Title: UAV Deployment Optimization in UAV-assisted Wireless Communications

Abstract: Due to the fact that the locations of base stations (BSs) cannot be changed after they are installed, it is very difficult to communicate directly with remote user equipment (UE), which will directly affect the lifespan of the system. Unmanned aerial vehicles (UAVs) offer a hopeful solution as mobile relays for fifth-generation wireless communications due to the flexible and cost-effective deployment. However, with the limited onboard energy of UAV and slow progress in energy storage technology, it is a key challenge to achieve the energy-efficient communication. Therefore, in this article, we study a wireless communication network using a UAV as a high-altitude relay, and formulate a UAV relay deployment optimization problem (URDOP) to minimize the energy consumption of system by optimizing the deployment of UAV, including the locations and number of UAV hover points. Since the formulated URDOP is a mixed-integer programming problem, it presents a significant challenge for conventional gradient-based approaches. To this end, we propose a self-adaptive differential evolution with a variable population size (SaDEVPS) algorithm to solve the formulated URDOP. The performance of proposed SaDEVPS is verified through simulations, and the results show that it can successfully decrease the energy consumption of system when compared to other benchmark algorithms across multiple instances.