Indoor positioning is a key component for location-aware mobile applications since GPS localization is not available in buildings or other indoor places. There exist several deductive and inductive methods, but many of them have high errors in their location estimation considering the issues related to the indoor environment and sensor measurements. In this paper, we present an algorithm for indoor localization based on probabilistic approach which provides large-scale and distributive platform for indoor positioning. It is tested by using the UJIndoor- Loc dataset and the results are compared with KNN method proving the accuracy and low-complexity of the algorithm. At the end, it is presented a real-time implementation of large-scale platform for indoor localization which is used as a presence management system in Faculty of Computer Science and Engineering.
Indoor positioning Wi-Fi N-Gaussian Peaks Smartphone e-FINKI.