Distributed Collaborative Tensor Beamforming via Gaussian Entropy Over Array Networks

Distributed collaborative beamforming has been intriguing considerable interest, whereas slow convergence would hinder practical applications of the conventional algorithms to networks of massive arrays. When non-circular complex-valued signals are involved, the conventional distributed beamforming algorithms would be sub-optimal. We consider herein adaptive diffusion approaches to enhance the beamforming performance for networks of massive (vector-sensor) arrays in potentially non-circular measurement noise scenarios. Formulating the tensorial array signal model, we develop a global adaptive tensorial beamforming algorithm and the corresponding distributed counterpart leveraging the minimum Gaussian entropy criterion for networks of massive arrays. The convergence behavior of the proposed algorithms in the small step-size regime is theoretically and experimentally demonstrated. Illustrative simulations validate the superior performance of the proposed algorithms, especially in the non-circular measurement noise scenarios.