摘要： 定位是无线传感器网络的基础工作。现有定位算法利用参考节点的位置信息对非参考节点进行定位，当该信息受到攻击或误差的影响时，将导致算法精确度下降。该文将传统最小二乘定位算法与Metropolis-Hasting抽样算法有机结合，提出一种改进的最小二乘定位算法。建造一个可能遭受攻击的模拟环境，在该环境下比较改进后的算法和原算法，结果表明，改进后的算法具有较好鲁棒性。

关键词: 无线传感器网络, Metropolis-Hasting抽样算法, 分布特性, 鲁棒性

Abstract: Location is foundation work in Wireless Sensor Network(WSN). Existing location algorithms need the position information of beacon nodes to locate the non-beacon nodes. When the position information is affected by the attacker or errors, the accuracy of these algorithms will degrade. This paper combines the LS localization algorithm and Metropolis-Hasting sample algorithm, and proposes an improved LS localization algorithm. It constructs a simulation circumstance which might be attacked and compares the improved algorithm with original algorithm. Experimental results demonstrate that the improved algorithm has better robustness than original algorithm.

Key words: Wireless Sensor Network(WSN), Metropolis-Hasting sample algorithm, distribution trait, robustness

中图分类号: 

• TP393