Yung P. Lee
Herbert A. Freese
John S. Hanna
Peter N. Mikhalevsky
Sci. Appl. Internatl. Corp., 1710 Goodridge Dr., MS T1-3-5, McLean, VA
Capon's maximum-likelihood method (MLM) has been used in many areas of signal processing, but it is known to be sensitive to mismatch. Several robust algorithms, such as the white noise constrained method and the ``reduced'' MLM, have been developed to reduce the sensitivity to mismatch. The objective of all techniques is to obtain an accurate measure of the energy originating in a specific resolution cell, whether this energy be considered signal or noise. In this paper, simulation results are presented for various adaptive matched-field processing methods used in conjunction with a tilted line array which have applications to both active and passive systems. In the simulations, the noise is modeled as having two components, one continuous, and the other discrete. The clutter is modeled as backscatter arising from many randomly distributed discrete sources. It is shown that adaptive matched-field processing, even with mismatch present, can provide significant suppression of both noise and reverberation in certain resolution cells, while at the same time maintaining almost unity gain for these cells.