A Mobile Localization Method for Wireless Sensor Network in Mixed LOS/NLOS Environments

Abstract

In Wireless Sensor Networks (WSNs), Non-Line-of-Sight (NLOS) propagation becomes the main challenge of mobile nodes localization. In order to solve this problem, this paper presents a Square Root Unscented Kalman Filtering-Convex Optimization (SRUKF-CO) method. The Square Root Unscented Kalman Filter (SRUKF) is first used to correct the measuring distance of the LOS and NLOS mobile nodes without prior information on the statistical properties of the NLOS error, which is independent of the physical measuring method. Then, the maximum likelihood localization method is used to estimate the position coordinates. Finally, the limit conditions are determined, and the convex optimization method is adopted to further reduce the NLOS errors. The simulation results show that this method has higher accuracy of positioning compared with unfiltered direct positioning (NF), Kalman Filter (KF) and Particle Filter (PF) under mixed LOS / NLOS environment, and it is robust to NLOS errors.