基于Wi－Fi室内定位的位置指纹算法

摘要

随着 Wi－Fi网络室内覆盖率的增加，人们对位置服务的需求也不断增加。利用现有的Wi－Fi网络进行室内定位，先分析了NN算法存在的不足，并提出了一种WNN位置指纹定位算法。整个定位过程分为数据采集、数据处理即奇异值处理和滑模滤波处理、实时定位即用WNN法对处理后的离线数据和实时数据进行对比、匹配，最接近的那组数据对应的区域号为定位结果。对几种信号处理方法进行了比较，得出滑模滤波的效果最优。实验表明，修正后的定位算法的定位准确率明显提高，达到了90％以上。%This paper uses the existing Wi-Fi network forindoor positioning and anaIyzes the deficiency in NN aIgorithm and then proposed a kind of WNN position fingerprint IocaIization aIgorithm.The whoIe positioning process is divided into data acqui-sition, data processing is singuIar vaIue processing and sIiding fiIter processing for the coI ected data,reaI-time Iocation we use the method of weighted nearest neighbor to make a comparison and match between the reaI-time data and offIine data which are aI processed.