EESM-based fingerprint algorithm for Wi-Fi indoor positioning system

Abstract

In recent years, great improvements took place in smartphone industry. Along with the development of cloud services and web applications, lots of developers also take great effort on developing smartphone applications. Among various applications related to social networks, LBS (Location Based Service) is the key technique which is the basic for social interactive activities. While GPS (global positioning system) works well enough outdoors, Wi-Fi RSS (receive signal strength)-based fingerprinting system is the most promising solution for indoors. A novel EESM-based fingerprint algorithm which improves the positioning performance by involving channel estimation for creating more robust fingerprints is proposed in this paper. Moreover, a cloud computing based indoor positioning system is also introduced for evaluating performances. Both simulations and experiments show that EESM-based fingerprints are more stable and representative in the multipath indoor environments. An up to 30% reduction in error distance in a classic indoor positioning system using k-NN matching algorithm is provided by our proposed EESM-based fingerprints.