Mobile Indoor Positioning System Utilizing WiFi RSSI and Motion Data

Recently, locating the position of people or objects in indoor environments where GPS signal is not available or lack light-of-sight (Indoor Positioning System) has attracted many researchers. Researches try to improve the performance of the indoor positioning system by proposing algorithms with the application of artificial models especially the deep learning models. This research focuses on studying the performance of indoor positioning systems utilizes WiFi RSSI fingerprinting for both static and mobile scenarios. In the static scenario, we use the traditional offline survey phase to collect the WiFi RSSI data, and then estimate the position of nodes in the online phase based on newly measured WiFi RSSI and data from the offline phase with various learning models. In the mobile scenario, we propose schemes to combine the result from a static scenario with the motion data to estimate the position of mobile nodes. In the proposed schemes, the estimated distance and angle from smartphone sensors (motion sensors) are analyzed to find the optimal adjustment to estimate the position of mobile nodes. The results from experiments with people carrying android devices show that our proposed scheme archives very good accuracy in indoor environments.