A Ubiquitous and Accurate Floor Estimation System Using Deep Representational Learning

Location-based services have undergone massive improvements over the last decade. Despite intense efforts in industry and academia, a pervasive infrastructure-free localization is still elusive. Towards making this possible, cellular-based systems have recently been proposed due to the wide-spread availability of the cellular networks and their support by commodity cellphones. However, these systems only consider locating the user in a 2D single floor environment, which reduces their value when used in multi-story buildings. In this paper, we propose CellRise, a deep learning-based system for floor identification in multi-story buildings using ubiquitous cellular signals. Due to the inherent challenges of leveraging the large propagation range and the overlap in the signal space between horizontal and vertical user movements, CellRise provides a novel module to generate floor-discriminative representations. These representations are then fed to a recurrent neural network that learns the sequential changes in signals to estimate the user floor level. Additionally, CellRise incorporates different modules that improve the deep model's generalization against avoiding overtraining and noise. These modules also permit CellRise to generalize to floors completely unseen during training. We have implemented and evaluated CellRise using two different buildings with a side-by-side comparison with the state-of-the-art floor estimation techniques. The results show that CellRise can accurately estimate the exact user's floor 97.7% of the time and within one floor error 100% of the time. This is better than the state-of-the-art systems by at least 17.9% in floor identification accuracy. In addition, we show that CellRise has robust performance in various challenging conditions.