ABSTRACT

In recent years, Unmanned Aerial Vehicle technology (UAV) has gained significant attention in modern telecommunications networks due to the various capabilities it offers. This paper investigates the joint optimization of the three-dimensional locations of UAV, user grouping on the ground, and power allocation for users in the uplink, with the aim of maximizing energy efficiency. At the same time, the users' quality of service (QoS) requirements are also considered. In this study, UAV is utilized as mobile base stations (BS), and information transfer occurs in the physical layer over non-orthogonal multiple access (NOMA).The main idea of this research is to present a communication system that: 1) combines the advantages of UAV communications with the capabilities provided by non-orthogonal multiple access, 2) maximizes energy efficiency, and 3) employs a Q-learning algorithm to solve the optimization problem. The performance analysis of the proposed framework demonstrates the effectivenessof optimal resource allocation approach. Numericalresults indicate that the proposed scheme, which includes optimal UAV deployment, user grouping, and optimal power allocation,significantly improves user energy efficiency compared to random grouping and uniform power allocation methods. Furthermore, results show that the Q-learning algorithm quickly converges to the optimal solution after the training phase.