An enhanced K-Nearest Neighbor algorithm for indoor positioning systems in a WLAN

With the rapid development and ubiquitous usage of Wireless Local Area Networks (WLAN), Location Based Systems (LBS) employing Signal Strength techniques have become an attractive area of research for location estimation in indoor environments. In this paper we propose a robust fingerprint method for localization based on the traditional K-Nearest Neighbor (KNN) method. Instead of considering a fixed number of neighbors, our approach uses an adaptive method to determine the optimal number of neighbors to be taken into account.. In order to prove the effectiveness of our method, we compare it with the traditional KNN approaches for a variety of number of Access Points (APs). Simulation results using Multi-Wall-Floor path loss model show that the proposed method yields an improved accuracy as compared with the traditional methods.