UAV-assisted NOMA Downlink Communications Based on Content Caching

In this paper, we investigate a UAV-assisted communication system based on data caching at the unmanned aerial vehicle (UAV). In the network, two ground users (GUs) request data stored in a local station (LS). However, direct links from the LS to the GUs are assumed to be unavailable due to unexpected or emergency circumstances such as natural disasters or complex terrains. Herein, a UAV is dispatched to cache the data in the LS and then transmit data to the two GUs by using non-orthogonal multiple access (NOMA) technique. In order to address the energy limitation issue, energy harvesting is applied in which the UAV can prolong its operation time by harvesting solar energy from the ambient environment. Besides, the UAV is assumed to move along a circular trajectory around a fixed center. We consider a joint caching and resource allocation algorithm to deal with the Markov decision process problem for maximizing the long-term data transmission rate of the system. In particular, we propose a novel scheme based on value iteration-based dynamic programming to obtain the optimal policy in which the UAV can allocate an optimal transmission power for the GUs in each time slot. The simulation results are presented to verify the effectiveness of the proposed method.