Communication-aware surveillance in mobile sensor networks

In this paper, we consider a surveillance problem, where a number of mobile nodes are tasked with surveying an area for the possible presence of stationary targets and reporting their findings back to a fixed base station, in the presence of realistic fading communication channels. We develop a mathematical framework for robust communication-aware surveillance, in order to survey the environment efficiently while maximizing the probability of connectivity of the nodes to the base station at all the time. More specifically, we show how to design local motion planning strategies that properly integrate both sensing and communication goals. By using Chernoff bound on the probability of detection error, we prove that the motion planning objective can be separated into a sensing function that maximizes the Kullback-Leibler (KL) divergence between the maximum uncertainty state and the current one, and a communication function, that maximizes the probability of being connected. The resulting motion trajectories provide the right balance between sensing and communication objectives and demonstrate interesting tradeoffs. Our simulation results then show the performance of our proposed communication-aware surveillance framework.