Computer scientist Biplab Sikdar is developing “immunization” techniques to safeguard against computer viruses. By Karen de Seve Printer-friendly PDF version Computer scientist Biplab Sikdar says he is used to waiting in line. “In India you wait in line for anything,” he explains. “There are lines everywhere over there, but here they are practically nonexistent.” Hailing from Calcutta, he insists that he doesn’t mind the wait, but when it comes to the Internet, Sikdar won’t tolerate delays. Instead of waiting in line, Rensselaer’s young assistant professor of electrical, computer, and systems engineering now spends his time online, perusing cyberspace for ways to keep the information superhighway free of roadblocks caused by malicious worms and viruses. These days, computer infections reach plague proportions, spreading from one host to another much more swiftly than the common cold. “Virus attacks are on the rise and the amount of damage is becoming larger,” says Sikdar, who has been a victim himself. “People are developing better viruses at the same time more people are conducting more business on the Web.” Taking a cue from biology, Sikdar uses mathematical epidemic models to uncover how these electronic pests affect a network’s ability to process and route data from one user to another. “I started looking at how a virus spreads,” he says. “Of course models in biology exist which talk about the spread of epidemics in a human or animal populations, and so I began looking into how these models could also apply to virus attacks on the Internet.” More specifically, Sikdar is investigating how these vandals propagate, where they are headed, and how they behave, with the goal of outsmarting and destroying the vermin before they do harm. The project, called “Biologically Modulated Models for the Dynamics of Computer Networks: Performance, Growth and Pathological Conditions,” attracted a five-year grant totaling $402,682 earlier this year from the National Science Foundation’s CAREER program, which supports researchers who are just gaining their footing in a particular field. “We need to better understand how these large-scale systems operate and he’s applying biological models to these big computer networks,” says Carl Landwehr, who oversees NSF grant applications under the CyberTrust heading. “It’s a rich enough area for Professor Sikdar to work in for several years to build a career.” Career Course Nearly 30 years old, Sikdar is in his third year as an assistant professor in Rensselaer’s Electrical, Computer & Systems Engineering department. Modeling the spread of computer viruses is an outgrowth of his work analyzing Internet traffic, which earned him a Ph.D. at Rensselaer in 2001. “The rate at which people get on the Internet has grown exponentially in the past decade,” he says. “I am trying to understand how to modulate the traffic, also called the data, so things don’t get overrun.” Sikdar says that if a router gets overloaded with information, it might just drop the data instead of processing and sending it. Computer viruses and worms sent through the Internet must also pass through regional and local router hubs on their way from their creator to their targets: usually PCs running Windows. Unlike antivirus and antiworm programs, which repair infected computers, Sikdar’s stealth profiling strategy will intercept Internet pestilence entering routers en route. “Viruses interfere with the way the routers work,” he says, explaining that he simulates virus attacks to study router behavior. “We are trying to find an invariance with the primary goal of preventing routers from going crazy when a virus comes along. It’s a preventative, rather than reactive, measure.” In addition to studying viruses under his NSF grant, Sikdar also will analyze life expectancy of pervasive wireless networks comprising small battery-powered sensors. “These could be used for civilian as well as military applications,” he says, describing their functionality as a form of “swarm intelligence.” Sikdar says ant colonies provide a good model for such networks, but he admits he hasn’t started an ant farm in his office. Goal Tending Although masked by his demure demeanor, Sikdar possesses the keen mind of a goal-oriented scholar, as his Rensselear resume reveals. “He finished his Ph.D. at RPI in a matter of three years, and that’s definitely impressive,” says Shivani Deshpande, one of five graduate students in Sikdar’s computer lab in the Jonsson Engineering Center. “He’s a workaholic, which is sometimes useful if I need some help in the middle of the night. I know he’s working around the clock.” Scientific endeavors aside, Sikdar escapes his heavy workload from time to time. “He likes to play computer games, and he has music going on his laptop all the time,” Deshpande reveals. But a soccer match is the one thing that’s sure to get Sikdar out of the office. He met his childhood ambition of going to a World Cup match in Japan. Yet as his Web site humorously attests, he and his cohorts were unable to replicate the results on Rensselaer turf. Perhaps some simulations are best left to the computer lab… More Information: |
||||||||||
|
||||||||||||
| Copyright ©2004 Rensselaer Polytechnic Institute. Published quarterly by the Rensselaer Office of Communications in collaboration with the Office of Research. |