Indoor pedestrian navigation fusing WLAN, INS, and map data using pattern matching

This paper presents an indoor pedestrian navigation scheme based on integrating different sources of information including wireless local area networks signals, inertial sensing, and map constraints. The proposed scheme is implemented and successfully tested on an Android mobile phone. Obtaining desired navigation performance solely based on inertial navigation systems (INS) and utilizing often inaccurate sensors of a phone is not feasible. The methods solely based on WLAN fingerprinting face the problem of inaccuracies resulting from the variant nature of the received signal strengths. The proposed approach fuses INS pattern matching with WLAN indoor positioning within the constraints of a map using a scoring function designed for a desired trade-off of positioning accuracy and complexity.