Map-Aware Indoor Area Estimation with Shortest Path Based on RSS Fingerprinting

With the widespread of smartphones, people can easily figure out where they are and enjoy other advanced services like searching nearby restaurant information or checking bus arrival time. Indoor positioning becomes a popular issue for location- based services used in shopping malls, hospitals, or largescale buildings. Contrast with spacious surroundings of outdoor, indoor environment is filled with obstacles and moving people, which impose great challenges to provide precise estimation of indoor positioning. This paper proposes Wi-Fi fingerprinting technique using received signal strength with consideration of map information to effectively eliminate unreasonable estimation outcomes. The proposed area estimation (AE) algorithms calculate the similarity of each area in the entire region to increase accuracy of distinguishing which area the user locates. Moreover, the shortest path with adjacent recognition (SPAR) algorithm further utilizes the concept of Dijkstra´s shortest path algorithm and the previous location information to predict user´s position. Experimental results show that the proposed AE with SPAR algorithms can provide better area estimation compared to conventional scheme.