As part of a newly formed alliance of international scientists, Rensselaer researchers will be exploring advanced technologies for wireless sensor networks in urban environments. The consortium, which is funded through the United States Army Research Laboratory and the United Kingdom Ministry of Defence, will receive up to $138 million over the next 10 years to rethink network computing and communications. Boleslaw Szymanski, professor of computer science and director of the Center for Pervasive Computing and Networking at Rensselaer, will lead a team supported by $1.85 million of the total project funding.

The International Technology Alliance (ITA) in Network and Information Sciences is led by IBM and includes top researchers from industry, academia, and government. The project’s goal is to enhance the ability of coalition forces to make flexible battlefield decisions using secure networks of sensors. But while it is mainly directed toward military applications, the fruits of the fundamental research are expected to find use in a broad spectrum of civilian contexts.

Szymanski is leading one of 12 projects within the ITA consortium. In collaboration with researchers from the University of Southampton, CUNY, LogicaCMG, and IBM laboratories in both the United States and the United Kingdom, he will investigate ways of managing complexity in sensor data infrastructures. “We are going to take what we already know about sensor network protocols and infrastructure and think creatively about the future designs,” Szymanski said. “The goal is to start from scratch and totally rethink every aspect of this technology.”

His team will be looking at networks containing a variety of different sensors, including audio, visual, radar, and chemical sensors. “With information coming from these different sources, we need to know how to make them collaborate to provide the best information while minimizing the chance that they will be detected,” Szymanski said. Each communication within a sensor network uses energy and resources, so the goal is to develop extremely efficient algorithms that keep the number of communications to a minimum.

“In urban environments, you can expect a lot of sensor failures,” Szymanski said. He compares the situation to a crowded lecture hall with one person communicating to many listeners: some people inevitably get stuck sitting behind a column; the speaker turns toward the chalk board and does not project his voice; or everybody in the audience tries to answer a question at the same time, creating chaos. Szymanski’s team hopes to design sensor network algorithms that address similar communication problems among nodes in a wireless sensor network. 

Three Rensselaer students will be working on the project under Szymanski’s supervision: Joel Branch and Mark Lisee, both doctoral students in computer science; and Kamil Wasilewski, an undergraduate studying computer science.

Published September 25, 2006