Deep Reinforcement Learning for Resource Sharing and UAV Trajectory Optimization in Multi-Operator UAV-Assisted Wireless Networks

Multi-operator unmanned aerial vehicle (UAV)-assisted wireless networks introduce new challenges in coordinating aerial base stations, ensuring fair resource sharing, and dynamically adapting to user mobility. Most existing studies either focus on single-operator scenarios or do not account for fairness in inter-operator load exchange. To address these limitations, we propose a hierarchical deep reinforcement learning framework that jointly optimizes inter-operator resource sharing and UAV trajectory design. At the operator level, we introduce a novel load exchange balance metric to ensure equitable cooperation and prevent disproportionate benefits. At the UAV level, deep reinforcement learning enables autonomous three-dimensional (3D) trajectory control to enhance throughput and reduce outages. Simulation results demonstrate that the proposed framework significantly outperforms benchmark schemes in terms of fairness, user rate, throughput, and outage, thereby validating its effectiveness in complex multi-operator environments.