Indoor wireless local area networks (WLAN) location system using a Kalman filter

There are many location determination techniques that are used to determine a mobile user's position. These include the globe positioning system (GPS), infrared (IR) based systems, radio frequency (RF) based systems and ultra wideband (UWB) based systems. Considering the indoor environments, RF based systems, especially those based on IEEE 802.11, have garnered considerable attention because these systems have an increasingly ubiquitous coverage area and they are able to work with the existing infrastructure without requiring additional location determination hardware. The most challenging issues concerning wireless local area network (WLAN) based location systems are multipath fading and interference caused by variations in the indoor propagation environment as well as interference from other devices working on the same unlicensed frequency band at 2.45GHz. This thesis addresses the issue of indoor location estimation using an existing WLAN infrastructure. The location detection technique investigated here builds a signal radio map in the test area using the received signal strength index (RSSI). To alleviate the effect of multipath fading and interference, the Kalman filter is adapted in this location determination system, which effectively filters out various noise sources in the test environment. The Kalman filter is also investigated in combination with an accelerometer for improving the accuracy when tracking mobile users. Experimental results using real data collected from an office environment indicate that the proposed Kalman filter based method improves the accuracy both in locating and tracking mobile users than the measurement without Kalman filters.