用于目标跟踪无线传感器网络的动态分簇算法

Dynamic Clustering Algorithm for Targets Tracking in Wireless Sensor Networks

针对跟踪精度和网络能耗的问题,提出了一种用于目标跟踪无线传感器网络的基于预测的动态分簇算法。把目标运动过程看作是高斯马尔可夫过程,根据目标历史轨迹,估计下一时刻位置坐标和运动速度,然后基于估计结果优化选择分簇的簇头和簇成员,形成一个动态分簇来实现目标跟踪。仿真结果表明:该算法使目标跟踪有较好的跟踪精度,能有效均衡网络能耗,延长网络寿命。

An algorithm based on prediction is proposed in this paper for target tracking in wireless sensors networks.The mobile movement is regarded as Gauss Markov progress and the cluster estimates the target＇s trajectory at the next time by history.And then clusterheads and nodes are selected for dynamic clusters.The simulation studies show that this approach has better localization accuracy and can balance networks＇energy consumption.