Deep Learning-based Floor Prediction Using Cell Network Information

Abstract

Location services are one of the most used applications today on mobile devices. The vast majority of localization systems propose solutions for locating the user in a 2D single floor environment. However, accurate estimation of the user's floor level, in tall multistory buildings, is a crucial basis for many applications, especially for emergency services. This paper presents a fingerprinting-based system that provides a low-cost floor localization service using the ubiquitous cellular signals received by the user's cell phone. Specifically, a convolutional neural network is trained to map the sequential change of the received cellular signals to the corresponding floor. Evaluation using different Android phones shows that the proposed system can track the user floor with at least 95.9% accuracy in different scenarios. This demonstrates the superiority of the system compared to the state-of-the-art systems in all experiments.