Distributed Channel Estimation and Optimization for 6D Movable Antenna: Unveiling Directional Sparsity

Six-dimensional movable antenna (6DMA) is an innovative and transformative technology to improve wireless network capacity by adjusting the 3D positions and rotations of antennas/antenna surfaces based on the channel spatial distribution. To achieve optimal antenna positions and rotations, acquiring statistical channel state information (CSI) is essential for 6DMA systems. However, existing works assume that a central processing unit (CPU) jointly processes the signals of all 6DMA surfaces. This incurs prohibitively high processing cost and latency for channel estimation due to the vast numbers of 6DMA candidate positions/rotations and antenna elements. Therefore, we propose a distributed 6DMA processing architecture to reduce the processing complexity of the CPU by equipping each 6DMA surface with a local processing unit (LPU). Furthermore, we unveil for the first time the directional sparsity property of the 6DMA channels with respect to distributed users, where each user has significant channel gains only for a (small) subset of 6DMA position-rotation pairs. Based on this property, we propose a practical three-stage protocol for the 6DMA system and corresponding algorithms to conduct statistical CSI acquisition for all 6DMA candidate positions/rotations, 6DMA position/rotation optimization based on statistical CSI, and instantaneous CSI estimation for user data transmission with optimized 6DMA positions/rotations. Simulation results show that the proposed channel estimation algorithms achieve higher accuracy than benchmark schemes, while requiring lower pilot overhead. Moreover, the proposed 6DMA system with statistical CSI-based position/rotation optimization achieves a higher ergodic sum rate than fixed-position and fluid antenna systems, even if the latter have perfect instantaneous CSI.