Sub-band Energy Based Collaborative Target Localization in Wireless Sensor Network System

In this paper, sub-band energy based collaborative target localization in wireless sensor network is presented. Three steps including target detection, target location candidates estimation and target location identification have been performed to realize the task of target localization. In the step of target detection, each sensor node detects the targets by filtered total acoustic energy based constant false alarm (FTE CFAR) detector, acoustic sub-band energy based constant false alarm (SE CFAR) detector and multi-modality energy based constant false alarm (MME CFAR) detector respectively. Region based target detection fusion is then performed based on all node detection results from the three detectors. Sub-band acoustic energy based algorithm is performed as the second step to estimate the target location candidates when targets are detected in the region. Finally, sequential bayesian estimation is used to identify the most possible target location from these candidate locations. Experiments have been conducted. Results show that this new approach is accurate and robust to the strong background noise. Besides, communication bandwidth and computation burden are not high compared with the other approaches.