Abstract

Unmanned aerial vehicles (UAVs), popularly known as drones,
have emerged as a promising solution for providing reliable and cost-effective wireless communications.
The use of UAVs as aerial wireless power transmitters (UAV-WPT), with additional flexibility and 3D mobility, is expected to provide efficient wireless power supplies to low-power and hard-to-reach devices. Due to their adjustable altitude and mobility, efficient line-of-sight (LoS) between UAVs and ground nodes (GNs) could be established, thus mitigating signal blockage and shadowing. Based on this feature, UAVs can be good candidates to charge battery-limited or hard-to-reach devices through radio frequency (RF) wireless power transfer (WPT), which will significantly improve the wireless charging efficiency compared to conventional ground charging stations at fixed locations. Although the deployment of UAVs as wireless power transmitters is promising, it comes with many design challenges and reliability problems. For instance, the energy efficiency (EE) of UAVs requires careful consideration as it significantly impacts the performance of UAV-WPT systems. Thus, there is a need for a comprehensive framework to optimize such networks, where the devices are wirelessly powered via UAVs to enable uplink data transmission.
In this thesis, we propose a detailed methodology to optimize the performance of the UAV-enabled WPT networks with different topologies and applications. We provide the required steps to be followed for most applicable networks, where specific considerations have to be considered for each case. The optimization problem's solution has two main steps; firstly, the path loss of the air-to-ground channels is minimized by optimizing the UAV position depending on the GNs' service demands. Secondly, using the optimized positioning and a closed-form expression for the EE, a resource allocation aiming to maximize EE is developed using Lagrangian optimization and gradient-descent methods.

We present five different system models, which reflect different practical cases and setups considering single and multiple UAV scenarios. These models are: UAV-enabled wireless powered communications network (UAV-WPCN), UAV-enabled wireless information and power transfer network (UAV-WIPT), UAV-enabled simultaneous wireless information and power transfer network (UAV-SWIPT),
multiple UAV-enabled wireless powered communications network (UAVs-WPCN), and multiple UAV-enabled simultaneous wireless information and power transfer network (UAVs-SWIPT). The results of applying the proposed scheme show significant enhancement in the EE for the non-orthogonal multiple access (NOMA) scheme compared to the orthogonal multiple access (OMA) scheme in most of the scenarios. However, the topology and distribution of the ground nodes play a vital role in figuring out the suitable access scheme to be used, where OMA or hybrid NOMA/OMA schemes could perform better.