GPS does not provide reliable accuracy in indoor and dense urban environments due to weak or blocked signals and multipath. In this research, an alternative indoor navigation solution using IEEE 802.11 WLAN (WiFi) and Low Cost MEMS-based inertial sensors based on an optimized adaptive version of the mixture Particle Filter (PF) is proposed. WiFi fingerprint positioning was used to provide wide indoor coverage. Reliable short-term accuracy of an Inertial Navigation System (INS) was used as a fine-tuner under general guidance of the WiFi solution. In addition, a multi-direction back-tracking zero velocity update algorithm that significantly reduces INS errors is proposed. Moreover, an adaptive weighting approach was developed to adjust the confidence in either INS or WiFi according to motion conditions. Physical experiments were performed using one MEMS-based gyroscope and two accelerometers integrated with a WiFi network in an indoor environment. Results showed significant improvement in overall navigation accuracy outperforming many state-of-the-art WiFi indoor navigation techniques.

Additional Metadata
Journal Navigation, Journal of the Institute of Navigation
Citation
Atia, M, Noureldin, A. (A.), Georgy, J. (J.), & Korenberg, M. (M.). (2011). Bayesian filtering based WiFi/INS integrated navigation solution for GPS-denied environments. Navigation, Journal of the Institute of Navigation, 58(2), 111–125.