Decentralized Joint Positioning and Path-Loss Exponent Estimation for Wireless Sensor Networks

Abstract

Wireless sensor network (WSN) localization has received a great deal of attention in recent years. Using the information of received signal strength (RSS) measurements, many decentralized localization methods have been proposed to decrease the sensor node’s energy consumption. Generally, the large-scale behavior of the signal received at each sensor node can be modeled as the distance-dependent path loss model in indoor environment, thus the performance of RSS-based localization methods will be affected by the estimation accuracy of path loss exponent (PLE). In this thesis, we propose a joint positioning and PLE estimation method based on recursive weighted least squares (RWLS) optimization for WSNs. We first formulate two recursive-in-time cost functions from a sum of the local cost functions and then prove that the minimization of the recursive-in-time cost functions can be realized by the joint positioning and PLE estimation method. In other words, our method is derived from the minimization of the recursive-in-time cost function and then realized in an iterative, decentralized manner. In the proposed method, the PLE and the target location are computed iteratively by taking a weighted average of the local estimates according to the reliability of the sensor nodes, and the reliability is concerning about the PLE and distance estimates. During each iteration, a sensor node computes a new PLE estimate and a new location estimate by its own observation and the most update passed over by the sensor node responsible for the pervious iteration. The newest estimates are circulated and updated among the sensor nodes in the localization procedure. Computer simulation results demonstrate that our method has better location accuracy than the recursive weighted least squares (RWLS) method.