Optimal Deployment of UAV-Mounted Reconfigurable Intelligent Surfaces for Millimeter-Wave Transmission

We study the deployment of reconfigurable intelligent surfaces (RIS's) on a cellular area. The RIS's are mounted on unmanned aerial vehicles that hover on the ground users at a fixed altitude. Severe blockage in millimeter-wave transmission is alleviated by the artifical Line-of-Sight (LoS) created by the UAV-RIS's. We formulate the optimal deployment of UAV RIS's as a binary linear optimization problem. We compare the performance of this solution with that of the particle swarm optimization (PSO)-based solution. We also compare the performance with that of the solution wherein the RIS locations are constrained to be equispaced and with equal distance from the base station. Results reveal that binary optimization leads to a good performance with much less run time. Finally we form a dataset and experiment with neural network-based solutions in order to obtain the optimal solution.