Random-process formulation of computationally efficient performance measures for wideband arrays in the far field

Random-process autocorrelations and crosscorrelations in three dimensions here provide computationally efficient mean-square-error and average-gain measures to control main beam and sidelobe characteristics in convex optimization of far-field beamforming coefficients for large wideband arrays. An example optimized with the ultra-efficient interior-point code LOQO illustrates.