LOCALIZATION ALGORITHM FOR MOBILE TARGET IN WIRELESS SENSOR NETWORKS

摘要

The paper presents a range-based localization method of the mobile target for wireless sensor networks (WSN). The new algorithm MPK filtering is made up of two stages: the position estimation of the mobile target predicted by the modified particle filtering (MPF) and the tracking of mobile target proceeded by unscented Kalman filtering (UKF). It utilizes the merits of these two filtering modules and evades their drawbacks. It not only has the high stability and fast convergence of PF but also has the small computation and high precision of UKF. When emergency events occur or the mobile target is lost, MPK filtering switches between the two filtering modules. The simulation results also show that MPK filtering has good performance in tracking precision but low communication cost.