This paper introduces a compressive sensing approach for single-snapshot adaptive beamforming. The observation data model is considered as source components in additive noise, and then a compressive sensing formulation is introduced to estimate the parameters of the interference signals. That is, a LASSO regression problem is proposed and solved, yielding the directions as well as the powers of the interference signals. On the other hand, the noise power is estimated by means of averaging the squares of the difference between the observation data and the estimate of the source components. Finally, the interference-plus-noise covariance matrix is reconstructed and used for adaptive beamformer design. Simulation results show better performance of the proposed beamformer than several existing beamformers, in the case of a single snapshot.

Adaptive beamforming; single snapshot; compressive sensing; covariance matrix reconstruction; sparsity