Decentralized source localization and tracking [wireless sensor networks]

This paper describes a new approach to the source localization and tracking problem in wireless sensor networks. A fast, easy-to-implement algorithm for localizing a source using received signal strength measurements is presented. The algorithm is based on incremental subgradient optimization methods. Using theory on the convergence rates of these methods, we characterize the amount of in-network communication required to achieve an accurate estimate of the source's location. In comparison to other localization and tracking algorithms described in the literature, the amount of communication (and thus energy and bandwidth) used by our algorithm is much lower than that used by other schemes, especially as network size grows.