基于室内地图环境信息的多楼层WiFi定位技术研究

摘要

In order to implement multi-floor buildings accurate location, the map information based indoor position scheme is proposed in this paper. Firstly, the map information model and affinity propagation method for fingerprints clustering are explored in offline phase. Then, the received signal strength indication (RSSI) threshold based floor identification method is put forward. Finally, the terminal positions are obtained by using the map information model and maximum posteriori position estimation method. Experimental results show that the proposed indoor position scheme can effectively enhance the location estimation accuracy and reduce the computation complexity for multi-floor buildings.%针对多楼层环境的室内定位需求，提出了一种基于地图环境先验信息的WiFi指纹定位方法。首先在离线阶段建立地图环境信息模型并对指纹进行仿射传播聚类，然后在线阶段采用RSSI阈值的楼层判别算法确定楼层，并结合地图信息模型和最大后验估计方法计算出终端位置。实验结果表明，相比于传统的指纹定位方法，室内定位技术不仅能提高定位精度，并且降低了在线阶段指纹匹配计算的复杂度。