Broad-band detection based on two-dimensional mixed autoregressive models

Two detectors for broadband plane waves in colored noise fields that use a linear array are described. The signal spectrum, arrival angle, and noise field coloration are not known a priori. Using a two-dimensional mixed autoregressive model, a generalized-likelihood-ratio test and its weak-signal equivalent, the Rao test, are derived. The performances of the proposed detectors are shown by computer simulations to be superior to those of conventional Fourier-based methods. Detection results for real data are presented