Performance of target search via track-before-detect for distributed sensor networks with heterogeneous sensors and imperfect communication links

Target detection is a surveillance problem of practical importance that is well suited to wireless sensor networks. In this paper, we study the target search via track-before-detect process for moving targets in large area utilizing a network of distributed sensor nodes. Specifically, we extend the track-before-detect process to heterogeneous sensors with imperfect communication links between the sensor and fusion center. We consider two different models: (1) heterogeneous sensors with different probability of detection and perfect communication links; and (2) heterogeneous sensors with imperfect communication links. To analyze the performance of searching process for the system, we derive the expressions for the probabilities of both successful search and false search. We also investigate the strategy to optimize the system by obtaining the minimal false search when meeting the requirement of the successful search performance.