Real-time path loss and node selection for cooperative localization in wireless sensor networks

RSS-based localization is considered a low-complexity algorithm with respect to other range techniques such as TOA or AOA. The accuracy of RSS methods depends on the suitability of the propagation models used for the actual propagation conditions. In indoor environments, in particular, it is very difficult to obtain a good propagation model. For that reason, we present a cooperative location algorithm that dynamically estimates the path loss exponent by using RSS measurements. Since the energy consumption is a key point in sensor networks, we propose a node selection in order to avoid the use of all the nodes in the network for positioning purposes. Hence, we derive a practical solution tailored to the strict requirements of sensor networks in terms of complexity, size and cost. We present results based on both computer simulations and experimentation with the Crossbow IRIS motes showing that the proposed scheme offers a good performance in terms of position accuracy.