© 1998 by Oxford University Press
Scanning the intricate structures of a cell's interior through a powerful electron microscope, a scientist in New York spies an intriguing piece of mitochondrial architecture and desires a closer look. Clicking on a computer screen, he directs the microscope to zoom in. Across the continent, at the National Center for Microscopy and Imaging Research at the University of California at San Diego, the microscope responds, sending three-dimensional images processed by the San Diego Supercomputer Center back to New York for study.
Impossible now, but this kind of collaborative effort is what some see as the ultimate role of the Internet in research -- a mechanism for high-resolution graphics, animation, or video that could provide for virtual laboratories and lecture halls, allowing equipment, data, and brainpower to be shared regardless of geographical location.
But the Internet in its current form is choked by commercial sites, millions of e-mails, and a seemingly never-ending swamp of the trivial and the mundane. Researchers, no longer with an exclusive backbone network, must now compete with every personal and business communication as they try to communicate data and collaborate on projects.
New initiatives, however, are bringing together elements from the government, academia, and corporate sectors, to develop faster and more robust Internet technology and to provide a new network backbone dedicated to the research community.
The two most prominent of these initiatives are the government-sponsored Next Generation Internet and the university driven Internet2. The two initiatives overlap in many ways, sharing similar goals, and using similar resources. Both involve partnerships among universities, government, and industry.
Internet2 seeks to accelerate the development of the Internet in academia. The initiative has over 120 member universities and institutions, all dedicated to "the development, evolution and use of advanced networking facilities and applications in the conduct of research and education," according to the Internet2 Website.
In mid-April, Vice President Gore unveiled "Abilene," Internet2's highspeed network reserved for academia. Abilene is to connect at least 100 universities and national labs at speeds that are 100 times faster than today's Internet, which clocks in at 56 kilobits per second. A smaller number of institutions are to be connected at speeds that are 1,000 times faster. It should begin working at the end of this year and be fully operational by the end of 1999.
Greg Wood, communications director for the Internet2 project, said Internet2 is concerned with more than just higher speed connections and data rates, but with the development and refinement of applications using a network. The broad goal is to provide universities and research institutions with the highest-quality networks possible, allowing them to achieve innovations in research, teaching, and learning.
More important than mere speed, though, will be work on faster routing and what is called "Quality of Service." The routers that receive and pass on packets of data along the network are the bottlenecks of the current Internet, and speeding them up will increase performance across the board. "Quality of Service" refers to a priority system for data. Video data places more demands on a network than do transmission of text files and e-mail. Currently, the Internet treats all packets of data equally. One of the I2 projects is to develop programs that are able to recognize what type of data is being sent, and route it to faster or slower connections as necessary.
The Next Generation Internet initiative was announced by President Clinton in October 1996. It's goal is to promote innovation in high-speed networking. High-speed connections will be used to experiment with the next generation of networking technologies and to develop applications useful for everything from medical research to national security. The National Science Foundation, along with the Departments of Defense and Energy, the National Aeronautics and Space Administration, the National Institutes of Health, and the National Institute for Standards and Technology will be key players.
A key element of both initiatives is the very high performance (or very high speed) Backbone Network Service. Run by the science foundation and MCI, the vBNS currently operates at 622 megabits per second, and is expected to reach speeds of 2.4 gigabits per second by the year 2000. Access to the backbone service is reserved for meritorious research applications only, which are determined on a case-by-case basis by a diverse peer review committee. There are currently 92 institutions that have received grants to make connections to the vBNS (77 of them are members of I2), with a future goal of 150.
"Our goal is not necessarily to hook up every institution in the United States," said Elizabeth Gaston, spokesperson for the vBNS program. The primary element in deciding whether or not a project gets permission to connect to the vBNS is whether the proposed research could be done only if access to this kind of speed was available.
Abilene and the vBNS are expected to interconnect at points in the future, although this is not yet certain.
The Collaboratory for Microscopic Digital Anatomy is a prime example of what these initiatives are trying to accomplish. Located at San Diego's National Center for Microscopy and Imaging Research, the Collaboratory is developing a joint computational environment that will offer access to the hi-tech microscopes available at the Center.
The project is focused on the intermediate high-voltage transmission electron microscope, one of the few in the United States made available to the biological research community. This instrument can obtain high quality images from thick subjects and render the results as 3-D images with the help of the supercomputer center in San Diego. The collaboratory team offers remote control of the IVEM instrument to users around the nation. Finding ways to send huge data sets while maintaining interactivity has been the driving force of the software advancement.
"We did it because it wasn't available," said Mark Ellisman, Ph.D., the principal investigator on the collaboratory project. As in the past, necessity has been the mother of invention. Tackling the problems of making an instrument such as the IVEM as interactive and responsive as possible has provided solutions that can be generalized to other imaging equipment.
"Just figuring out that scales to all sorts of projects of this sort," Ellisman said.
David Spector, Ph.D., of Cold Spring Harbor Laboratory, N.Y., is interested in the functional organization of the mammalian cell nucleus, particularly how factors involved in RNA splicing find the RNA.
"It [the collaboratory] has impacted on my research," Spector said. "The equipment there is not accessible to the individual investigator." Originally, the researcher was sent a survey of his sample, and then sent in commands to the IVEM for specific scans, which it executed. The resulting scans were then placed on a web page for the researcher to examine.
But hard work and greater speed has resulted in a new version, not yet available, that allows for the IVEM to be controlled remotely through a web browser. The researcher would be able to see the scans in real time and send orders to the IVEM directly.
"We'll actually see the primary data," said Spector, who has seen a demonstration of the new software. "Now with this set up, one would have the opportunity to see the sample directly [along] with the operator."
Microscopy, molecular modeling, and radiology archives are only some of the uses to which researchers are putting the power of high-speed networks. From biology to cosmology to music, the number of fields that can benefit seem limitless.
"What we're aiming for here is not just an infrastructure, but a test bed," said Gaston. "We want to always stay a couple of steps ahead." As the technology evolves, the government would shift it into the commercial sector. That is what happened in 1995 with the National Science Foundation's NSFnet, which had been dedicated to research and today forms the backbone of much of the commercial Internet. Gaston expects the same to eventually happen with the vBNS and other technologies developed by the I2 and NGI projects. The government provides for the basic research which eventually comes out into the common market. And by that time, they'll be working on an Internet3.
-- Laurent Castellucci
New Initiatives for Researchers Rev Up the Internet
Traffic Bypasses
Collaborative Research
The Test Bed
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||