Sensor-Target Assignment Strategy for Multi-Target Collaborative Tracking under Low Detection Probability

Abstract

Aiming at the difficulty of sensor resource assignment in multi-target collaborative tracking under low detection probability, a strategy for sensor-target assignment is proposed. Firstly, in order to solve the problem that tracking precision of covariance assignment strategy is decreased due to serious data loss under low detection probability, the nearest sensor-target assignment strategy which gives priority consideration to detection probability is proposed. Secondly, in order to solve the problem that the sensor network has low node density and limited number of nodes for tracking under low detection probability, the nearest sensor-target assignment strategy for mobile sensor nodes is proposed by repositioning the mobile sensor nodes. Finally, in order to solve the problem that the nearest sensor-target assignment strategy could lead to low tracking precision and the covariance assignment strategy could lead to data loss, the nearest-covariance assignment strategy is proposed, which considers both the detection probability and tracking precision to repeated nodes. Simulation experiments are used to verify the validity of the proposed algorithms, which indicate that the algorithms can distinctly improve the target detection probability of sensor network as well as the precision and stability of target tracking.