An Improved Indoor Positioning Method Based on Affinity Propagation Clustering

The fingerprint-based wireless local area network (WLAN) positioning has gained significant interest in recent years. Indoor localization methods based on WLAN and RSS with advantage of low cost are most widely used. In this paper, we propose an improved indoor positioning method based on Affinity Propagation (AP) algorithm to obtain better accuracy: k coordinate points measured through KNN algorithm at the stage of orientation are clustered through Affinity Propagation (AP) algorithm into the largest cluster whose center node serves as the ultimate coordinate point. Experimental results conducted in the real environments show that our proposed algorithm can obtain improvement of the mean error distance of 18.68% and 34.72%, compared with the KNN method and traditional NN method, respectively.