- The Washington Times - Thursday, August 16, 2001

Sequels rarely outshine the original. For researchers working on Internet2, though, the goal is just that — to improve upon the existing Internet by breaking down technological barriers such as narrow bandwidth and inconsistent service.
Bad sequels often abandon the formula that made the original a success. That isn't the case for Internet2. Its backers are re-creating the atmosphere that helped the first online system grow — banding academic groups together outside the commercial world to foster a nurturing environment where innovation can thrive.
Internet2 is run by a not-for-profit consortium based in Ann Arbor, Mich., that features about 187 U.S. universities. Each partner pays $25,000 in annual dues to keep the consortium, which has a $6 million yearly budget, operational.
Guy L. Jones, chief technology officer at George Washington University, says the original Internet wasn't about the breakthrough technology, but about fostering creativity.
"It was the mind-set that you would provide a communal service that drives innovation," Mr. Jones says. "It was that openness that allowed the growth and sharing to occur."
George Washington is one of several local universities — including the University of Maryland, George Mason, Gallaudet and Georgetown — that participate in the consortium.
At full strength, the speedy Internet2 makes traditional applications look like so many sluggish games of Pong. Nationwide, Internet2 connections share the same consistent, high-speed capacity, which removes the chance of bottlenecks from data moving from one fast network to another, slower one. The network also promises reliable service, unlike the existing Internet, where data routinely breaks up before being received by the user, with incomplete video or audio materials.
That opens up the bandwidth, the amount of data that can travel in a network over a set time period. Bandwidth limitations are seen as a key drawback to the current online world, barring streaming audio and video from becoming the norm.
Potential uses include providing shared access to isolated resources such as telescopes and microscopes, letting doctors advise on surgeries performed hundreds of miles away, delivering on-demand video and audio into homes and offices.
Internet2 relies on existing cables, from fiber-optic wires to ethernet cables. The difference, says Internet2 spokesman Greg Wood, is in the enhanced computers, which send and retrieve the data over such wires.
One of many examples of Internet2 at work was a partnership between Gallaudet and Georgetown universities that delivered video of American Sign Language with synchronized audio, text and graphics. Existing Internet visuals couldn't capture the nuances of hand movement vital to ASL.
Another Internet2 collaboration — between the Pittsburgh Supercomputing Center, Carnegie Mellon University and the University of Pittsburgh Medical Center — used the network to connect and convert scanned data of the human brain almost instantaneously into an animated, three-dimensional image illustrating which brain parts activate during mental activity. Similar projects before would have demanded significant time delays of at least a day or so.
Internet2 relies on two major backbones, Abilene and MCI Worldcom's high-performance Backbone Network Service, known in Internet circles as vBNS. The former is named after the Kansas railhead that opened the Old West for settling.
Speed estimates vary for Internet2, but they start at 155 megabits per second, 100 times faster than an average college connection and nearly 3,000 times faster than a conventional dial-up modem. A megabit is about a million bits, and bits are the smallest units in computer lore, meaning either a zero or a one.
One-fourth of consortium members connect directly to one of the two backbones; the rest tie into giga-pops (gigabit point of presence), high-speed access points scattered nationwide. Giga-pops allow universities to hook up to Internet2 at one point, rather than many, which puts less strain on the backbone.
Mr. Jones says academics leaned heavily on the original Internet from the mid-1980s to about 1994 as a new, revolutionary research tool. Then the commercial World Wide Web erupted, and suddenly the system became too busy for scholars to rely on it for massive transfer of data. Traffic jammed. Government and scholastic researchers, once the sole proprietors of the Internet, were left wanting.
"Suddenly, we were a minority shareholder," Mr. Jones says of the educational community. "At that time, some higher educational groups decided to find out what the next generation will be."
Internet2 was born in 1996 in Chicago, begun by a group of 34 interested universities. Each would pony up the annual membership fee and also pour in thousands more to make sure their campuses could handle the data streams to be generated.
In the past few years, companies flush with dot-com venture capital committed themselves to researching new, better ways to leverage the Internet. Now, with many Internet-based companies as extinct as the raptor, funding for such groundbreaking research has dried up.
Enter the Internet2 consortium, where such research is encouraged.
Mr. Jones sees Internet2 sparking creativity and using the Internet concept beyond what has been done before.
As with the beginning of the first Internet, not much is known about precisely what the new system will bring to the public. No one knows exactly how the new service will be folded into the existing Internet, or when.
"The difference is, now it's widely understood that the Internet is a very valuable vehicle for business," Mr. Wood says.
Mr. Jones, though, says the researchers knocking down barriers won't be the ones to take advantage of such leaps.
"Turning a good idea into a sustainable business model is not something higher education will do," he says. About 60 for-profit companies, such as Cisco and IBM, are committed to Internet2 and certainly will have a say in how the technology eventually reaches the masses.
Internet2 has corporate partners that can advise as to the commercial implications. In three or four years, new computers will be adaptable to Internet2 technology, Mr. Wood says. Older computers also should be adaptable to the new system.
Internet2 researchers also work closely with the Next Generation Internet (NGI) initiative, a government-funded project with similar goals. A key difference between the two is that NGI focuses mainly on the needs of federal groups such as the Energy Department and the National Aeronautics and Space Administration. Because university research occasionally overlaps with federal work, the two groups share a working relationship.
Anthony Conto, executive director of MAX, or Mid-Atlantic Crossroads, says Internet2 requires its consortium members to continually upgrade its services. MAX is a regional network founded by Georgetown, George Washington, the University of Maryland and Virginia Tech that connects to Internet2.
Currently, MAX has an OC12 bandwidth, or 622 megabits per second to Abilene. In the next few weeks, the system will be upgraded to OC48, or 2.4 gigabits per second. A gigabit is about a billion bits.
"This allows people's imaginations to run. If there's no capability, then it's only fantasy," Mr. Conto says. "Having a high-performance network allows your imagination to be creative."
One example of such ingenuity at the University of Maryland is a project letting geography students collect data from a NASA satellite that maps the vegetation canopy of the world's forests. Students can control some of the satellite's data collection and analysis. Mr. Conto says such broad efforts likely were impossible without Internet2 technology.
Another project in development at the University of Maryland is teaching courses in real time at both its College Park campus and a university in Japan.
MAX serves as a regional giga-pop, allowing its 18 members to connect as one unit to the backbone service.
Ardoth Hassler, associate vice president for university information services at Georgetown, says the costs to rev up to Internet2 speeds can be daunting.
"If you had to cost-justify it when you did it, you probably couldn't afford it," Ms. Hassler says.
But the benefits already are paying off. Georgetown researchers are in touch with the National Institutes of Health and National Cancer Institute, among other groups, sharing information online. Ms. Hassler says Internet2 use, as a whole, also is increasing on campus. The university commits from $250,000 to $300,000 annually toward Internet2 efforts.
She sees the technology offering on-demand learning, such as a student accessing video of Franklin D. Roosevelt's Dec. 8, 1941, "Day of Infamy" speech at 2 a.m. from a laptop.
"It's an essential research tool. It's gotta be there," she says.
Bandwidth problems could be eradicated over the next few years courtesy of Internet2 technology. Mr. Conto says researchers are seeking to transmit data over wavelengths, or color frequencies, to allow for more streams of information to travel along wires.
"Internet2 is an envelope-pushing project, and it always will be," he says.

LOAD COMMENTS ()

 

Click to Read More

Click to Hide