Researchers seek to reduce vulnerability of nation's communications networks

By Kim Gudeman, Coordinated Science Lab
March 24, 2008

R. Srikant

In the event of a terrorist attack or natural disaster, the nation’s communications system would play an integral role in re-establishing order. But what if those networks also sustained damage?

ECE Professor R. Srikant is designing new technology that would enable communications networks to recover in the event of a wide-spread outage. Srikant recently received a three-year, $1.3 million grant from the Defense Threat Reduction Agency, part of the U.S. Department of Defense, to study new algorithms that would allow networks to overcome significant destruction.

Srikant, a researcher in the Coordinated Science Laboratory and Gilmore Family scholar, is the project’s principal investigator. His team also includes two former PhD students, Atilla Eryilmaz and Lei Ying, now assistant professors of electrical and computer engineering at Ohio State University and Iowa State University, respectively.

"Currently, we have no real understanding of what would happen to the communications systems should a portion of the network disappear," Srikant said. "We’re trying to create a system that can repair itself during outages that are impossible to plan for in advance."

The key to the project is creating new algorithms that allow a network to quickly repair itself. The algorithms could be used for any communications infrastructure, including the Internet, telephone and wireless networks. The team also will examine how adaptive re-routing of traffic using mobile wireless nodes can speed up the return of connectivity.

The project will take place in three phases:

• Phase one will include developing a theoretical framework to better understand the fundamental limits of the performance of WMD recovery strategies for a large class of networks, including the Internet, MANETs, and heterogeneous networks.
• Phase two will address the development of adaptive reconfigurable networking algorithms for WMD attack recovery that achieve the fundamental limits derived in step one.
• And finally, phase three will include the design of low-complexity, distributed, and scalable implementations of the algorithms developed in step two.

The most significant challenge facing Srikant and his team is to create distributed algorithms that not only work, but are easy to implement.

"If the solution is so complex, it will take too long to execute," he said. "Our goal is to achieve good performance without sacrificing efficiency."

Editor’s note: Media inquiries should be directed to Brad Petersen, assistant director of communications, at bradp@uiuc.edu or (217) 244-6376.

Index of Articles