Exploring Single Target Tracking Using Acoustic Wireless Sensor Networks

Wireless sensor network is a self-organizing network which consists of large number of intelligent nodes; it has ability of sensing the surrounding cooperatively. Single target tracking is a hotspot in wireless sensor networks, however, existing node group management and target localization algorithms face some difficult problems such as balance between energy consumption and localization accuracy, fault tolerance, target lost etc, especially for high-speed target, it is a great challenge to solve these problems in target tracking based on acoustic wireless sensor networks.After a deep study about the single target positioning and tracking in acoustic wireless sensor networks based on the vehicle tracking, a positioning and tracking algorithm for single target tracking on acoustic wireless sensor networks is proposed. The main contributions of this paper are:1. Proposes a WCA-CFAR target detection algorithm for single node based on CA-CFAR.2. Improves a node self-organizing algorithm for single target tracking of acoustic wireless sensor networks which elects and transfers cluster head in terms of the acoustic energy detected by sensor nodes.3. Proposes a weighted-centroid target localization algorithm based on the conception that the energy received by node is inversely proportional to the square of the distance between the node and target.4. Develops an emulator for single target tracking of acoustic wireless sensor networks in Visual C++ 6.0 and MATLAB 6.5, and extensive simulations have been conducted to characterize performance of the node self-organizing algorithm, weighted-centroid target localization algorithm and trilateration. Simulations results show the node self-organizing algorithm and weighted-centroid target localization algorithm have higher localization accuracy and lower energy consumption.