November 1, 1996
The parable about the Blind Men and the Elephant applies very well to the current variety of views on what the term "Internet-II" means. Some in the higher education community see it as a separate network exclusively for higher education; for others it is viewed as a research testbed; still others believe it is a publicity ploy to get more Federal money for higher education; many outside of higher education think it must be a solution for the problems of the current Internet. In the view of many of us involved in the planning for Internet-II, there is some degree of validity in each of these views but they all miss the essential point of the initiative. This paper will try to convey the important fundamental nature of the Internet-II initiative as seen by the information technology community that has developed and nurtured it.
It is undeniable that the quality of service as perceived by users of the existing commodity Internet has deteriorated over the last several years. This has had a serious impact on researchers and scholars who must rely on the Internet for their work, must transfer large amounts of data between remote sites, or who try to use streaming applications such as packet video for conferencing or collaboration. However, for the majority of users the occasional delay in delivery of a Web screen is more of a frustration than a strong deterrent. Thus, although the higher education community is concerned with this problem, it alone would not have generated the strong consensus and momentum behind developing a greatly enhanced Internet infrastructure.
The higher education community has incorporated information technology into its academic and research programs for over 30 years. It is not uncommon today for a campus, particularly in the research community, to have a high speed ubiquitous network and at least one computer on most faculty, staff and graduate student's desks. The Internet itself grew up in this environment long before it became popular in the commercial world.
This long familiarity with the use of information technology continues to inform and expand the vision of how such technology might transform teaching and learning, scholarly collaboration and research, and ultimately the university itself. Today teachers are able to have far greater contact with students through the convenience of electronic mail. Class assignments and background materials can be found on the Web so there is no longer the worry about misplacing those essential pieces of paper. Collaboration between students and among researchers takes place over the network asynchronously and far more effectively than might have been achieved before. A number of experiments have shown that it is possible to deliver an entire course in certain fields remotely over the Internet. Yet all these enhancements are merely scratching the surface of a world of compelling possibilities.
The information technology leaders of the major US research universities, in response to increasing demands from teaching and research faculty, have articulated a common vision of what is needed in order to realize the next generation of network-based applications of importance to the higher education community. These include new modes of interactive collaboration and new ways of enhancing learning through information technology, the integration of complex distributed digital library collections with academic programs, greater access to expensive specialized research facilities such as accelerators and supercomputers, and life long scholarly pursuits facilitated through ready access to learning materials from homes, offices, or anywhere convenient to the learner. Electronic access to a broad range of new types of information as well as "raw data" will engender new possibilities for the synthesis of ideas that can ultimately lead to the creation of important new knowledge.
The university as a center of knowledge and learning must become integrated more fully into the fabric of industry and commerce. Modes of learning and collaboration developed during periods of residential education, our traditional period of introduction to higher education, can be perpetuated if sufficient access to learning resources is available. A basic component of this will be an ubiquitous advanced communications infrastructure that can enable teaching and learning to remain a part of everyday life. The future American work force thus equipped will continue to vitalize industry and commerce and help to maintain our world leadership in critical emerging fields.
It is this vision of new modes of higher education enabled through a set of advanced distributed information systems that motivates and energizes the higher education information technology community. The realization of this vision will require major advances in pedagogy, supporting resources, computer software and technology infrastructure. A fundamental component of this complex picture, without which many of the other pieces can not even be developed, is the availability of a rich set of communications services in support of advanced applications requirements.
In the commodity Internet today, there is exactly one "quality of service" - best effort delivery of most data packets at some unpredictable time to a single destination. All current applications are designed with this limitation in mind. The new advanced applications that are envisioned will require greater guarantees of predictable delivery, within a defined time window, and with an acceptable transport capacity. Some applications will require the delivery of the same data to many recipients. Some will tolerate asymmetrical service quality and others will require closely symmetrical service. Each application must be able to request and acquire the required services (or wait until they are available) dynamically.
It is not at all certain that the commercial commodity Internet is prepared to move quickly to deploy the range of advanced Internet communication services that higher education seeks. Higher education now represents only a small fraction of the Internet consumer market. A successful business must focus on developments that "maximize the bottom line" and for the Internet market this means the broad consumer customer base and the high end commercial and industrial customers. These customers are not (yet) demanding the types of data communications services that are needed to support the new types of applications envisioned by higher education.
Given this perspective, information technology leaders in the higher education community have come together with partners from the Federal government and industry to develop a joint strategy to catalyze and accelerate the development of "next generation" applications and the infrastructure required to enable them. The focus of this broad effort is on realizing the vision described above. The first, and in many ways easiest, step is to cause to be put in place a new set of advanced Internet communications services and make them readily available to the community of applications developers and users.
This advanced communications infrastructure has come to be called "Internet-II" because it builds on the common bearer service of the existing Internet - the Internet Protocol - with extensions and enhancements now in the later stages of development in labs and vendor shops. Unfortunately the same terminology has been applied to the broader goals of the applications space as well which obscures the primary meaning of the Internet-II initiative.
Concern has been expressed that higher education is abandoning the commercial Internet industry. Far from it. It is estimated that more than 70% of the Internet traffic entering or leaving our campuses is exchanged with non-university sites. The advanced Internet infrastructure will serve only as a proving ground among those institutions developing the new applications. Furthermore, it is anticipated that success within this precompetitive arena will lead to early adoption of the new technologies and service delivery models by commercial Internet providers. Only the private sector that can build and support the infrastructure required to bring the advanced services into homes and offices to reach the ultimate beneficiaries of the new applications. Therefore it is clear that higher education not only will continue to rely heavily on the commercial Internet for most of its communications services today but is planning for technology transfer to the private sector as soon as possible.
The enhanced Internet architecture, as currently conceived, includes a new infrastructure component that will enable the higher education community to continue to make use of existing Internet services as well as gain access to the new advanced Internet-II services. This component has been termed the "GigaPOP" but might be described more usefully as an Internet-II Services Center (I2SC). One might think of an I2SC as the Internet equivalent of a regional shopping mall: a vendor-neutral place where consumers in the region can come to acquire a variety of Internet transport services. A campus or research institution will be able to lease a single high capacity connection to an Internet Services Center at which it can gain access to differentiated services from one or more pre-competitive as well as commercial service providers. This should lead to healthy competition among Internet service providers as well as prove to be a highly cost effective way for Internet consumers to acquire services.
It must be emphasized that although the I2SC's will be able to make available a wide variety of Internet services, the Internet-II Project community will be defined as those institutions that not only connect to the I2SC's but also provide campus infrastructure support for end-to-end advanced Internet services between participating research and user locations. However, a very important aspect of proving the value of the new applications will be to observe their efficacy in a variety of school, community and professional environments. Therefore it is expected that the community of participants will expand as the project matures and applications become available for testing among a wider audience.
The Internet-II Project is focused on developing and proving the effectiveness of a wide range of new teaching, learning, research and collaboration technologies. Participants will commit considerable financial and intellectual resources towards the successful outcome of this project. The project also will require the availability of advanced Internet services and much higher capacity and "intelligence" in the network than exists today. Achieving this infrastructure will require the continued close cooperation of the entire Internet-II Project community.
Seed money will help move this vision towards reality but ultimately the result must be self sustaining and commercially viable. Higher education looks forward to working closely with partners in the information technology and communications service industry to achieve these goals in the most cost effective and broadly applicable manner.