Diffuse Multipath Exploitation for Adaptive Radar Detection

We deal with the problem of detecting point-like targets in diffuse multipath environments, modeling the target echo as the superposition of a deterministic signal with an unknown scaling factor (due to the direct path) plus a zero-mean complex circular symmetric Gaussian random vector with an unknown covariance matrix (accounting for the echoes from the glistening surface). We devise a constrained Generalized Likelihood Ratio Test (GLRT) for the resulting hypothesis testing problem, enforcing the primary data covariance matrix (due to both interference and multipath echoes) to belong to a neighborhood of the secondary data sample covariance matrix. Remarkably, the proposed decision scheme ensures the desirable Constant False Alarm Rate (CFAR) property with respect to the unknown parameters of the interference. The performance assessment, conducted on simulated data in terms of detection probability also in comparison with existing solutions, highlights the effectiveness of the new approach to cope with diffuse multipath phenomena.