A Study on the Target Tracking in Sensor Networks Based on Improved Particle Filtering

Particle Filter(PF) algorithm is analyzed,and an improved PF algorithm(UPF: Unscented Particle Filter) is discussed in this paper.Importance density function is generated by Unscented Kalman Filter(UKF).Sensor nodes are organized into clusters,UPF and PF algorithms are applied to target tracking in wireless sensor networks(WSNs),to track single target which moves in uniform rectilinear motion.Finally,the comparison of two algorithm's performance in target tracking is presented and the simulation results are also given. From these results we can see that UPF increases the utilization ratio of particles,has better tracking accuracy and better tracking performance.