Internet2 and Networking Futures

The Internet 2 project is higher education's effort to move the Internet and all of the associated uses of the Internet to a new plateau of function and capability.

We not only want to develop this new technology and demonstrate it, but also deploy it to make sure that the technology we've developed will scale and meet the demands of the world. Just as we did with the original Internet work that higher education pioneered for all of us, we believe the time has come to develop a new generation of Internet technology that will enable a whole new field of applications to be utilized in support of education and research. Despite the fact that much of what I say will be focused on the network and the development of technology by research institutions, the ultimate objective is to make that technology and research available much more broadly. As we learned in the development of the Internet, we have to not only do the research that's required to bring a new technology to life, but we also have to demonstrate that it will scale and be useful, not just to hundreds, not just to thousands, but to tens of thousands and hundreds of thousands of people before the rest of the telecommunications industry will take the application seriously and begin to move these capabilities into the private sector.

We've gained experience with what we think of as a technology development cycle, where technology starts in the research labs, moves into a few departments in the institution; then is adopted in several research universities and finally deployed across the research and community. At that point, we get to the level of scale that allows us to demonstrate that the technology be useful for an even larger community and we're able to roll the technology out for all of higher education. We've had that experience before; we're expecting to repeat the experience with the Internet 2 project.

The principles, then, of the Internet 2 project are, first of all; that it's designed from the beginning to benefit all segments of the education community, K-12 schools, museums, community libraries, colleges and universities. That broad benefit will come only after we have developed the technology and deployed it broadly enough to prove its capability.

As a result, the Next Generation Internet/Internet 2 effort is really our next step in achieving these benefits for the broad educational and library community. It focuses, first of all, on protocols and applications that are not yet developed. Second, it maintains the connection to today's Internet so that the new technology can be implemented in a seamless fashion. Finally, it addresses the performance problems we're experiencing in today's Internet; and supports new applications.

As the effort succeeds, history shows that all of higher education will benefit. As a result, we need to think about how federal and state policies should encourage this innovation and make sure the innovation is disseminated across publicly accessible telecommunications networks.

The federal government's research and development process will be an important adjunct to this effort. Just as with the original Internet effort, where the federal government financed only a small portion (less than ten percent) of the total funding that was necessary to create the Internet, but played a key role in providing the incentives that kept the process going forward, we expect in the Internet 2 environment to benefit from the same type of federal investment. But that investment will not succeed unless it once again generates a partnership between government, industry and higher education.

Therefore, our objectives for the project are, first of all, to respond to the needs of research and education, whether it's making it possible for a scholar to collaborate with her colleague somewhere else in the world or making it possible for a teacher to teach a class that is not actually in one physical space, but spread out across the world. We want to make sure that the Internet 2 project focuses on the needs of research and education. In order to do that, we understand that we are going to need to innovate new applications and demonstrate their capabilities. Today, we can do audio over the Internet. We can do video over the Internet -- not quite so well. We can do electronic mail over the Internet. We can access our libraries over the Internet. But what we can't do very well today is create an environment in which all of those things can happen simultaneously, the way they would happen in a single physical environment. For instance, we can teach using interactive video to a remote physical classroom, but we don't have very good tools to gather 20 people together from all over the world, each sitting in front of their own computers, and teach as if that were a classroom.

The whole promise of applications is to take things that today have to be physically real and make them virtual, so that when researchers need to work together, they don't have to come to the same place; but they can stay apart and work together. In order to undertake those applications innovations and to demonstrate their capability, we have to have significant development in the network itself. We need reliable, broad- band connectivity from a desktop in Ann Arbor to a desktop in New York City, or from a studio in New Orleans to a set of people distributed all over the United States. That connectivity not only needs to connect our campuses together, but it also needs to work seamlessly with the technology that we have implemented within our campuses. We furthermore want to make sure that, while we're developing this technology and demonstrating it, that we can remain in control of the capabilities that are required to carry out effective development and demonstration; and that, whenever we need to reach out into the rest of the Internet, that it can be done transparently.

Finally, since only about ten percent of our Internet traffic goes from one of our campuses to another, we need to make sure that what we develop transfers rapidly into the commercial sector, that is going to allow us to reach not just people on other campuses, but students in their homes and researchers, wherever they happen to be working -- whether they are at a university or in some private company.

In order to accomplish these objectives, we have developed an architecture for the Internet 2 which is focused on aggregating bandwidth in particular locations, which we have chosen to call gigapops.

A gigapop allows a university or a set of universities to collaborate in creating a very high speed connection and control point for their interface to the rest of the Internet 2 environment. We anticipate that these switching and control points will be distributed across the United States, each one having somewhere between two and ten universities connected. Those universities will usually come together in regional collaborations which will be led by the participating institutions. Between the gigapops, we expect that we will be able to work with the telecommunications community to find the lowest cost, highest function connectivity that meets the needs of the Internet 2 effort. Today, we're thinking about those connections starting at OC3 or 155 megabit speeds, but we expect rapidly to move to OC12 and OC48 -- up to 600 megabit and even higher.

Even more important than the speed of the connections, though, are the new protocols that we need to develop to manage the overall capacity in Internet 2. A major innovation in the Internet 2 will be to move away from a best efforts approach to delivering connectivity towards a guarantee of service quality. The protocols to accomplish those objectives are just now being developed, and will need to be experimented with and deployed over the coming years as we move forward in this project. Perhaps the most difficult part will be making sure that the protocols actually communicate from the application operating in a computer on one desktop all the way through the network to an application that's operating on another desktop.

The Internet 2 project had its beginning in a set of discussions that took place in the fall of 1996, where we involved a number of people throughout higher education and focused on how higher education in today's environment could take a hand in addressing the issues that we had jointly discovered as we tried to use the Internet to accomplish the objectives that are critical to the higher education community. Those discussions were carried out under many auspices, but perhaps most importantly by the Networking and Telecommunications Task Force (NTTF) of Educom. In October of 1996, more than 30 universities came together in Chicago and agreed to each commit $25,000 of funding to provide the initial investment needed to start the Internet 2 project, and further, to commit to an investment on their campuses of a half million dollars per year for three years to actually implement the new technology for their campuses. Gary Auguston at Penn State was the leader of this effort and Mike Roberts, vice president at Educom, has served as the chief executive officer of Internet2 as we've moved this project forward. Today, more than 100 research universities are committed to participate.

Shortly after the meeting in Chicago, the Clinton administration made an announcement that they were going to commit $100,000,000 per year at the federal level for a five year period to create what they called "the Next Generation Internet project." (NGI) That commitment provided us with a linkage between higher education's efforts and the efforts of the federal government. Through the combination of higher education on the one hand and the federal government on the other hand, we're now working closely with industry to secure the third critical leg in this partnership, the major industrial commitments to moving this network forward.

As we go forward, we're going to confront a number of issues. First of all, we're going to have to decide just how much formal organization we need in order to move the Internet 2 project along and coordinate it effectively with the federal government's NGI effort. Today, we really just have a task force, led by a group of information technology leaders from higher education. Planning is underway to examine the possibility of creating a new corporation or some other organization that will provide us with a long-term ability to address these issues. More and more, we're coming to the conclusion that we need to focus on that long-term role because we expect the technology we develop in the Internet 2 to be informed by research that's going on in a federally funded effort that we're calling the "Internet 3" today, and we expect the technology developed in the Internet 2 to be available to the broader Internet. And as time goes forward, we expect that all of these technologies will continue to migrate into the commercial sector.

Our most important challenge as we go forward will be maintaining the focus on the applications that we want to use for the benefit of higher education and how those applications have an impact on the broader society. There's always the temptation to get focused on the network for the network's sake, when, in fact, the network is simply a vehicle to accomplish a much broader set of objectives. Finally, we need to make sure that we focus carefully on the relationships that we need to maintain with industry and government. If we develop new protocols as we develop Internet 2, they won't be useful unless they're adopted by industry and unless they become part of a standards environment. If we don't find a way to migrate the technology in a way that industry invests, then we're not ultimately going to be successful in having it available to the broad community so that education can reach out and serve people whenever and wherever they need to learn.

In the final analysis, the Internet 2 project is a project about the future of higher education and how we are able to deliver the value we offer to the society at large. But beyond that, as we learned with the Internet, once we create applications that allow people to work together at a distance and share knowledge and information, the whole world realizes that those applications are critical to their success as well. We very much expect that we'll be able to make a similar contribution with Internet 2 to the one that we made with the original Internet research. We very much intend that higher education's Internet2 effort will make a contribution to society similar to that which came from the original Internet effort.

About CREN

© CREN, 1999

Contact Us