Indoor Positioning System Based On WIFI Fingerprinting Method

With the advent of the Global Positioning System(GPS),the question of positioning in the outdoor environment is considered to be satisfactory.However,the issue of positioning in the indoor environment is still a massive challenge for researchers.While GPS is almost impossible to use in indoor environment,many alternatives have been proposed.One of the most potential and high popularity solutions is the fingerprint-based solution.It includes a training phase for collecting radio signal strength samples(RSS)in schools that need localization into a database(called a radio map).Radio signals can be from WiFi access points,GSM base stations or other RF-based networks,but due to the high popularity of WiFi,it is often chosen as the primary signal for the indoor positioning systems.Then,in the positioning phase,the mobile device that interested in its location can collect its current RSS samples and compare it to the radio map set up during the training phase to determine its possible position.Our indoor positioning system based on WiFi fingerprinting method has been developed on the Android platform.We present a robot-based system and many cheap RFID tags deployed on the ground to automate the training process and,more importantly,regularly update radio maps to reflect current RSS templates.The experiment was conducted in a lab with 108 reference points on 58 square meters.The system achieves an average accuracy of 3 to 6 meters with the training database collected manually by human and 1 to 5meters with the training database automatically collected by robot.Experimental results have shown the advantages of automated data collection,which not only significantly reduces human labor but also maintains the positioning accuracy of the system.