基于PCA-LSSVR算法的WLAN室内定位方法  

Indoor positioning algorithm for WLAN based on principal component analysis and least square support vector regression

作　　者：张勇[1,2] 黄杰[1] 徐科宇 

机构地区：[1]合肥工业大学计算机与信息学院,合肥230009 [2]芜湖创业园留学人员博士后科研工作站,芜湖241000

出　　处：《仪器仪表学报》2015年第2期408-414,共7页Chinese Journal of Scientific Instrument

基　　金：国家科技支撑计划(2013BAH52F01);国家级大学生创新创业训练计划项目(201410359025)资助项目

摘　　要：针对WLAN室内定位系统中存在的接收信号强度指示(RSSI)时变特性降低定位精度的问题,提出一种基于主成分分析(PCA)和最小二乘支持向量回归机(LS-SVR)的PCA-LSSVR定位算法。该算法首先利用PCA对采集的各接入点(AP)的原始RSSI信号进行数据降维和去相关处理,提取主要的定位特征数据;然后利用LS-SVR构建指纹点的定位特征数据与其位置的非线性关系,并利用此关系对测试点的位置进行回归预测。实验结果表明,该算法的定位精度优于几种传统的定位算法,是一种性能良好的WLAN室内定位算法。The time-varying characteristic of the received signal strength indication（ RSSI） degrades the indoor positioning accuracy in wireless local area network（ WLAN）.A novel indoor positioning algorithm PCA-LSSVR based on principal component analysis（ PCA） and least square support vector regression（ LS-SVR） is proposed to address the problem.Firstly,the proposed algorithm employs PCA to reduce the dimensions of the original RSSI and their correlation,which are sampled from each access point（ AP）,and the main location features are extracted.Secondly,LS-SVR is employed to build the nonlinear relationship between the location features of reference points and their locations,and then test points＇ locations based on the nonlinear relationship are predicted.Experimental results show that the proposed algorithm is superior to several traditional indoor positioning algorithms with better positioning accuracy.

关 键 词：WLAN 室内定位 主成分分析 最小二乘支持向量回归机 接收信号强度指示 

分 类 号：TN92[电子电信—通信与信息系统] TP393.17[电子电信—信息与通信工程]