Eugene, Oregon
May 13-15, 1998

The Future Network: Transforming Learning and Scholarship

Internet2: Background and Mission

Douglas E. Van Houweling, President
University Corporation for Advanced Internet Development (UCAID)

Thank you for inviting me here today. This is an extraordinarily important assemblage of people who have, in my view, perhaps the most important set of missions for guarding the future of our enterprise of any group that I've had a chance to address since I assumed this position as leader of the University Corporation for Advanced Internet Development. The opportunities that the new network capabilities I'll be discussing offer to all of us will almost certainly outstrip anything that we can imagine today.

When we were talking about this at dinner last night, it became even clearer to me that the partnership between the research library community and the network and applications community will be absolutely fundamental to the ability of the higher education community--and ultimately the population of the world--to take advantage of the new facilities that we can make available. So I will try to move rapidly across the terrain of what we talk about when we discuss Internet2 and then leave time for some discussion so we can make sure that we've begun to knit the relations that will be required for our future.

I want to talk first about Internet2 and what it is. Second, I want to provide you with a bit of history and background. I want to discuss with you some of the limitations that we live with in this phase of the Internet, some barriers to progress that we all confront, and the projects that have arisen from that. And I want to move into applications--I have a number of examples--to talk about the network requirements and our recent announcement with Abilene, and finally I will focus on some implications and receive some comments and questions from you.

The University Corporation for Advanced Internet Development (UCAID) is a not-for-profit corporation. It was formed in order to support the Internet2 project and related activities. Today there are two projects in the University Corporation, the Internet2 project and the Abilene project. These projects are only the beginning of the set of efforts that we expect will be included in the University Corporation's future.

The mission of UCAID is to provide leadership and direction for advanced networking and development within the university community. The organization is built around a board of university chief executive officers who represent all the regular members. We structured this as a very agile organization capable of responding to rapid change, because that's what this industry demands.

We have three councils that advise the board: A Council on Applications, a Council on Policy and Operations, and a Council on Network Research. The chairs of all of those councils sit on the board. Dues from member institutions provide the income base, which now amounts to somewhere in the neighborhood of three million dollars per year. It's an extraordinary collection of individuals spread quite broadly across the higher education community and includes people with extraordinary knowledge, expertise, and interest in our activities.

There are now 126 regular members, 20 affiliate members, and 30 corporate members in UCAID. We have an open application process, which means that any organization that commits to the objectives and commits the resources required to reach those objectives can become a member. The most important of those is that all of our regular members are committed to making sure that on their campuses adequate infrastructure is deployed and adequate connections are made to the national facility to support advanced applications. For most institutions, that's an investment of somewhere in the neighborhood of one million dollars a year.

In addition to our regular members we also have affiliate members, groups that tend to be organizations that bring members together, like regional networks and corporate members. In addition, we have a category we call "corporate partners," which means they committed one million dollars or more across a three-year period to supporting the Internet2 project. Most of those funds go directly to your institutions in support of the infrastructure that you need to build on your campuses and the activities you need to support your campus environment.

The Internet2 mission is slightly more complicated than the UCAID mission because it's more specific. We are focused on building an environment that supports advanced network-based applications and the services that are required to support those applications. Furthermore, we are focused on making sure that technology becomes available throughout the global Internet on a wide basis. So our goals are, first of all, to enable the new generation of applications; second, to recreate a leading edge research and education network capability; and, third, to transfer that capability to the global production of the Internet.

A lot of you know networking history. Research universities really started this business in collaboration with the National Science Foundation when we developed NSFnet. Up until that time, as most of you know, people thought of the Internet as a scientific curiosity focused on the needs of computer scientists and a few others. As a result of the work that we did together and the not-for-profit corporations we founded, we were in a position in 1994 when the World Wide Web and the WWW browsers began to become available to support what has become an explosion in development--creation, development, and access to information on a worldwide basis.

In 1994-1995, the higher education community was encouraged by our colleagues in the commercial Internet business to withdraw from the business of providing Internet services. First of all, it was stated that it was inappropriate for not-for-profit organizations to be heavily engaged in the competitive for-profit industry that was at that time developing. Second of all, we were assured that the for-profit environment would create the kind of capabilities that we needed for our future. We were happy to withdraw. Our business, after all, is not primarily at deploying services; our business is trying to focus on inventing the future.

We discovered, however, that we left an enormous gap when we did withdraw. The focus of the commercial Internet industry is on meeting the demand of the next three to six months. The focus of the higher education community is on thinking what we need to be doing two, three, or four years from now. As a result of that, a set of activities were undertaken to create the capability for us, along with our partners in the federal mission agencies, to reenter the advanced Internet technology environment. In 1997, both the Internet2 project and the federal government's Next Generation Internet Initiative were commenced.

Let's talk a little bit about why that turned out to be necessary. Today's Internet, as you all know, is growing at a rate of ten to 15 percent a month. It's been growing at that speed for about a decade. If you stop and think about any enterprise that has that kind of a growth rate across a decade, you begin to understand why this environment is having such an enormous impact on the way we all live and do our work.

The problem, of course, is that there is no technology that grows at a comparable rate. The consequence of this growth rate, combined with the slower growth and speed of the base technology, means that the Internet has become a heavily congested environment as capacity lags the applications. That's especially true for applications that are focused on direct human interaction.

You've all tried to browse publications on the Web, and you've discovered that that's a lot slower than turning pages in a book. And the random access facilities really don't work all that well. If you really want to do shared offering in real-time with somebody at a distance, the time lags and the lack of dependability in the environment get in the way. We would very much like to have virtual meetings and seminars, but at today's level of performance the Internet and its applications aren't capable of supporting that. And, of course, our research community would like to undertake distributed large-scale computing and database applications, but today's Internet doesn't support that effort, either.

In fact, because of these problems and because of some basic liability problems, we hesitate to pursue mission-critical applications on the public Internet. Very few faculty members are brave enough to take the Internet into the classroom and in real-time show their students things that are happening across the world simply because, in that kind of a situation, the Internet doesn't approach the reliability of the chalkboard. Furthermore, we don't know and we are increasingly concerned that the sites we believe we are connected with are really what they are presented to be. We don't have the right level of authentication capability in today's Internet.

So what we have done, especially in the commercial sector, is invented Intranet and Extranet. The Intranet--which most of us have in some fashion or other on our campuses, quite often to support administrative applications--is a part of the Internet that is walled off from the rest of the world and can meet the performance and reliability criteria we need for mission-critical applications. Extranet is the same thing, only linking a large number of institutions together.

The problem with both of those solutions, though, is that they are walled off from the public Internet. So, if your objective is to reach out to the broader population involving the Internet's current challenges, using Extranet and Intranet will not meet the objectives. And, of course, higher education has the need to reach the broad population. So the barriers to our progress are that the providers are swamped attempting to match the passage in demands, and, partly because they're swamped and partly because it's a very competitive industry, there is no large-scale development environment available to support advances in the base technology. Furthermore, because there's no way to pass guarantees of service back and forth between Internet providers, there is not today a large incentive for commercial providers to focus on major investments to improve performance. Thus, we can't deploy the advanced applications we would like to deploy.

So the Next Generation Internet and the Internet2 projects are responses to that need. The first is focused on the needs of the federal mission agencies; the second is focused on our needs, the higher education community's. Fortunately, the two collaborate rather closely together, both at the technical level and as the NGI Initiative is funding a portion of the work of the Internet2 project through the National Science Foundation's division for advanced networking.

What are the attributes, then, of the applications we'd like to undertake using this technology? First of all, we'd like to be able to support interactive research collaboration and instruction. We'd like to provide real-time access to remote scientific instruments. We'd like to provide for large-scale computations. And we'd like to provide for shared virtual reality and combinations of the above. To just give some examples, I will run through a number of the applications that are currently using Internet2 technology.

At Gallaudet there's an effort to teach American Sign Language using Internet2 capabilities, which naturally requires very high-definition video so you can actually see the sign language at a distance. The University of Michigan is making access to a scanning electron microscope available under control at a distance by Internet2 technology. The University of Missouri at Columbia has built a spreadsheet for looking at and managing large volumes of distributed images.

The whole area of neuroscience and the ability to actually understand how the brain and the nervous system operate in response to various stimuli in real-time is being undertaken. We have also been able to bring researchers from all over the world together to cooperatively run an upper atmosphere radar facility and make it possible through links among those people and instruments to run experiments with the worldwide research team.

The University of Chicago is working on tele-immersion activities where people can come together in a shared virtual space. One of the exciting activities now is the educational environment that has been created that helps people study the function of the inner ear, actually looking at how the inner ear functions on a virtual basis. We're also looking at the issue of how the ecosystem operates both in simulated and real spaces.

What it takes for all these applications to work is engaging the applications developers in this new environment and building a network that delivers the functionality they require. The priorities then are to focus on the content and to make sure that we put the middleware in place that allows these applications developers to do the work without having to constantly reinvent all of the basic systems. We need a network that provides performance guarantees in response to what the applications require, that provides authentication and security, and that supports the new business models that will be required as we differentiate service.

In some sense, what we are talking about doing here is moving from a single lane road, which is what the Internet is today, to a multi-lane highway with special purpose lanes for different applications, with access control to control congestion where necessary, and toll systems to recover the costs where that is required. The solutions are obtained in obviously bigger network bytes, new protocols that allow us to support and guarantee quality of service, support for large numbers of people across the world simultaneously viewing a single event through Internet-based broadcasts, and new middleware. Our approach to all of this is that we will in the Internet2 environment develop and test those capabilities and then--through our partnership with industry--have industry deploy these in the commercial, global Internet environment.

Part of our efforts is Project Abilene, which was just announced last month in the White House. Abilene is a high-capacity backbone network that is designed to guarantee that the larger pipes we need will be available to our community. It's directed overall by the University Corporation. Our partners are Qwest, which is providing the fiber; Nortel Networks, which is lighting the fiber; and Cisco Systems, which is routing the packets in this network. We are open to other contributors, and there will be some announced in this area as we move forward.

The Abilene network has extraordinary capacity. It starts at 2.4 billion bits a second among the Internet2 gigaPoPs, and we are planning to move that to 9.6 billion bits per second within a year. We are providing connections to universities at 622 million or even faster--or 155 million bits per second. The result of this is that within easy telecommunications reach of almost every large university gigaPoP we now have a very high-capacity backbone network available to meet the needs.

This is actually the second Internet2 backbone. The first is the vBNS, which is being supported by MCI and the National Science Foundation (NSF). One of our objectives here is to provide a broader spectrum of alternatives in the backbone environment. Of course, we are also going to make sure that all of these networks peer with one other, and we are working with NSF and other federal agencies to accomplish that. We expect that we'll be delivering service on Abilene by the end of this year. And because these agreements are five-year agreements, our application developer can now count on the fact that there will be very high-speed, uncongested network access available for the foreseeable future.

We're also working in computer science communities to make sure that we are doing the kind of research that will take us to the next level of networking in the future. Now, of course, this has generated a lot of interest on an international scale, so there are global issues to consider, and we are working on research partnerships across the world and on collaborating to achieve the basic network capabilities we need. We are using the National Science Foundation's STAR TAP (Science, Technology and Research Transit Access Point) to implement these capabilities, and we are developing an international membership policy for the Internet2 project that will focus on collaborative arrangements with organizations similar to UCAID and in other parts of the world.

STAR TAP is in Chicago. It already links us directly to a number of other networks across the world. In particular, we assigned a Memorandum of Understanding with CANARIE, the Canadian advanced network that is sponsoring CA*net II. That connection is now operational and scholarly collaboration has been initiated. We already have projects in digital music and teleimmersion moving across our board.

So what are the results? Well, I think we need to think about a future in which a lot of things we now think of as separate communications and information access services will be subsumed within this larger Internet environment. Everything from television to movie rentals will now be thought of as being done in the Internet environment. In fact, on the way out here I sat next to a gentleman whose business is in opinion polling, and we got to talking about how opinion polling was going to change in this new world. The business of calling people on telephones to get their opinions probably has a limited future and, while he was somewhat daunted by that prospect, he also understands that most of you would like that very much.

Well, I think the results will be, first of all, that we move from a mass communication world to a personalized communication world, and that interactive electronic communities will proliferate and thrive in this new world. Further, this new network capability is designed for immediate interaction where that's needed, rather than delayed interaction. The media and the transport for all of this will become much less regular and, as a result of increasing capabilities, will move from casual uses--even important ones--to mission critical uses of the network.

The implications for us include at least the following questions: When our patrons go to the Web, where are we? In some sense, we all are a little revolted at the lack of organization, the lack of authority, in the web environment. But, unfortunately, when most of us go there, we don't find our libraries there in leadership and addressing those issues.

Second, we need to ask ourselves what types of information will be most assisted by these advanced Internet capabilities. Just as in the classroom it doesn't make sense to use a computer to turn pages, it probably doesn't make much sense to use the computer to turn pages in the World Wide Web. But if what you really want to do is show a simulation of the neurological process, then that might be an application that we would like to move into this environment. As we see more multimedia products that support the scholarly world, it seems to me that we need to think about which of those products will be best deployed in the new Internet environment. This will be a special challenge for us because we haven't adapted very well to those new media types even within our existing library structures.

The other question is: How shall we utilize the Internet's global reach? If you really think hard about how you're going to hold and secure the intellectual assets of the world in this new environment, how do you approach collection management on a global basis? Will it be possible for us to greatly enrich the access to global resources, especially regional sources? Or will we, by operating in a less coordinated fashion, with all of us at various places around the world, duplicate the core collection and not enhance the diversity of access that would otherwise be available?

And perhaps most interestingly, because of the enormous changes that are going to take place in the publishing and information worlds, does this give us an opportunity to recapture the scholarly communications enterprise for higher education? We are now in a position where the world will change quite dramatically over the next decade. The interesting question for us, I believe, is how to take advantage of that change and begin to assert what many of us believe are the key roles that the scholarly community should play in the knowledge world of the future.

We need to make sure that the advanced infrastructure is broadly deployed and available on our campuses. There's an enormously good start on that already. We need to participate in these advanced Internet efforts, and we need to focus on our assets; try to leave our information and understanding of information and how it's used and reconnect with a kind of creativity that comes out of the higher education environment--which is difficult to find anywhere else in the world--and to engage the applications development process. The library community needs to be deeply engaged in creating new applications for the advanced Internet, and we need to focus on the breakthrough applications that will move us into a new environment, as opposed to simply replicating the environment we understand and work with today.

It's quite an agenda. I'm delighted that I could be here with you today so that we have a chance to chat a little bit about what we might do together. You're critical to the success and the positive impact of this investment that's being made on a global basis, and I'd be delighted to hear your comments and questions. Thank you very much.