Optimizing Artificial Neural Network for Beacon Based Indoor Localization

In this paper, we study the problem of indoor localization with emphasis on precision. There are already various technologies that can be used to gather data and more or less precisely estimate the user's location. In our work, we use Bluetooth Low Energy beacons. When these transmitters are properly placed in the environment, their signal can serve as a data input. We designed a feed-forward artificial neural network that processes this data and produces estimated coordinates that denote the position of the user. Experiments show, that our approach achieves the best average precision of 1.21 meters which is comparable to results reported by other known approaches.