Learning from the sky: Design of autonomous radio-enabled unmanned aerial vehicles in mobile cellular networks

The use of UAVs in wireless networks has recently attracted significant attention. The first part of this thesis aims to investigate current works of UAV-aided wireless communications and develop novel methods for both the placement and path design of a UAV as a flying RAN in wireless networks. We highlight how the use of city 3D maps can bring substantial benefits for the reliable self-placement of flying radios.Regardless of the placement or path design, all algorithms operate on the basis of an array of information such as node GPS location, the city 3D map, etc. allowing the prediction of radio signal strengths. While such data may be collected via the network before the actual UAV flight, part or all of the information may also have to be learned by the UAV. In this regard, a part of this thesis is devoted to discussing how to learn such information from the UAV-borne measurements.Assuming the availability of safe cellular connectivity, UAVs are becoming promising for a wide range of applications such as transportation, etc. The main challenge in these areas is the design of trajectories that guarantee reliable cellular connectivity all along the path while allowing the completion of the UAV mission. Hence, in the second part of this thesis, we propose a novel approach for optimal path design between an initial and terminal points by leveraging on a coverage map. Lastly, we discuss the experimental verification of the placement algorithm of a UAV relay in LTE networks.