Data-Driven Based Positioning Technique for UAV Aided NOMA System

Unmanned aerial vehicles (UAVs) have attained prevalent attraction in the mobile networks as a reliable technique which can improve the network capacity and provide efficient communications for ground users during emergency situations. Using UAVs in conjunction with non-orthogonal multiple access (NOMA) can greatly improve the performance of the overall network. In this paper, we study the maximization of the overall achievable cell sum rate in a UAV-aided NOMA network by optimizing UAV positioning vector using the real-time observations. We propose a low complex model-free data driven based approach to find a near-optimal UAV positioning vector in a single cell NOMA system. The proposed approach is based on a dynamic linearization data model with a time-varying pseudo gradient parameter. Numerical simulations show that the proposed algorithm provides the performance very close to the exhaustive search algorithm with low computational complexity. The simulation results show that the proposed algorithm provides the performance very close to the exhaustive search algorithm with low computational complexity.