UNIVERSITIES AND COMPANIES ARE TEAMING UP TO DEVELOP NETWORKS 100 TIMES FASTER THAN TODAY'S.

Research scientists have an insatiable need for speed—at least when it comes to Internet access. Faster transmission times translate into the ability to share information and collaborate on huge projects, such as battlefield simulations, weather forecasting, folding proteins for cancer drug research, computational fluid dynamics, and finite element analysis. Broadband is only the beginning. In the world of home computer users, we talk about speeds measured in megabits per second. But scientists are already lusting after Internet access speeds 100 times faster, measured in multiple gigabits per second. Even today's most advanced Internet technology may not accommodate wide use of these kinds of capabilities, which is why researchers aren't waiting for telecommunications companies to build the infrastructure for them. Universities are teaming up with technology companies, such as Cisco, to develop their own lightning-fast national fiber- optic networks, with a spirit reminiscent of the Internet's earliest days.

It's not surprising that today's scientific community is taking the initiative when you consider the Internet's history. In 1969, the first four nodes on the Department of Defense's Advanced Research Projects Agency's ARPANET, the initial incarnation of the Internet, were research institutions (the Network Measurement Center at UCLA, Stanford Research Institute, UC-Santa Barbara and the University of Utah).

One new project in keeping with the collaborative academic spirit of the Internet's roots is the National LambdaRail (NLR), a consortium of U.S. universities and technology companies that is creating a national fiber-optic networking infrastructure using dense wave division multiplexing (DWDM) and 10 gigabits per second ethernet (LAN PHY) technologies. The idea is to foster networking research and next generation network-based applications in science, engineering, and medicine. For example, using NLR, scientists at universities and Department of Energy (DOE) labs around the country will be connected to the $1.4 billion Spallation Neutron Source, an accelerator-based neutron source being built by six DOE labs in Oak Ridge, Tenn., expected to be completed in 2006. As the creation of NSFnet led to commercialization of the Internet, one goal of the NLR is to enable the technology transfer into commercial development and the creation of new markets.

The beauty of the NLR is that it puts the control of experimental network infrastructure in the hands of the nation's scientists and researchers. That the NLR is user-owned makes it, and other networks like it, inherently threatening to the existing telecom infrastructure. "Such networks represent an inherently disruptive innovation," states the IEEE-USA's position paper "Accelerating Advanced Broadband Deployment." "Their deployment is, therefore, having and is likely to continue to have direct impact on the business models of current telecom providers." But, they argue, nothing less than the country's competitiveness is at stake.

"Our telecom infrastructure is falling rapidly behind the rest of the world, especially behind our most likely competitors: Japan, China, and South Korea," explains Alan McAdams, professor of managerial economics at Cornell University and the architect of the position paper. That's why it's important for the United States to adopt policies that will ensure that user-owned fiber-optic networks "are given a fair marketplace opportunity to prove themselves on their merits as contributors to enhancing the country's national productivity, homeland security, and international competitiveness." Look no further than the Internet itself for proof that the strategy works.

Mary Kathleen Flynn has covered technology for more than 15 years for a variety of media outlets, including Newsweek, the New York Times, U.S. News & World Report, CNN, and MSNBC.