On the global convergence of a class of distributed algorithms for maximizing the coverage of a WSN

We consider the problem of finding a periodic schedule for the wake-up times of a set of nodes in a Wireless Sensor Network (WSN) that optimizes the coverage of the the nodes are deployed on. An exact solution of the problem entails the solution of an Integer Linear Program and is hardly viable on low power nodes. In this paper, we study the convergence of an efficient decentralized algorithm for node scattering by casting the problem into one of asymptotic stability for a particular class of linear switching systems. We present asymptotic stability results for generic WSN topologies and an application of the algorithm to the coverage problem to show the effectiveness of the proposed solution. ©2009 IEEE.