Smartphone based Indoor Path Estimation and Localization without Human Intervention

The growing commercial interest in indoor localization-based services has stimulated the development of many indoor positioning systems. Despite extensive research on localization, system requirements, such as site survey, user intervention, or specific hardware/software, place limitations on the widespread deployment of localization. To overcome these limitations, we propose a path estimation and localization system for indoor environments, termed PYLON, that runs on a smartphone and a server without any human intervention. PYLON uses an actual floor plan and measurements from widely deployed WiFi access points (APs) and Bluetooth Low Energy (BLE) beacons to estimate the user's path. It creates virtual rooms according to received signal strength indicator (RSSI) values and matches them to actual rooms in the real-world floor plan. After room mapping, PYLON uses door passing times to precisely refine a user's estimated path. Unlike conventional path estimation and localization systems, PYLON works independently of device types. We implement PYLON on five Android smartphones and conduct evaluation with three users in an office building. Our experimental results show that PYLON achieves 97% floor plan mapping accuracy with a localization error of 1:42 m.