Moreover, new technologies and deregulation are contributing to the almost constant investment in new European infrastructure. The move to optical networks that will employ dense wavelength division multiplexing (DWDM) technology for packing enormous amounts of data, voice, and video onto fiber and provide easy access to metropolitan areas is creating an environment that I expect will render the older networks obsolete in a few years. Telecom deregulation of national networks provides opportunities for new carriers in Europe, such as MCI WorldCom, to challenge the newly privatized government-owned national telephone monopolies for key business customers. In an unpublished study I conducted early this year of dedicated Internet access prices in Europe, MCI WorldCom's UUNet subsidiary in Italy was found to be offering prices below what its parent company offered in New York, far undercutting Telecom Italia's prices.

Similar changes are beginning in Asia, too. During 1999, we have begun to see major shifts in the Asian mesh network, with large new high-capacity bandwidth deployed between major centers for the first time. Even as of late 1998, most bandwidth from Asia still flowed to, through, and out of the U.S.

Back to Top

Merged Voice and Data

The reasons for changes in the U.S.-centric picture of the Asian international Internet are quite different from those in Europe. In Europe, corporate business and early moves toward telecom deregulation propelled the shift to investment in high-capacity bandwidth Internet infrastructure. In Asia, the motivations are more complex. First, there has been an increase in corporate use of the Internet, speeding growth in intra-Asian international Internet connectivity. In Asia, there is an active group of smaller firms using the Internet intensively to rapidly build up their businesses. Second, and perhaps more important, international carriers are building an Asian mesh network to provide merged voice and data services among Asian countries. This expansion is intended to move Asian voice traffic onto a cheaper backbone network so traditional carriers can continue to compete with low-cost voice-call companies leasing excess capacity.

Consequently, during 1999, Asia is seeing an increase in international Internet infrastructure that also handles voice traffic. It is the same type of DWDM optical network that has been deployed across the Atlantic and in Europe. The infrastructure —consisting of a synchronous optical network (SONET), a communication format allowing different types of formats to be transmitted on the same communications line, and the synchronous digital hierarchy (SDH) data transmission standard—will support asynchronous transfer mode (ATM)-based networks similar to the ones carriers have built in the U.S. and Europe. The difference will be that once voice traffic moves onto the Asian network, there should be a dramatic increase in backbone capacity between Asian countries for both voice and data.

However, unlike the case in Europe, where some carriers have described their ATM backbone networks in some detail, there is little public information about what will happen in Asia. My recent discussions with representatives of MCI WorldCom indicate the company will deploy additional bandwidth from the U.S. to Tokyo, Sydney, and Hong Kong this year. MCI WorldCom will also begin to connect major Asian centers, with connections running at 10–45Mbps between Tokyo, Hong Kong, and Sydney. In addition, MCI WorldCom will begin to deploy an Asian mesh infrastructure that will join other Asian centers, including Manila, Singapore, Jakarta, and Bangkok, to the first major loop.

Other global telecom carriers, such as AT&T, Teleglobe, Sprint/Global One, Cable & Wireless, and Qwest, will likely consider similar optical-based infrastructures for Asia, but they have not disclosed their plans to the public. Nearly all of these carriers have dramatically upgraded the features and business services they offer in Asia over the past year. For example, AT&T now provides virtual private network (VPN) capabilities for large multinational businesses, as do Sprint/Global One, Teleglobe, and others. Equant, a new international carrier headquartered in Amsterdam, which before being spun off as an independent company was the airlines' global communications network, has evolved its old X.25 network into a competitive frame-relay network and plans to provide other international data network services to multinational businesses. In sum, Asian networks have been evolving and becoming more sophisticated. The breakthrough capacity to a more Asian-centric Internet infrastructure has yet to occur but should soon.


The owners of the networks may no longer be the telecom giants of old, but the bandwidth will be far greater and much cheaper.


A critical business issue that could potentially slow future growth in Asia is whether U.S. dominance in e-commerce might provide the basis for a reversal in the decentralization process. The business-to-business e-commerce that will dominate Internet traffic during the next three years will motivate development of even more connectivity than before. U.S. firms that want to use suppliers in Japan, Australia, and Europe will want their partners to have high-capacity bandwidth Internet infrastructure that facilitates interconnectivity. Japanese global consumer electronics firms, such as Sony, will also press for high-speed connections to the U.S. and Europe. The owners of the networks may no longer be the telecom giants of old, but the bandwidth will be far greater and much cheaper.

Back to Top

Conclusion

The new meshed networks in Europe and Asia are ending the U.S.-centric nature of the international Internet. They will also result in greater intra-regional connectivity. New, optical-based networks, similar to those in the U.S., are becoming commonplace around the world. During the next two to five years, they will support corporate data, voice, and video services that have traditionally been segregated on isolated networks.

The new international Internet infrastructure should yield multiple benefits. First, prices should fall dramatically, especially for intraregional connections. The Internet's U.S.-centric nature was created by the economics of international connectivity. If it were two or three times more expensive to lease or pay for intraregional connections, Internet traffic would flow to the U.S. on cheaper links. As more intraregional structures are built—even if they are driven by the need for cheaper voice traffic in Asia—Internet connections will cost less. More such structures should mean the end of U.S. dominance over regional traffic flows (see Petrazzini and Kibati's "The Internet in Developing Countries" in this issue).

Second, more regional users of the Internet will join the large international population already avidly using the Internet. Further increasing the international Internet population will introduce new corporations and consumers to the Net, creating a more diverse, less U.S.-oriented World-Wide Web. A more internationalized Web will in turn have important implications for the future of e-commerce, especially of the business-to-business variety. It will mean that selling and buying between non-English-speaking firms will be more important. It will create additional opportunities for indigenous Asian and European Web developers and content creators. Over the next two or three years, the nature of the Web could also become less U.S.-centric in a cultural and business sense.

And third, new investments in the Internet could create opportunities for U.S. Internet firms to increase their business in Europe and Asia. It might also create opportunities for Asian and European Yahoos, Netscapes, and America Onlines.

Back to Top

References

1. Cohen, R. European Internet backbone connectivity. In Telegeography 1999, G. Staple, Ed. Telegeography, Washington, D.C., 1999; see www.telegeography.com.

2. Paltridge, S. Internet Traffic Exchange: Developments and Policy. OECD Committee for Information, Computer, and Communications Policy, Paris, 1998; see dsti/iccp/tisp(98)1/final.

3. Williams, C. WorldCom's International Strategy. Presentation to investors, 1997 (formerly available at investor relations part of WorldCom Web site, no longer available).

Back to Top

Author

Robert B. Cohen ([email protected]) is an adjunct fellow in the Economic Strategy Institute in Washington, D.C. and president of Cohen Communications Group in New York.

Back to Top

Tables

T1Table 1. Estimated bandwidth from major European gateway centers, late 1998.

T2Table 2. Estimated bandwidth from major Asian gateway centers, late 1998.

Back to top


©1999 ACM  0002-0782/99/0600  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 1999 ACM, Inc.


 

No entries found