Target Tracking Approximation Algorithms with Particle Filter Optimization and Fault-Tolerant Analysis in Wireless Sensor Networks

In order to process target tracking approximation with unknown motion state models beforehand in a twodimensional field of binary proximity sensors, the algorithms based on cost functions of particle filters and near-linear curve simple optimization are proposed in this paper. Through moving target across detecting intersecting fields of sensor nodes sequentially, cost functions are introduced to solve target tracking approximation and velocity estimation which is not similar to traditional particle filters that rely on probabilistic assumptions about the motion states. Then a near-linear curve geometric approach is used to simplify and easily describe target trajectories that are below a certain error measure. Because there maybe some sensor nodes invalid in practice, so a fault-tolerant detection is applied to avoid the nodes’ reporting fault and also improve accuracy of tracking at the same time. The validity of our algorithms is demonstrated through simulation results.