Background

"to assist the U.S. research and education community (R&E) in meeting its needs for next generation international Internet services." . . . "HPIIS will seek high-performance connectivity between the NSF's very High Speed Backbone Network Service (vBNS) and high-performance networks of major international research partners."

High performance R&E Internet services (both international and domestic) will be based on high-bandwidth communication links and Internet features that use the NSF's very High Speed Backbone Network Service (vBNS) and the high-performance networks of major international research and educational organizations. The NSF's goal is that these networks would operate and interconnect with a minimum bandwidth of 155 Mb/s (OC-3).

Work is currently underway in the U.S. and other countries to develop national, continental, and intercontinental high-performance Internet services for research and education. The NSF's solicitation seeks to broaden U.S. participation in these global developments and, by extension, provides an avenue for future, potential collaborative projects that build upon the HPIIS.

This white paper presents some issues and concerns that may help the international research and education communities in planning for projects and/or initiatives that can assist in defining the scope and functionality of the HPIIS as it is being developed.

Current State of International Networking

It is projected that the number of Internet users in the research and education community will continue to grow, and that these users will continue to require new levels of connectivity, performance, and services. New applications involving distributed high-performance computing, remote visualization and imaging, and telecollaboration, together with growth in aggregate traffic, make the provision of an increasingly high-performance network infrastructure necessary [1].

U.S. R&E Networks and the vBNS

• network connections at research and education institutions;
• regional networks;
• international connections in support of the global activities of U.S. research and education; and
• a very high speed Backbone Network System called the vBNS, first implemented to connect the NSF Supercomputing Centers -- currently
  • the Cornell Theory Center
  • the National Center for Atmospheric Research
  • the National Center for Supercomputing Applications
  • the Pittsburgh Supercomputing Center
  • and the San Diego Supercomputing Center
  a growing number of U.S. universities -- currently
  • the University of Pennsylvania
  • the University of Illinois at Chicago
  • the Northwestern University
  • the University of Chicago
  • the University of Colorado at Boulder
  • the Penn State
  • the Carnegie Mellon University

  (a complete, updated list is maintained by NSF and another list is maintained at the vBNS web site)

  and the NSF-sponsored NAPs:

  • the Ameritech NAP
  • the MFS NAP
  • the Pacific Bell NAP
  • the Sprint NAP

The vBNS and Internet2

This effort is being undertaken because the existing Commodity Internet connections available to many universities are insufficient for their needs, and because it is not possible to experiment with the Commodity Internet as would be the case if the universities controlled their own network.

Rather than each university being attached directly to the I2 Backbone, member universities are expected to aggregate into new regional networks (called gigapops), with each gigapop attaching to the I2 Backbone. Each university would have to fund both the I2 Backbone and the cost of joining a gigapop.

This new network infrastructure can be thought of as "an Intranet for U.S. universities". Essentially this new network infrastructure is being modeled after the old NSFnet paradigm.

Initially, the I2 Backbone will be the NSF's vBNS network, and in the long run, I2 usage may provide long term funding for the vBNS with the possibility that the vBNS is reformulated exclusively as the I2 Backbone. This potentially greatly expanded role of the vBNS within the U.S. makes the vBNS an even more important portion of any future international R&E Internet.

International Access to the vBNS

There are two fundamental infrastructure issues that must be dealt with to achieve the NSF's goals.

The first is to obtain OC-3 facilities to interconnect the various national R&E networks. Very few national networks and even fewer international networks operate at dedicated OC-3 speeds. References in Appendix A point to several example diagrams to illustrate the state-of-the-art in national networking.

The second issue is that IP routing must be engineered in such a way that no commodity IP traffic flows through the vBNS from any of the HPIIS institutions; only approved R&E traffic may flow over the vBNS. Figure 1 illustrates the logical separation of commodity and HPIIS traffic among HPIIS institutions.

A Current International Internet Access Point

Funded for three years by the National Science Foundation, the STAR TAP project will provide a persistent connection point for U.S. and international high-performance research networks and will improve the speed and performance of the applications that will run over these networks.

It is possible that other international access points to the vBNS will also be established.

Sample Access Proposal

Access via STAR TAP is one possible route that a designated institution might use in creating an international Internet. One such proposal, from the country of Singapore, is currently being reviewed by the NSF to permit that country to connect to the U.S. vBNS. An excerpted version of this proposal provides some insight into the kinds of projects, resources, and other requirements necessary to meet the AUP for connecting to the U.S. vBNS.

Conclusion

As these developments are proceeding, opportunities to team with these consortia or to provide test-bed projects that demonstrate how the international Internet might benefit the research and education communities around the world will become available. It is important that such opportunities are seized and that projects (both national and international) be designed to help direct the evolution of the newly emergent international Internet services.

[1] Excerpted from: Networking and Communications Research Infrastructure