Multiple Target Tracking in Wireless Sensor Networks Based on Sensor Grouping and Hybrid Iterative-Heuristic Optimization

A novel hybrid method for tracking multiple indistinguishable maneuvering targets using a wireless sensor network is introduced in this paper. The problem of tracking the location of targets is formulated as a Maximum Likelihood Estimation. We propose a hybrid optimization method, which consists of an iterative and a heuristic search method, for finding the location of targets simultaneously. The Levenberg-Marquardt (LM) algorithm is used for iterative search, while the Particle Swarm Optimization (PSO) is used for the heuristic search. We use the maximum sensors separating distance-grouping algorithm (G-MSSD), which was introduced in our previous work, to generate initial guesses for search algorithms. The estimates of both methods are compared and the best one is selected as the final estimation. We demonstrate the accuracy and performance of our new tracking method via simulations and compare our results with the Gauss-Newton (GN) method.