Research On Target Tracking Algorithm In Wireless Sensor Networks

The target tracking technologies in military and civilian fields are both important and valuable for theoretical and practical research, and the emergence of wireless sensor networks provides a new solution for target tracking. In wireless sensor networks, the nodes are typically of small physical dimensions and low cost, the communication between nodes adopts wireless technology and realizes the deployment very flexible, these traits make the sensor networks have a special advantage in the application of target tracking. At the same time, the sensor nodes are operated by battery power, making energy consumption a major concern, the communication between nodes may become unreliable, even the nodes failure happens commonly during monitoring, All of these limitations bring about unique challenges that are not encountered in conventional target tracking systems.This thesis mainly studies the target tracking algorithm in wireless sensor networks, and proposes a new tracking algorithm based on regional beacon selection to solve the problem of low localization accuracy and high energy consumption. The algorithm takes the work-sleep status of the sensor nodes into account as well as the topological relationship between nodes. The simulation results show that the proposed location algorithm has excellent performance in terms of tracking accuracy, energy consumption and stability. The main contributions of the thesis are as follows:Firstly, introduce the basic concepts of wireless sensor networks and its characteristics, and describe the process of target tracking and previous achievements.Secondly, research on the optimization strategy of coverage problems in wireless sensor networks, and propose a sleep scheduling scheme based on Voronoi diagram, simulation results show that the scheduling scheme is efficient and energy-saving.Thirdly, analyze the topological relationship between nodes, and propose a beacon selection method to solve the problem of low accuracy in traditional location algorithm. Simulation results validate the effectiveness in respect of location accuracy and stability; it also provides a feasible method for selecting optimal beacons to reduce the localization error. Fourthly, propose a new tracking algorithm based on regional beacon selection, which combine beacon selection method with Voronoi diagram sleep scheduling scheme. Comparing with the general approach, the simulations demonstrate the proposed algorithm have great improvement in terms of tracking quality, energy consumption and stability.Finally, the work in this thesis is briefly summarized and some future work is presented.