Indoor Fingerprint Localization Based on Fuzzy C-Means Clustering

Accuracy of the global positioning system (GPS) cannot meet the demand of indoor service. To address this issue, a radio frequency (RF) based system named RADAR for locating and tracking users inside buildings is presented, using fingerprint architecture. However, the traditional system is still sensitive to multipath and body movement. Furthermore, it costs much computing time. In this paper, we propose a fingerprint algorithm based on fuzzy c-means clustering. It uses clustering to reduce the computing time. We have evaluated the system in underground parking area. The results show that the technique reduces the computing time below a reasonable degree and enhances the accuracy of previous system slightly.