Indoor Positioning Based On Smart Mobile Devices

Global Positioning System (GPS) as an outdoor positioning technology has been widelyused in vehicle navigation, logistics, disaster response, and military strikes. However, in theindoor case, due to the block of buildings, the GPS positioning accuracy will be greatlydecreased, even can’t work. So a more rational indoor positioning solution is needed. On theother hand, since2007the first iPhone launch, in2008the first Android phone on sale, in2010the first iPad tablet on sale, just a few years, smart mobile devices have rapidly inpeople’s lives plays an indispensable role. The traditional Internet business includes socialnetworking, online games, immediate chatting is rapidly migrating to mobile devices. For themobile characteristics of mobile devices, people increasingly want to be able to get their ownlocations and others’ locations through mobile devices. Our study of indoor positioning focuson two main aspects based on mobile devices: WiFi positioning and acoustic positioning.For the WiFi positioning, based on the research of history, we find it is necessary todevelop a WiFi positioning research toolset based on mobile devices. This paper also analyzesthe components of WiFi positioning research process. Designed and implemented a toolset forWiFi positioning research. The toolset covers the “definition of test bed”,“data collection”,“emulation and comparison” and “verification”. On top of these processes the toolset alsocontribute a C/S architecture positioning prototype to researchers.For the WiFi fingerprint will be out of date because of the environment changes, wepropose a WiFi location fingerprint update algorithm. The algorithm uses user feedback datato score the AP in the location fingerprint. The algorithm can detect new installed andremoved APs, especially considers the unstable AP. Second, it can eliminate the temporaryAP’s impact on traditional method s. Finally, this algorithm can accelerate the fingerprintdatabase’s update by detecting the AP’s RSS change cause by AP moving and furniturereorganizing. The experimental results show that this algorithm is superior to the traditionalalgorithm on the efficiency of database’s update.For the acoustic positioning, this paper using the Doppler effect of the received sound,when the mobile device moves. This paper researches the direction finding using smartmobile devices firstly. We design a direction finding scheme that uses software PLL approachto break down the insufficient resolution of signal that was suffered by traditionaltime-frequency analysis. The scheme also uses linear regression to handle the noise of inertialsensors. The accuracy of acoustic direction finding is around2.1degrees within the range of32m. The movement style needed by this scheme is gentle and acceptable.Based on the accuracy acoustic direction finding, this paper also installed some soundanchor in the room to do mobile devices indoor positioning and tracking. Because of theinaccuracy of the compass in the mobile device, we use the difference of the angle to arrivalto do indoor positioning. For indoor positioning, the performance of our method is90-percentile errors are under0.92m. After getting initial position, this paper also uses therelationship of Doppler shift and device distance shift to do real time indoor tracking.