Detection and localization of multiple sources in noise with unknown covariance

A technique for detection and localization of multiple sources in the presence of noise with unknown and arbitrary covariance is presented. The technique is applicable to coherent and noncoherent signals and to arbitrary array geometry. It is based on Rissanen´s (1983) MDL minimum description length principle for model selection. Its computational load is comparable to that of analogous techniques for white noise. Simulation results demonstrating the performance of this technique are included