The Fast Collection of Radio Fingerprint for WiFi-Based Indoor Positioning System

Abstract

Mobile communications and Internet have become one of the most important services nowadays. Most mobile devices, such as cell phone, PDA, and laptop computer, equipped with the WiFi, GPS, and Bluetooth as their standard, build-in equipment, and people have got used to use mobile services in their daily life. One interesting service is Wifi-based indoor positioning system (IPS) that attracts many researchers devote their effort to it. Many researches in IPS determine the user location by the method of scene analysis. This method needs to collect the RSSI of APs from the interested place beforehand to build the building's WiFi radio fingerprint database and this task is time-consuming. Meanwhile, it also needs to resample very often in order to keep the accuracy of the positioning result. In this paper, we design and test a fast setup algorithm for collecting those sampling information. The Android smartphone and its build-in motion sensors were used to help collecting the AP's RSSI. We use the motion sensor to detect the pace while walking and collect the AP's RSSI in every step. The method gets through the sampling work in a walking duration which is much shorter than traditional method. This paper also compares the fast setup method with the traditional method from the positioning accuracy point of view. Experiments show that there is no significant difference on positioning accuracy between them.