Target Tracking Algorithm Research Based On Underwater Wireless Sensor Network

Underwater sensor network(UWSN) target tracking is one of the important frontiers of UWSN application research. Underwater target tracking has very important applications in the fields such as defense, resource exploration. UWSN target tracking involves target identification and detection, signal processing algorithms, sensor networks, underwater acoustic communication technology. This paper studies a distributed Kalman filter tracking algorithm based on UWSN mesh topology and valid node selection scheme, it also studies a distributed particle filter tracking algorithm based on UWSN cluster topology and a track-before-detect particle filter algorithm based on weak target detection. This paper describes the principle and experimental results of these algorithms.The main work of the thesis:1、 Summarizes the Bayesian algorithms for UWSN target tracking, analyses the Kalman filter in distributed network applications combining with the UWSN mesh topology model,using effective node selection strategy to choose effective sensor nodes that reduces the number of working sensors. Simulation experiments of this algorithm are performed.2、 Analyses the UWSN distributed communication model for tracking algorithm combining with the UWSN cluster topology model. Introducing the underwater target tracking algorithm performance evaluation system. Analyses the disadvantages of particle filter resampling process, and using an improved resampling algorithms.3、 Introducing the target tracking thinking of track-before-detect for the strong underwater background noise, discuss the track-before-detect principle in detail, describe the particle filter tracking algorithm based on track-before-detect,analyses the weak target detection problems of underwater sonar sensor. Simulation analysis are performed with strong clutter noise and the detection characteristics of sonar sensors.