On the RBF-based positioning using WLAN signal strength fingerprints

Abstract

Provision of accurate and reliable location estimates is the key issue for the proliferation of indoor location oriented services and applications. Our positioning method is based on Radial Basis Function (RBF) networks and we exploit Received Signal Strength (RSS) measurements from several WLAN Access Points (AP). We incorporate the RSS covariance matrix into the estimation method and couple that with a methodology that indicates which APs can be ignored during positioning without sacrificing accuracy. We evaluate the RBF method in a real-life setup, and experimental results suggest that the proposed approach performs well.