摘要： 定位技术是无线传感器网络的支撑技术之一。针对无线传感器网络低成本、低能耗的要求，提出一种抽样定位和求精的分布式算法。算法在第1阶段基于收到的锚节点信息进行抽样，形成节点的初始位置估计。在第2阶段对节点初始位置进行求精。仿真实验结果显示了该定位算法可以在1~2次求精情况下达到收敛，在样本量为20左右的较低阈值下实现较高的定位精度，在12%左右的锚节点比例下实现95%以上的定位覆盖率，并且与dv-hop和dv-distance定位算法比较，证明该算法分别可以提高20%和5%的定位精度。

关键词: 无线传感器网络, 定位, 抽样定位, 求精算法

Abstract: Localization is one of the supporting technologies in wireless sensor networks. In this paper, a novel sampling localization and refinement distributed algorithm is described and evaluated. The algorithm includes two stages. In first stage, sampling is carried out based on received anchor information and the initial position estimate of sensor node comes into being. In second stage, the initial node position is refined. Simulation results show this algorithm is convergent under 1~2 refinement times and achieves high localization accuracy in low sample size threshold as 20. It also gets 95% localization coverage when anchor ratio is about 12%. Comparing with dv-hop and dv-distance methods, it improves localization accuracy by 20% and 5%.

Key words: wireless sensor networks, localization, sampling localization, refinement algorithm

中图分类号: 

• TP393