A real-time algorithm for fingerprint localization based on clustering and spatial diversity

Reliable and real-time location information is important for location-based services (LBS) of the mobile terminals. Existing location techniques have poor performance in urban environments due to severe multipath and NLOS propagation. Fingerprint localization based on wireless-LAN (WLAN) has been preferred for those types of environments. However, the most challenge in fingerprint positioning is the fluctuation of the received signal strength (RSS) and the online computational burden or latency for a metropolitan-scale radio map. To alleviate this problem, we propose a real-time algorithm including two key techniques: (1) The micro-cell-based radio map construction method exploiting spatial diversity is used to tackle the fluctuation of RSS and enhance positioning accuracy. (2) The RSS-based clustering of the radio map is presented to reduce the computational burden and improve the performance of the locating system. The practical implementation of positioning mobile terminals at different speeds in urban environments is described and conducted. Experimental results show that our algorithms can significantly improve the performance of WLAN-based positioning system.