Co-array Processing Assisted Bayesian Beamforming (CABB): A Nonlinear Beamforming Technique for Joint Aerial Layer Network (JALN) Backbone

摘要

In this paper we describe a novel non-linear digital beamforming technique that we refer to as Co-array processing assisted Bayesian Beamforming (CABB) for directional line of sight networks like USAF Joint Aerial Layer Network (JALN) backbone. CABB leverages recently proposed signal processing techniques that can extract directions of arrivals of O(N2) sources using an antenna array that has only O(N) elements with non-uniform inter-element separation. The benefits of CABB for directional LOS networks are (1) reduced size of antenna array, (2) improved spectral efficiency (3) adaptation to changing channel conditions as a result of mobility. The main disadvantage in comparison to linear digital beamforming techniques is increased computations. In this paper we validate these tradeoffs via high fidelity simulations. We also argue that since most practical USAF directional LOS networks are sparse, it is possible to realize the CABB technique using existing hardware despite high computational requirements.