Indoor pedestrian integrated localization strategy based on WiFi/PDR

In view that WiFi localization has poor instability in indoor pedestrian localization, an improved K-nearest neighbor algorithm is proposed to overcome this problem. A real-time updated step-length model and a heading estimation algorithm based on indoor environmental features are proposed to improve the positioning accuracy of pedestrian dead reckoning. In addition, a self-adaptive particle filtering algorithm is used to integrate the WiFi with the pedestrian dead reckoning. An adaptive factor is used to automatically adjust the influence of WiFi observations on particle movements. A series of experiments were implemented on mobile phone, and the results show that the proposed integration localization strategy achieves 0.66 m location accuracy which is better than that of the traditional particle filtering algorithm.