摘要
定位技术是无线传感器网络的关键技术之一,移动节点的定位则是定位技术中的难点。目前以蒙特卡洛算法为基础的几种定位算法普遍存在对锚节点数量要求较高、定位误差较大的问题。针对实际应用中一般的运动模型,提出一种基于参考节点选择模型的蒙特卡洛定位算法(MCLAS),采样时通过参考节点选择模型,将相邻节点纳入参考节点选择范围,尽量选取距离定位节点较近且分布均匀的参考节点构成采样盒,位置估计时利用样本与节点依据运动模型预测的上一时刻运动方向的角度确定权重,避免低质量样本的过度使用,提高定位精度。仿真结果表明,算法在定位精度、定位覆盖率和时间复杂度等方面具有较好的性能。
Localization is an important technology in Wireless Sensor Networks（WSN）,and the localization of mobile node is a nodus.There lies some problems,such as need for many anchor node and high error of localization,in present localization methods based on Monto Carlo Localization Algorithm（MCL）.This paper propose a MCL algorithm based on Anchor node selected model（MCLAS）,which bring neighbor nodes into reference nodes in course of sample,and select the nearest nodes which distribute symmetrically to construct sample box.Next,in course of evaluating position,MCLAS decide the sample＇s weight according to the angle of sample and moving direction of node forecasting by motion model,which avoid the excessive use of bad sample,raising precision of localization.The simulation demonstrates that the proposed algorithm provides better performance in localization precision,localization coverage,and time complexity.
出处
《传感技术学报》
CAS
CSCD
北大核心
2011年第2期264-268,共5页
Chinese Journal of Sensors and Actuators
关键词
无线传感器网络
定位
蒙特卡洛
权重
wireless sensor networks
localization
Monto-Carlo
weight
