GLRT-based adaptive detection-estimation of Gaussian sources in coloured noise fields

Detection-estimation of uncorrelated point Gaussian sources in spatially coloured noise is considered for uniform linear arrays. We assume that a secondary (training) dataset that contains only the coloured noise samples is provided in addition to the primary dataset that has samples with a mixture of coloured noise and signals. We propose a GLRT (generalised likelihood-ratio test) technique based on likelihood-ratio (LR) optimisation.