一种基于锚节点分簇的传感器网络节点定位方法

Node localization method based on anchor clustering in wireless sensor networks

提出一种基于锚节点分簇的节点定位方法,该方法以锚节点为中心自适应构建成簇,并将处于多个簇重复覆盖的节点作为伪锚节点。在测距方法构建节点间距离矩阵基础上,结合多维尺度分析和锚节点位置信息实现了对节点进行相对位置定位,并通过伪锚节点与锚节点位置信息进行坐标变换实现了绝对位置智能定位,提高了定位的准确性。仿真结果表明,该方法定位精度高,对测量误差、节点连接度以及锚节点比例的变化有很强的鲁棒性。

This paper proposed a method of nodes localization based on anchor clustering（LAC）,improved the localization accuracy by using this method in the wireless sensor networks.All the nodes was divided into many clusters which anchor as a center autonomously and the fake anchor defined as the node located on margins of clusters.Built the distance matrix between the nodes by using measuring methods,obtained a local relative map of a cluster by using multi-dimensional scaling（MDS） algorithm and anchor positional information,implemented the global map by using of nodes positional coordinate transformation of anchor and fake anchor.Simulation results demonstrate that LAC is more robust to disturbance effect and higher accuracy performance in the nodes localization.