An Adaptive Beamformer Based on Adaptive Covariance Estimator

Based on the Minimum Variance Distortionless Response-Sample Matrix Inversion (MVDRSMI) method, we propose a novel Adaptive Covariance Estimator (MVDR-ACE) beamformer for adaptation to multiple interference environments. The MVDR-ACE beamformer iteratively determines a minimum number of data samples required while maintaining its average signal-to-interference-noise to be within 3 dB from the performance of a theoretical optimum MVDR beamformer and meeting an instantaneous interference cancellation requirement. Finally, based on numerical simulations, we analyze and validate the performance of the MVDR-ACE beamformer. We also compare its performance to the conventional MVDR-SMI beamformer that uses a fixed data sample in its covariance estimator.