Abstract: Node self-localization is one of the crucial techniques for Wireless Sensor Networks（WSN）.The traditional Sequence-Based Localization（SBL） only focuses on 2D networks.A new localization algorithm called 3D sequences and centroids algorithm for 3D WSN is presented which divides the localization space into distinct regions such as edge，face and body by the perpendicular faces of reference nodes’ pairs.It uses the triangular area enclosed by the centroids of the three “nearest” location regions and chooses the centroid of the triangular represented as location estimate of the unknown node.It doesn’t need any improvement in hardware.Simulation experiments prove that it can obviously improve the localization accuracy compared with SBL and can satisfy the need of localization for 3D WSN.

Key words: Wireless Sensor Networks（WSN）, node self-localization, 3D sequences and centroids algorithm, localization sequences

摘要： 节点自定位技术是无线传感器网络的关键技术之一。三维序列重心算法利用锚节点两两之间的垂直平分面将定位空间分为边、面和体三类区域，缩小了未知节点可能存在的范围，并在所在范围内再次求出离未知节点最近三点组成的三角形的重心作为未知点位置的估计。该算法改善了二维序列算法误差较大的问题，且不需要增加硬件设施来实现特殊的功能。仿真结果表明，该算法可以达到较高的定位精度，能够满足三维空间中未知节点定位的应用需要。

关键词: 无线传感器网络, 节点自定位, 三维序列重心算法, 定位序列