Spatial and Spectral Filtering as Applied to Array Processing.

摘要

Algorithms have been developed that can be flexibly applied to the processing of discrete directional interference, out-of-band look-direction interference, and in-band look-direction interference, all of which are frequently encountered in array processing. Discrete directional interference is handled by a recursive adaptive LMS processor which steers linear array pattern nulls to spatially filter the interference. The form of the processor is controlled by choosing the elements of a selection matrix in its interference estimation channel. The selection matrix chooses subsets of the input array for use in forming an estimate of the interference which is subtracted from the output of a uniform array which is used to focus on the look-direction waveforms. Look-direction waveforms remain undistorted whether using a constrained or unconstrained form of the spatial processing algorithm. Out-of-band look-direction interference is processed by choosing the appropriate filter order out of a set of preprogrammed complex digital Butterworth filters. In-band look-direction interference is handled by an LMS processor that uses time correlations on an FIR to steer a spectral notch to perform the required filtering. The processor may be implemented by either a direct algorithm that correlates the time samples and uses a recursive algorithm to solve the correlation equations involved or by a recursive adaptive algorithm. The properties of each of the processors are developed and performance capabilities illustrated.