Target Tracking Based On Distributed EM Algorithm

Wireless Sensor Network is composed of a set of small wireless sensor which can sense and monitoring environment,and it is a sensor network system that communicates with others sensor to get real time target motion state.Its application has been expended from the military fields,such as education,medical treatment,transportation,environmental monitoring and so on,which has made great contributions to the development of the society.But in WSN,the sensor node energy and bandwidth are limited,and in WSN target localization and tracking have 80% energy is used to mutual communication between the sensor nodes.Thus it is very important that adopt effective communication strategies and data processing methods to reduce the energy consumption of sensor nodes and prolong the lifetime of the whole network.In the process of target monitoring and tracking,all the sensor nodes in the monitoring area can improve the accuracy of target tracking by exchanging information with each others,but the energy consumption of the nodes is very large.Therefore,the use of distributed EM(Expectation Maximization)algorithm based on particle filter for target tracking is very suitable.This method not only can effectively track the target,but also can reduce the communication of sensor nodes and prolong the service life of the whole network.The premise of target tracking is to locate the target,using centralized EM based on Gauss mixture algorithm for target location is a good choice,and the simulation results show that the centralized EM algorithm has good positioning performance.There are two kinds of distributed communication strategies based on distributed EM algorithm: incremental strategy and consensus filter strategy.Through the analysis and comparison of the network topology and communication mode of the two distributed communication strategies,it is found that the consensus strategy has good performance and practicability,but with the increase in the number of nodes in the entire network,communication traffic is very large,it will bring more energy consumption.Therefore,we adopt the clustering grid communication strategy.The strategy reduces the communication energy of the node by changing the network topology and communication method of the consensus filter strategy to reduce the network energy consumption and prolong the service life of the network.Finally,we simulate the target tracking effect of the distributed EM algorithm based on particle filter,the results show that the algorithm can effectively track the target.At the same time,compare with consensus filter strategy and cluster grid strategy about communication traffic of single nodes,the results show that the proposed strategy can reduce node energy consumption to a certain extent.