Aerial 6D Movable Antenna-Enabled Cell-Free Networks

In this paper, we propose an aerial movable antenna (AMA) architecture with six-dimensional (6D) spatial degrees of freedom (DoFs) to enhance the capacity of cell-free networks. Unlike conventional terrestrial access points (APs) with fixed-position antennas, the proposed AMA-enabled APs can flexibly adjust the three-dimensional (3D) unmanned aerial vehicle (UAV) positions and 3D array rotations. To overcome the high-dimensional movement-design challenges posed by multiple APs and antennas, we develop a low-overhead, low-complexity, and scalable distributed processing optimization framework to maximize the achievable uplink sum-rate of the proposed aerial cell-free network. Specifically, the team minimum mean square error (TMMSE) algorithm is proposed to design receive combiners using partial channel state information, while the weighted critic update multi-agent twin-delayed deep deterministic policy gradient (WCU-MATD3) algorithm efficiently optimizes 3D positions and 3D rotations through distributed AP collaboration. Simulation results demonstrate that the proposed AMA-enable scheme achieves a 37.1% performance gain over conventional cell-free networks with fixed antenna position and rotation by exploiting 6D spatial DoFs, while the proposed algorithm exhibits satisfactory convergence and exploration capabilities.