Real-Time Outdoor Localization Using Radio Maps: A Deep Learning Approach
This paper deals with the problem of localization in a cellular network in a dense urban scenario. Global Navigation Satellite Systems typically perform poorly in urban environments, where the likelihood of line-of-sight conditions between the devices and the satellites is low, and thus alternative localization methods are required for good accuracy. We present LocUNet: A deep learning method for localization, based merely on measured received signal strengths (RSS) from Base Stations (BSs), which does not require any increase in computation complexity at the user devices with respect to thedevice standard operations, unlike methods that rely on Time of Arrival (ToA) or Angle of Arrival(AoA) information. In a wireless network, user devices scan the base station beacon slots and identifythe few strongest base station signals for handover and user-base station association purposes. In theproposed method, the user to be localized simply reports such received signal strengths to a centralprocessing unit, which may be located in the cloud. Alternatively, the localization can be performedlocally at the user. Using estimated pathloss radio maps from the BSs, LocUNet can localize users withstate-of-the-art accuracy and enjoys high robustness to inaccuracies in the radio maps. The proposedmethod does not require pre-sampling of the environment; and is suitable for real-time applications,thanks to the RadioUNet, a neural network-based radio map estimator. We also introduce two datasetsthat allow numerical comparisons of RSS and ToA methods in realistic urban environments.