Energy Efficiency Optimization for Multi-UAV-Assisted Cell-Free Aerial Access Networks

Abstract

Aerial radio access networks, enabled by unmanned aerial vehicles (UAVs), have emerged as a promising paradigm for extending wireless coverage and enhancing service continuity. Among various UAV platforms, fixed-wing UAVs offer advantages in coverage range, making them particularly suitable for scenarios lacking terrestrial infrastructure. This article investigates a UAV-based aerial access network based on cell-free systems, where multiple fixed-wing UAVs cooperatively serve user terminals (UTs) as distributed access points (APs). Despite their potential, two major challenges arise: 1) the trajectory design of multiple UAVs becomes increasingly complex due to their high mobility and coordination requirements, and 2) energy management becomes inefficient under conventional designs, especially considering the limited onboard energy of UAVs and their continuous propulsion requirements. To address these issues, this article proposes an energy-efficient joint trajectory and transmit power optimization algorithm to fully exploit the spatial degrees of freedom (DoF) of mobile UAV-APs. In particular, with fixed transmit powers, the nonconvex communication rate is approximated via first-order Taylor expansion, while the nonconvex energy consumption model is converted into a convex lower bound based on fixed-wing propulsion characteristics. The trajectory optimization problem is then reformulated into a concave-convex fractional programming (CCFP) framework to derive approximately optimal trajectories. Then, with fixed trajectories, the energy efficiency (EE) maximization problem reduces to a rate maximization problem, where the nonconvex communication rate is similarly approximated using a first-order Taylor expansion to optimize power allocation. Finally, an alternating optimization (AO) framework iteratively updates the trajectories and transmit powers to maximize EE while ensuring robust communication. Extensive simulation results demonstrate the superiority of the proposed method over benchmark schemes, validating its effectiveness in energy-efficient aerial radio access networks based on cell-free systems.