Indoor Localization on Smartphone Using PDR and Sparse Deployed BLE Beacons in Large Open Area

Mei Wang,Nan Duan,Zou Zhou,Hongbing Qiu,Xiaopeng Li,Fei Zheng
DOI: https://doi.org/10.1109/iwcmc51323.2021.9498670
2021-01-01
Abstract:A lot of work has demonstrated that satisfactory localization accuracy can be achieved in a typical office environment (e.g. teaching building, hospital and office building) which contains many small rooms on both sides of narrow corridors. However, it is observed that the accuracy degrades a lot when it comes to the large open environment, such as halls, supermarkets and airports. Another problem is the higher hardware cost and computational burden caused by the large localization area. In this paper, we analyze the feasibility of each technique in large open environment and then propose a practical indoor positioning system, in which PDR cumulative error is intermittently reduced by sparsely deployed BLE beacon. One of our contributions is to propose a single-beacon-based positioning method that is based on the RSS peak and user's walking direction. Furthermore, we adopt a robust RSS peak detection method to recognize the real RSS peak accurately. Meanwhile, to compensate for the delay in RSS peak detection, PDR is used to infer the corresponding displacement. Finally, unlike the traditional step length estimation methods establishing an empirical formula, we propose a novel parameters-free method which doesn't require offline training and is feasible for everyone. Experimental results show that our proposed system exhibits 90% accuracy within 3.5m for a total distance traveled of 377m in the large open environment.
What problem does this paper attempt to address?