Abstract:A method for adaptively selecting training data is proposed to improve the performance of adaptive beamforming. The method first measures the contribution of each snapshot to the covariance matrix required by the beamforming using the sparse iterative covariance-based estimation technique. Then, those snapshots making larger contributions are selected as the final training samples. The homogeneity of training samples can be improved significantly, and thus results in an evident performance improvement in adaptive beamforming. Simulation results demonstrate the effectiveness of the proposed method.

Training data selection method for adaptive beamforming