A distributed localization with unknown attenuation coefficient in wireless sensor networks

Recently, a localization attracts much attention in sensor networks. One of major localization algorithms uses Received Signal Strength Indicator (RSSI). This algorithm needs some prior knowledge of an environment and thus requires prior experiments. To avoid troublesome prior measurements, this paper proposes a decentralized localization algorithm with unknown attenuation coefficient. The proposed algorithm has no need to measure radio propagation characteristics, because the algorithm estimates attenuation coefficient besides a source location and a signal energy at unit-distance away from the source. The proposed algorithm has three calculation methods, because they estimate three unknown parameters. We evaluate the computational complexity and localization accuracy of the proposed calculation methods. Finally, we show location estimation accuracies of the proposed algorithm by results of computer simulation and experiment.