Perched conspicuously on the top of Janardhan Iyengar’s bookshelf in Stager Hall is a Mac Classic II, the last black-and-white compact computer produced by Macintosh. The assistant professor of Computer Science smiles as he looks up at the machine.
“It was being used as a doorstop downstairs,” says Iyengar. “One of my students wanted to rip out the insides and revamp it, and we might do it.”
Iyengar might also revamp the architecture of the Internet itself, thanks to a $171,619 grant from the National Science Foundation. Awarded in July, the grant supports Iyengar’s project titled “Tng, a Next Generation Transport Services Architecture.” The project is part of a $499,879 collaborative grant with Yale University, and is funded under the American Recovery and Reinvestment Act of 2009.
“It’s an exciting project,” says Iyengar, who will partner with Bryan Ford, assistant professor of Computer Science at Yale University on the research. “Much of the Internet’s architecture was conceived in a different day, during the 1970s and 1980s, when the Internet was a different beast. The whole face of the Internet has changed since then.”
Iyengar will investigate ways to change Internet architecture, the transportation system through which information travels from computer to computer. The original architecture does not take into account new players known as “middleboxes”—firewalls, performance enhancers and network address translators.
“The Internet was made so everything could be converted into small packets and sent,” says Iyengar, who will hire a postdoctoral research fellow in the coming months. “I can create an e-mail, Web page or voice message and everything gets sent as packets. The Internet doesn’t care what’s going through it. That’s the crux—it shouldn’t care.
“But new players joined the network that had other needs. Organizations joined, and so did spammers, viruses and security hackers. So folks built firewalls to protect users and organizations. The original creators of the Internet didn’t think about firewalls, so the original Internet architecture had no place for them.”
Middleboxes such as firewalls pose problems because they are invisible to computers and users, the endpoints at which intelligence exists. The middleboxes must make assumptions about communication between the endpoints, which often creates significant hurdles in deploying innovative technology at endpoints.
“Our goal is to have the endpoints interact with the middleboxes and make them part of a new architecture,” Iyengar says.
The professor also hopes to have student exchanges between Franklin & Marshall and Yale as part of his collaboration with Ford, with whom he published Breaking Up the Transport Logjam in 2008. The two researchers speak via Skype on a regular basis, technology far removed from the Mac Classic II sitting on Iyengar’s bookshelf.
“We want users to drive innovation,” Iyengar says. “The network in the Internet is becoming more intelligent, and we want to create a new architecture that understands and deals with this intelligence.”