On Target Tracking In Wireless Sensor Network

Wireless sensor network is one of the ten changing the world technologies and one of the most influencing technologies in the 21st century. For the wide applications of wireless sensor network, target tracking is its basic capability. Wireless sensor network is very suitable for target tracking. This dissertation deals with the problem of target tracking in wireless sensor network, the main contributions are as follows.1) The TDoA and FDoA measurements are nonlinearly related to the target pa-rameters. On the basis of the principle of total least square, the total least square solution for target tracking using TDoA and FDoA is provided. And the Cramer-Rao Lower Bound for target tracking using TDoA and FDoA is studied. Simulation re-sults prove that the total least square solution can attain the approximate Cramer-Rao Lower Bound. Additional, the Simulation experiments has also been carried to study the relationship between the number of sensor node and Cramer-Rao Lower Bound for target tracking using TDoA and FDoA. The analysis results show that more sensor nodes do not imply the high tracking accuracy. And we can adjust the average distance between and average velocity of sensing nodes to improve the tracking accuracy.2) Compared with the computation complexity and extra requirements for wireless sensor network of TDoA and FDoA methods, binary sensor network provide a simple frame for target tracking. On the basis of problems existing in the available algorithms, distance weighted target tracking algorithm is presented. According to the principle of distance weight, the prediction based distance weighted target tracking algorithm is proposed. The performance of distance weighted algorithm is better than the available algorithms and can suit for real-time target tracking, no matter what movement model is adopted by the target. The prediction based distance weighted algorithm not only avoids the transmission of distance information, but also has the same error interval with the distance weighted algorithm.3) Connected coverage is one of the evaluation parameter for target tracking. Three different connected coverage scenarios are given. In the scenario of connected coverage, the performance of the distance weighted algorithm is analyzed. In the anal-ysis, we find that the performance of distance weighted algorithm is closely related to the form of distance function. The optimal distance weighted function is found out and the simulation results show that the performance of optimal distance weighted func-tion is outperform that of the same distance weighted function in all kinds of sensing fields. In addition, the performance of prediction based distance weighted algorithm has been improved by the application of the optimal distance weighted function.4) Euclidean distance or Euclidean norm is equivalent to the 2-norm, which is ap-plied widely in researches and real applications. In order to improve the performance of distance weighted algorithm for binary sensor network, the optimum p-norm weight is presented on the basis of the principle of p-norm. The optimal p-norm weight is determined by simulation. The simulation results show that the optimal p-norm weight not only improves the performance of distance weight in small error interval, but also reduces the probability of tracking error in big error interval. And the performance of prediction based distance weighted algorithm has also been improved by the applica-tion of the optimal p-norm weight.