Performance-Guided Reconfiguration of Wireless Sensor Networks that Use Binary Data for Target Localization

Available resources limit the performance and lifetime of a wireless sensor network. In order to increase network lifetime in the face of depleting resources, networks that can be reconfigured to meet specified performance criteria are of interest. For a target localization scenario using binary sensor data, we present an approach in which network reconfiguration is guided using a performance criterion derived from the Cramer-Rao lower bound on the covariance of the errors in target location estimation. Inactive sensors are strategically placed within the sensor field to provide redundancy. When a sensor fails, sets of inactive sensors are identified as candidates for activation to replace it. The performance expected to be achieved by the activation of each candidate is determined, and decisions about which candidates to activate are made based on criteria of interest for the given application. Example scenarios are provided and some possible applications are discussed