12 March 1992 /internet/legislative.actions/hearing.12mar92/habermann.testimony Testimony of Dr. A. Nico Habermann and Dr. Stephen S. Wolff Committee on Science, Space and Technology Subcommittee on Science March 12, 1992 Part 1: Testimony of Dr. A. Nico Habermann Mr. Chairman, thank you for the opportunity to appear before your committee today to provide information about the exciting program in networking supported by the National Science Foundation (NSF) and several other agencies of the U.S. government. I am privileged to serve as the Assistant Director of the NSF for the Computer and Information Science and Engineering Directorate, which has responsibility for broad national research, infrastructure and facilities programs in computer, communications, and information sciences and engineering. Included among my responsibilities is leadership of the overall NSF High Performance Computing and Communications(HPCC) Program with its important components in national Supercomputer Centers and the NSFNET, the subject of our discussion today. In this latter respect, I am pleased to be accompanied by my colleague Dr. Stephen S. Wolff, Director of the Division of Networking & Communications Research & Infrastructure (NCRI). Dr. Wolff has provided leadership for this division since its inception and in this capacity has lead the creation and development of the NSFNET and the emerging NREN program. Before turning to Dr. Wolff, to elaborate on the NSFNET, I would very much like to place this activity in the larger context that it impacts. Background The President's High Performance Computing and Communications Program, which was announced on February 5, 1991, consists of four components, one of which is the National Research and Education Network (NREN). The NSFNET activity is part of the NREN component. The NREN is also a major subject of the High Performance Computing Act of 1991 (P.L. 102-194) that was signed by the President this past December. This Act, that your Committee was instrumental in drafting, provides important impetus to the presidential HPCC initiative. Leadership and direction for the HPCC Program is provided by the Office of Science and Technology Policy, through the FCCSET Committee on Physical, Mathematical, and Engineering Sciences (PMES). The High Performance Computing, Communications, and Information Technology (HPCCIT) subcommittee is chartered under the PMES and is composed of an executive council and four task groups to coordinate science and engineering computing, computer research and development, Federal networking and communications, and education. Since October 1991, I have served as the Co-Chairman of this Networking activity. As described in the Supplement to the President's FY 1993 Budget, "Grand Challenges 1993: High Performance Computing and Communications", NSF is designated as the coordinating agency for the NREN program. As the NREN title indicates, to quote from the Grand Challenges report, "The NREN program is both a goal of the HPCC Program and a key enabling technology for success in the other components. The NREN is the future realization of an interconnected gigabit computer network system supporting HPCC." If we are successful in deploying this technology for the research and education community, then aside from supporting current science and technology Grand Challenge Applications that are important to federal mission agencies, it will broadly influence communications technology development. However, it is important to bear in mind, that the government program, as its name implies, primarily supports computer and communications networking for research and education, not general purpose usage. Nonetheless, the NREN component incorporates important testbeds and research for new communications technologies. The NREN component is dedicated to promoting communications among researchers, educators, and students in the U.S. The NREN activities contribute directly to the goals of the High PerformanceComputing and Communications Program in three ways: 1) by extending U.S. technological leadership in computer communications; 2) by enhancing the dissemination and application of computer and communications technologies to enable advances on applications such as, Grand Challenges; and 3) by demonstrating innovative new means of communication to spur gains in U.S. productivity. In order to achieve these goals, the NREN program consists of two sub-components: one that supports the development and enhancement of network backbone services, which serves the purpose of connecting a large number of regional research and education networks - the Interagency Interim NREN; and a second sub-component, which supports basic and experimental research in the design of large-scale, high-speed networks for future use (gigabit networks R&D). The first NREN sub-component, developing connections between existing and growing regional networks, includes three network backbones supported by NSF, DOE and NASA. The backbone currently supported by NSF is the NSFNET, which connects a large number of regional networks at a variety of educational and research institutions throughout the U.S. The NSFNET backbone, all of whose services are competitively procured from the private sector, provides a networking superstructure that enables scientists and educators to communicate across the boundaries of their regional networks. The second NREN subcomponent, supporting networking research, includes a collection of five gigabit testbed networks, connecting experimental sites across the entire nation. At each step of the development of the NSF NREN program, we must ask ourselves why the government should continue to be involved with the private sector in developing computer network infrastructure to support the research and education community. Although the private sector plays an increasingly important role, there are indeed cogent reasons why the government should stay involved in important aspects of a host of activities in network development and research. In order to put further discussion regarding NSF's role in networking in perspective, it seems proper to list here the main reasons, as we see them, for NSF's continued involvement in support of technology development and deployment. The proper and effective use of very high speed computer networks, and the connection between networks will require innovative research across various disciplines and technologies that government, industry, and academia working together are uniquely capable of providing. It is certain that the capabilities of networks can be increased at least a hundred-fold to support a mode of interaction we can only dream of today. (Imagine, for example, the impact of a hundred-fold increase in both aircraft speed and passenger capacity on travel, military, and on airports, etc!). To meet these challenges, industrial and academic R&D, coordinated and focused by the Federal Government, will concentrate on the advanced generic technologies required to realize a very high speed network. Since there is practically no limit to further development of networking technology, the research and education community should be stimulated to find and explore innovative ways of communicating with each other and with growing information sources. At this time, we think that the development will lead to the use of networks for remote, interactive, real-time computing. However, experience with the ARPANET, designed in the mid-seventies, has shown that the outcome may well be both broader and richer than our original expectations. Networks help broaden the participation for the entire country by providing equal access to advanced computer facilities, such as, the supercomputer centers, for remote and relatively isolated parts of the country and similarly help increase the involvement of minorities and under-represented groups in the research and education enterprise. This enables all scientists and students to more fully participate in leading-edge research and education opportunities that otherwise might not be affordable. My final point supporting NSF involvement in this technology development and deployment relates to the need to encourage all educational institutions, including K-12, to explore the networking capabilities that allow them to access and use the tools that researchers develop and utilize in the work on the Grand Challenges. This can lead to more excitement in education and may stimulate more students to enter science and engineering. NSF welcomes the opportunity to work with the private sector on these and all other aspects of networking to the benefit of our science and education community in the interest of the Nation's future. And now with your permission, I would like to turn to my colleague Dr. Stephen Wolff to provide an overview of the current state of NSF's networking program and summarize the management and development plan and associated policy issues.