Research On Target Tracking Technology In Wireless Sensor Networks

Target tracking, which is one of the typical and complex applications of wireless sensornetworks，has become a hotspot in recent years. Due to the special nature of WSN, studying suitablemethods for target tracking in WSN is of great significance. In this paper, the algorithms of targettracking in WSN are studied, and a target tracking platform is initially built. The main contents are asfollows:Aiming at the problems of computational complexity and sample impoverishment that exist inparticle filter, a new particle filter based on improved resampling is presented. In this algorithm, thecycle comparison operation is avoided, and more effective particles are generated through mutationaccording to the diffusion factor. Simulation results show that the proposed algorithm has betterperformance both in estimation accuracy and real-time performance.Aiming at some problems exist in self-organizing algorithm, an energy-efficient dynamiccollaborative self-organizing algorithm is proposed, in which the performance index is consideredmore comprehensively. The cluster head is adaptively switched according to the accuracy threshold,so is the selection of task nodes. Simulation results indicate that the proposed algorithm can not onlyreduce energy consumption, but also prolong the life time of the network while meeting the precisionrequirement.A multi-target tracking algorithm based on signal separation is proposed considering the problemof signal aliasing. The algorithm is based on dynamic clustering. The aliasing signals are separated bythe common members and the positions are calculated by cluster heads using improved particlefiltering algorithm. Simulation results show that the proposed algorithm is able to accurately track thetargets which are close to each other.A target tracking platform based on WSN is established, including data acquisition andtransmission based on WSN and monitoring software. The former is used to send and receiveinformation among the nodes. The latter is used to display the position of the target after collectionand analysis of data from base station. Tracking of a moving target is initially realized by test on theplatform.