The Target Tracking of Wireless Sensor Network Using an Improved Unscented Particle Filter

Based on the analysis of Particle Filter (PF), this paper proposed an improved unscented particle filter (UPF) algorithm by utilizing unscented Kalman filter (UKF) to obtain an importance density function. This algorithm clustered the sensor network nodes through dynamic organization. Moreover, the single target moving uniformly and linearly in the network was tracked by applying the UPF into the target tracking of Wireless Sensor Network (WSN). Finally, a simulation comparison between UPF and PF was conducted using MATLAB. Simulation results showed that the improved UPF was capable of improving the utilization efficiency of particles and presented a stable tracking performance.