Analysis of linear least square solution for RSS based localization

Positioning of wireless devices has received a great deal of interest from researchers in the last decade. In order to locate nodes in low complexity and power efficient networks, the received signal strength (RSS) based positioning systems have been the center of focus. RSS based localization needs no additional hardware and hence is favored for low complexity and cheap localization networks. A major source of error in RSS location estimation is due to shadowing effects in multipath wireless channels. In this paper we analyze the performance of RSS location estimator based on the linear least square approach. We derive expressions for mean square error (MSE) and bias of location estimates. The theoretical analysis is compared with simulation results and it is observed that the analysis accurately predicts the performance of the location estimation. We also discuss the impact of reference node placement on estimation bias.