RSSI-based positioning in unknown path-loss model for WSN

In low complexity networks such as wireless sensor networks (WSNs), localization of nodes becomes essential for many applications. In order to obtain location estimates with minimal cost, the received signal strength indicator (RSSI) based localization has been a preferred approach. For accurate location estimates via RSSI, the knowledge of the network path-loss model is of high importance. Accurate modeling of the channel, prior to system deployment is a laborious task, hence there is a recent trend to estimate the power-distance gradient or pathloss exponent (PLE) alongside the location coordinates. In this study, we put to test our previously proposed linear least squares (LLS) based algorithm using NXP´s IEEE 802.15.4 compliant JN5148 nodes. Experiments are carried out in outdoor and indoor environments. Our results indicate that the performance of joint estimation of the PLE and the location coordinates surpasses that of the results obtained by using the distance estimates as provided by the vendor for positioning.