Research On Cooperative Localization In Wireless Sensor Network

Location information is critical to the surveillance of the network in wireless sensor network. The location of the event or the location of the node from which information is acquired are important information in sensor network monitoring. Monitoring without location information is usually meaningless. Therefore, estimating the location of the event or the location of the node is the fundamental function of WSN. The problem of locating nodes in sensor network has become a hot issue. TOA-based cooperative localization technique has been investigated in this work.In WSN, the accuracy of cooperative localization is related to many factors. In this work, Fisher Information Matrix in Cramer-Rao Lower Bound has been used as tool for analysis. Several factors related to accuracy are analyzed. In addition, the benefit from cooperative localization compared with non-cooperative localization is proved.A novel distributed, high-accurate and low complexity localization algorithm is proposed. Firstly, the thought from DV-hop is employed. By counting hop from mobile node to fixed node, approximate distance can be estimated. After distances to all fixed node are estimated, the approximate location can be calculated, which is a good initial value for iterative method. Then every node is assigned a weight value according to their CRLB and hop counts to nearest fixed node. When solving nonlinear least square, the item with the largest weight will have the highest priority to have its error minimized. Simulation shows that the algorithm converges quickly, and global optimum solution is guaranteed. In general situation, the proposed algorithm performs better than currently best SDP method.