Joint Detection and Tracking via Path Planning in the Mobile Underwater Sensing Network

Abstract

Due to the complexity of the underwater acoustic environment, it is a great challenge to jointly detect and track a target by passive methods. A path planning algorithm is presented to improve the performance of the passive detection and tracking for the mobile underwater wireless sensor network(MUWSNs). The target state is modeled as a random finite set, and the Bernoulli filter is utilized to detect and track the target. Moreover, the filter is associated with the path planning algorithm based on two reward functions: the Rényi divergence and the Fisher information gain. The simulations are performed with some practical underwater acoustic environmental parameters, and the results show that the proposed path planning algorithm can improve the detection and tracking performance of the MUWSNs. For the comparison of two reward functions, the results indicate that the algorithm with Rényi divergence has lower position errors.