Reinforcement Learning-Based Joint Optimization of Transmit Power, CCA, and UAV-AP Deployment in Aerial Wi-Fi Network

Unmanned aerial vehicle (UAV)-based air-to-ground (A2G) Wi-Fi network have attracted considerable attention due to their potential to overcome the limitations of ground access points (APs) and the flexibility in network deployment. In this paper, we propose a hierarchical multi-agent Q-learning (HiMAQ) framework that jointly optimizes transmit power, clear channel assessment (CCA) threshold, and UAV-AP position to enhance sum throughput, fairness, and energy efficiency in A2G Wi-Fi network. Simulation results demonstrate that the proposed HiMAQ framework outperforms existing baseline models in terms of the proposed reward considering various performance measures such as sum throughput, number of outage users, Jain’s fairness index, and energy consumption.