ECE department receives four NSF grants

By: Melissa Wiegand

Posted: 11/6/09

Stevens' Electrical and Computer Engineering (ECE) department recently received four National Science Foundation (NSF) research grants, totaling more than $1 million for faculty projects in the areas of Wireless Communications, Network Security and Cognitive Radio (CR). Each grant will fund research in the specified project area for three years. Department Director Yu-Dong Yao is pleased with his faculty's achievements and is confident that the projects will "contribute significantly to advances in both CR and wireless network security."

Yao himself received a grant in early September for his project, entitled "xBeam: Cross-Layer Beamforming Against Denial of Service Attacks in Wireless Networks." A common example of Denial of Service (DoS) is the error message one receives when the server of a high-traffic website is overloaded. It is called a DoS attack because the overload is often caused by a hacker, who bombards the server with requests for access. "DoS attacks are a major issue for both commercial and military organizations," says Yao. DoS attacks are easier to generate in wireless networks, since they are less secure and have less bandwidth available than wired networks do. The problem is also more difficult to solve within wireless networks, because one cannot simply isolate the hacker's location or disconnect a cable.

Yao's research focuses on combining various wireless protocol layers to "detect and eliminate DoS attacks." He plans to examine various DoS attacks, develop xBeam algorithms, evaluate the effectiveness of xBeam in deterring DoS attacks, and validate the algorithms using a wireless test bed. He and Research Associate Hongbing Cheng will work together to develop a novel defense approach against DoS attacks in wireless networks. Their ultimate goal is to deter DoS attacks across different networking layers, multiple DoS attacks mounting from different spatial directions, and attacks mounted by malicious attackers who are mobile. Both graduate and undergraduate students will have the opportunity to gain research experience by contributing to this project.

In August, Professor Hongbin Li received a grant for his work on "Data-Driven Adaptive Quantization for Distributed Inference." With the help of Research Associate Dr. Jun Fang, Li plans to develop an integrated signal processing framework for wireless sensor networks. Such systems, composed of a many low-power smart sensors with integrated sensing, processing, and communication abilities, can be deployed close to the phenomenon of interest, thus offering unprecedented sensing opportunities. Wireless sensor networks are useful for civilian applications such as wildfire detection and monitoring of endangered species, as well as military surveillance. According to Li, one problem with these sensors is that "they are powered by batteries, oftentimes irreplaceable, and yet they are expected to run over a long period of time, ranging from days to even months." This calls for efficient signal processing algorithms to reduce the amount of information that needs to be exchanged within the network. "The data compression techniques currently in use have difficulty handling such a distributed sensing task, where the goal is not to recover the original sensor observations but, instead, to extract some useful information, such as to detect the presence of a target, from the observations," says Li. His goal is to "develop novel compression techniques that are specifically designed to solve such underlying sensing or interference tasks." Such a joint design approach is expected to result in significantly more efficient solutions in terms of both power and bandwidth consumption. The project also addresses a fundamental challenge of quantization for distributed inference in a sensor network environment, where the optimum quantizer generally cannot be implemented due to its dependence on unknown parameters associated with the random event being monitored by the sensor network.

These are only two of the four NSF grants given to the ECE faculty here at Stevens in the past few months. Professor Rajarathnam Chandramouli received a grant for his work in "Human Behavior-Inspired Cognitive Radio Network Design," which focuses on the study of communication protocols involving CR networks that resemble human behavior and psychological interactions. Professor K.P. Subbalakshmi also received funding for her research on "Denial-of-Service Attacks and Counter Measures in Dynamic Spectrum Access Networks." The designs proposed through this project will contribute to the development of first responder networks.

The Institute's ECE department has strong research programs and receives contracts from various commercial and government agencies. "We currently have faculty teams actively working on projects for the Air Force and Department of Defense, especially for their Defense Advanced Research Projects Agency (DARPA)," says Yao. According to Yao, the department devotes its research to several focused areas of electrical and computer engineering in order to "present undergraduate and graduate students with the opportunity to become experts in specific areas." Since these four grants are for projects in closely related concentrations, he believes that students will gain a deeper understanding of each research area by participating in the various projects.