基于无线传感器网络动态簇的目标跟踪

着重研究了基于动态簇目标跟踪的两个关键环节：动态簇的组建和目标状态估计。在动态簇组建方面，提出了综合考虑节点感应信息和节点电量的簇头竞选方案；在目标状态估计方面，定义了包含信号测量强度的最近点( CPA)事件，并提出了基于该CPA事件的新型估计算法。结合这两个关键环节，本文给出了基于动态簇的无线传感器网络目标跟踪过程，包括动态簇的初始化、簇内数据聚合、目标状态估计、动态簇的重组、目标丢失检测以及目标恢复等。仿真结果表明：本文建立的目标跟踪系统具有失效概率低、网络寿命长和跟踪精度高等优点。

Target Tracking with Dynamic Clusters in Wireless Sensor Network

The two key issues in target-tracking applications with dynamic clusters, i.e. cluster con-struction and target estimation, were emphatically studied. For the former issue, a voting approach was developed for the cluster head selection by taking both the node sensing information and the node power into consideration. For the latter one, a new closest point of approach(CPA) event incorporat?ing the measured signal strength was defined and a target estimation algorithm was presented. A dy?namic cluster-based target-tracking process was described, wmch contains cluster initialization, in-clus?ter data aggregation, target estimation, cluster reconstruction, target loss checkup, target recover, and so on. The simulation results show that the presented tracking system has low failure probability, long network life and high estimation accuracy.