Error-Ellipse-Resampling-Based Particle Filtering Algorithm for Target Tracking

In this paper, an error-ellipse-resampling-based particle filter (EER-PF) algorithm is proposed for target tracking in wireless sensor networks. In order to improve the effectiveness of the particles, in the process of resampling, the error ellipse of different confidence levels is established according to the error covariance matrix of particles. The particles are divided into different levels based on the geometrical position, and then the particles are screened and optimized. The effectiveness of the proposed method in a cumulative error optimization was verified by comparing with the performance of posterior Cramér-Rao lower bound (PCRLB). Experimental results show that the proposed algorithm can effectively solve the problem of sample degeneracy and impoverishment, and has higher positioning accuracy.