Moving Target Tracking in Three Dimensional Space with Wireless Sensor Network

In three-dimensional space, current target tracking algorithms based on wireless sensor networks are mainly non-iterative and operated with only current measurement result. A typical example is the least square algorithm. Compared with iterative algorithms which use historical information, such as extended Kalman filter, non-iterative algorithms always achieve lower accuracy but can avoid the dependence upon prior knowledge of system noises. In this letter, we firstly proposed a minimum residual localization algorithm based on particle swarm optimization, which is a non-iterative algorithm. Then, a data-fitting strategy is adopted to convert the non-iterative algorithm into iterative one without knowledge of system noise. Hence, the historical information can be used to improve the accuracy of non-iterative algorithm significantly. Simulation results show that the proposed algorithm acquires better localization result with strong adaptability for different motion.