Conventional distributed collaborative beamforming would generally wrestle with the challenges of slowed-down convergence rate, prohibitively high computational costs and inefficient communication when applied to large-scale networks equipped with massive arrays. To address these challenges, we herein reformulate the distributed collaborative beamforming problem from the tensor perspective. By exploiting the inherent algebraic structure present in the problem, we develop a fully distributed collaborative beamforming algorithm incorporating the diffusion scheme for arrays endowed with the property of multi-linear translation invariance (MLTI). We also derive the convergence constraint of the proposed algorithm. Illustrative simulations validate the superior performance of the proposed algorithm, with notably accelerated convergence rate, reduced computational complexity and enhanced communication efficiency.

当应用于配备有大规模阵列的大规模网络时，传统的分布式协作波束成形通常会面临收敛速度减慢，计算成本过高以及通信效率低下的挑战。为了解决这些挑战，我们在这里从张量角度重构了分布式协作波束成形问题。通过利用问题中存在的固有代数结构，我们开发了一种完全分布式的协作波束成形算法，该算法结合了具有多线性平移不变性（MLTI）特性的阵列的扩散方案。我们还导出了所提出算法的收敛约束。说明性仿真验证了所提出算法的优越性能，并显着加快了收敛速度，