Abstract:

In order to achieve high tracking precision of moving target positioning in wireless sensor networks (WSNs) with low energy consumption and bandwidth consumption, we propose a WSNs Bayesian localization and tracking algorithm based on Kullback-Leibler divergence filtering. Firstly, we use the Gaussian and Wishart distribution without considering the speed limits and restrictions of movement direction to construct a mobile localization Bayesian state evolution model for WSNs, as well as a moving target positioning observation model based on the path loss model. Secondly, we use the Kullback Leibler to construct a divergence filtering error calculation model, which can estimate the position of the goal of the mobile node through activating the surrounding nodes. The recursive probability calculation process we designed integrates prediction and updates process, and realizes synchronous target localization and tracking. Simulation results show that the proposed model has advantages in tracking accuracy and resource saving.

Key words: Kullback-Leibler bifurcation, divergence filtering, wireless sensor networks(WSNs), moving target, localization and tracking