基于层次聚类的WiFi室内位置指纹定位算法

提出一种利用WiFi信号指纹实现对室内区域进行定位的CL-KNN(complete linkage K-nearest neighbor)算法.该算法先采用层次聚类方法对测试环境进行区域划分,再根据相应的WiFi信号指纹信息进行匹配,最后通过加权计算确定定位结果.实验结果表明,在WiFi热点数量足够多的情况下,与原始KNN算法和kmeans-KNN算法相比,CL-KNN算法可以获得更高的定位精度和准确率.

WiFi indoor position fingerprinting localization algorithm based on hierarchical clustering

This paper presents an implementation of positioning of the interior regions of the CL-KNN (Complete-Linkage K-NearestNeighbor) by using the WiFi signal fingerprint algorithm. The algorithm first performs a region division to a test environment with hierarchical clustering method,and it according to the WiFi signal fingerprint information corresponding to match , and finally through the weighted calculation to determine the location results. The experimental results show that, in the cases that WiFi access point is sufficient ,CL-KNN algorithm can achieve higher positioning precision and accuracy compared with the original KNN algorithm and kmeans-KNN algorithm.