Indoor location with Wi-Fi fingerprinting

Abstract

There are many applications for indoor location determination, from the navigation of hospitals, airports, parking garages and shopping malls, for example, to navigational aids for the blind and visually impaired, targeted advertising, mining, and disaster response. GPS signals are too weak for indoor use, however, making it necessary to investigate other means of navigation. Most approaches such as ultrasound and RFID tags require special hardware to be installed and remain expensive and inconvenient. The solution proposed in this paper makes use of commonly available Wi-Fi networks and runs on ordinary smart phones and tablets without the need to install special hardware. It comprises a calibration stage and a navigation stage. The calibration stage creates a "Wi-Fi fingerprint" for each room of a building. It minimizes the calibration time through the use of waypoints. The navigation stage matches Wi-Fi signals to the fingerprints to determine the user's most likely location. It uses maximum likelihood classification for this matching and takes the building's topology into account through the use of Bayes' Theorem. The system is implemented as a mobile Android app and is easy to use. In testing, it took only an hour to calibrate a home or shopping mall, and the navigation stage yielded the correct location 97.5% of the time in a home and 100% of the time in a mall.