Band-diagonal regularization of Gaussian interference covariance matrices ML estimates

The convergence rate of a number of adaptive processing algorithms for coherent signal extraction from Gaussian interference is compared. Our main attention is paid to a relatively new hybrid method of regularization of the maximum-likelihood estimate (MLE) of the covariance matrix (CM) of Gaussian vectors with discrete and continuous power spectra. Relatively high convergence rate and efficient lattice-filter-based implementation for the introduced technique are demonstrated.