一种基于离群点检测的定位算法  

A LOCALIZATION ALGORITHM BASED ON THE OUTLIER DETECTION

作　　者：刘广聪[1] 郝艳茹 刘铮 Liu Guangcong;Hao Yanru;Liu Zheng(College of Computer,Guangdong University of Technology,Guangzhou 510006,Guangdong,China)

机构地区：[1]广东工业大学计算机学院,广东广州510006

出　　处：《计算机应用与软件》2018年第3期130-134,共5页Computer Applications and Software

基　　金：广东省科技计划项目(2015B090901017);广州市科技计划项目(201508020030)

摘　　要：DV-Hop算法在无线传感器网络节点分布不均匀时定位误差较大。针对上述问题,利用离群点检测算法提高计算未知节点坐标的精度。在采用多边测量算法估计未知节点的坐标位置时用离群点检测算法LOF对未知节点的估计坐标进行分析和筛选,最终确定未知节点位置。仿真实验表明,该方法能提高节点的定位精度,减小定位误差。The DV-Hop algorithm has a large positioning error when the network nodes are not evenly distributed.According to the above problem,this paper uses the outliers detection algorithms to improve the calculation precision of the unknown node coordinate.In the process of using multilateral localization algorithm to estimate the unknown node coordinates,the outliers detection algorithm LOF is utilized to analyse and screen the estimated coordinates of unknown nodes and then confirm the final location.The experiment indicates that the improved algorithm can improve positioning accuracy of the nodes by reducing errors

关 键 词：DV-HOP 加权 离群点检测算法 LOF 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]