| dc.description.abstract | Unmanned Aerial Vehicles (UAVs) have been used to provide aerial networks by mounting base stations on them, due to advantages such as superior Line of Sight (LoS), fast deployment and flexibility when operating them, to mention but a few. However, need has arisen to make use of UAVs in applications without the need of human control, or deployment in a centralized network i.e. decentralization. In this report, we employ the use of deep reinforcement learning, a machine learning method, to achieve this decentralization property, where the UAV is placed in an environment it has never seen and expected to navigate it successfully, reaching the ground terminals (i.e. mobile phone users) and finding the optimal position to provide a good Quality of Service (QoS), depending on their distribution. | en_US |
