A team led by P. Vijay Kumar of the University of Southern California has received the Best Paper Award at the 4th IEEE International Conference on Distributed Computing in Sensor Systems (DCOSS '08).
The paper entitled On the Average Case Communication Complexity for Detection in Sensor Networks considers the problem of detecting certain events via algorithms that seek to minimize the probabilities of mis-detection and of false alarm as well as the energy expended in executing the algorithm. Kumar, Venkatesan Ekambaram, and Tarun Agarwal proposed ways of minimizing the energy consumed by inter-sensor communication, a key sensornet problem. Sensornets are arrays of unsupervised data-collecting devices monitoring variables like temperature, pressure, or seismic activity connected by large-scale networks.
In addition to examining the problem of "clutter" in sensor-network-based distributed intrusion detection, ther researchers present results from a QualNet simulation of the algorithms, including intruder tracking using a naive polynomial-regression algorithm.
Coincidentally, Christos Papadimitriou was a keynote speaker at DCOSS '08, addressing the topic of The Algorithmic Lens: How the Computational Perspective is Transforming the Sciences. Papadimitriou was recently interviewed by Dr. Dobb's Journal.