Abstract: Node localization is an important problem for wireless sensor network, however most localization approach can not evaluate position accuracy for each sensor node. This paper proposes a range-free localization method which bounds actual sensor position within a region, and takes the area of this region as an evaluation criterion of localization accuracy. Non-convex of the network is then exploited to improve localization accuracy. Simulation results show that 90% sensor nodes can be well positioned in a network of 250 nodes and 20% known nodes if non-convex constraints are used.

Key words: wireless sensor network, node localization, non-convex constraint information

摘要： 节点定位是无线传感器网络中的关键性问题，大多数定位方法无法评估每个节点的定位精度。该文提出一种无需测距的定位算法，将传感器节点的真实位置限定于一个区域中，使用该区域的面积评估传感器节点的位置精确度，利用网络中的非凸约束提高定位精度。仿真结果表明，在使用非凸约束的情况下，对于节点总数为250，20%为已知位置节点的传感器网络来说，90%的节点能较好地被定位。

关键词: 无线传感器网络, 节点定位, 非凸约束信息

CLC Number: 

• TP393.03