An Adaptive Indoor Positioning Method Using Multisource Information Fusion Combing Wi-Fi/MM/PDR

Abstract

With the extensive application of microelectromechanical systems (MEMS) in smartphone, indoor positioning based on smartphone has attracted immense attention over recent years. An indoor positioning method based on a single positioning information source has limitations, and it is difficult to balance the accuracy and cost. Therefore, it is worth further research to reduce the positioning error and improve the robustness of a system by integrating multiple sensors. The direct fusion of multiple information sources may introduce new errors. Therefore, the information sources should be screened to make their advantages complementary. In this paper, a hierarchical indoor positioning method integrating Wi-Fi, magnetic matching (MM), and pedestrian dead reckoning (PDR) is proposed. Firstly, a region is divided into several subregions by a cluster algorithm. Then a magnetic matching suitability (MMS) model is proposed to evaluate the subregional MMSs, and different strategies are adopted for the information fusion according to the evaluation results. Finally, an adaptive extended Kalman filter (AEKF) with innovation is proposed to integrate PDR and the absolute localization result. Experimental results show that the proposed positioning method can make full use the advantages of each positioning information source and has better positioning accuracy and stability.