Fairness-Aware Cooperative Communication With Wireless-Powered UAVs Under Uncertain User Locations

Unmanned aerial vehicles (UAVs) have emerged as a promising solution to enhance the flexibility, adaptability, and coverage of mobile service provisioning in wireless communication systems. However, their limited onboard energy and the uncertainty of service target locations in practical environments pose significant challenges to achieving reliable and fair quality of service (QoS). In this article, we study the problem of fairness-aware UAVs cooperative communication (FUCC), where multiple wireless-powered UAVs cooperatively plan their trajectories to search for and serve users without any prior knowledge of their locations, aiming to achieve long-term fairness under energy constraints. To solve this problem, we first formulate it as a multiobjective optimization (MOO) problem and then transform it into a partially observable Markov decision process (POMDP). Then, we propose a deep reinforcement learning (DRL)-based cooperative control framework for UAVs to achieve long-term fairness and communication efficiency in dynamic and uncertain environments. In the algorithm, we introduce an inter-UAV communication mechanism with an observation-sharing strategy and an advising scheme to address suboptimal cooperation under environmental uncertainty and partial observability. In addition, to enable UAVs to make charging decisions, we independently design discrete and continuous action modules in each UAV’s decision network to handle flight control and charging decisions at each timeslot. Simulation results show that our proposed algorithm significantly improves the communication service fairness by 22.41% and service efficiency by 20.04% compared to state-of-the-art DRL-based baselines.