quotation:		[Copy]
		[Copy]

This Paper:Browse 1954 Download 9
基于WiFi的分级室内定位
王行娟
0 Fontlarge +\|Default\|Small
(武汉华夏理工学院信息工程学院，武汉 430223)

摘要:

在经典的K最近邻（KNearest Neighbors,KNN）的WiFi定位方法中，其算法复杂度随着定位区域和定位区域内的WiFi接入点（Access Point，AP）的增加而增加，无法满足实时定位的要求。为此，提出一种分级WiFi定位算法。算法分为粗定位和精定位阶段，首先通过AP的可见性利用汉明距离寻找可能的子区域，再用KNN算法在子区域内（利用信号强度欧氏距离）进行精定位。经过实测数据验证，平均单次定位时间在KNN算法下下降了约95%，在最大后验算法下下降了约96%，表明所提分级定位框架具有延迟低的优点。

关键词: 室内定位 WiFi定位 K最近邻分级结构

DOI：

基金项目:湖北省科研计划项目(B2020314，B2020309)

WiFi-based indoor positioning by adopting a hierarchical structure

WANG Xingjuan

(School of Information Engineering,Wuhan Huaxia University of Technology,Wuhan 430223,China)

Abstract:

WiFibased indoor positioning has great potential applications in locationbased service(LBS),due to its independence of extra hardware.However,in the classical Knearest neighbors(KNN) algorithm adopted in WiFibased positioning,the computational cost increases with the size of the area and the number of access points(APs),and it cannot fulfill the requirement of realtime implementation.In this paper,a hierarchical positioning method is proposed for WiFibased positioning.The method includes two phases:coarse positioning and accurate positioning.In coarse positioning phase,some possible subregions are found according to the Hamming distance.In accurate positioning phase,the classical KNN algorithm is adopted.The proposed method can decrease the computation cost and accelerate the positioning.Experiment shows that the proposed method has a decrease of about 95% and 96% in average positioning time under the KNN method and the maximum posterior probability method,respectively.

Key words: indoor positioning WiFi based positioning K nearest neighbors hierarchical structure