A GLRT-like CFAR detector for heterogeneous environments

In this paper, we propose a new solution for the detection problem of coherent target in heterogeneous environments. Specifically, first assume that clutter returns from different range bins share same covariance structure but power levels. This model meets experimental evidence related to non-Gaussian and non-homogeneous scenarios. Then, unlike existing solutions are based upon estimate plug methods, an approximation generalized likelihood ratio test where maximizers likelihoods obtained through alternating estimation procedure. Remarkably, also prove such procedure leads architecture possessing constant false alarm rate (CFAR) when specific initialization is used. The performance analysis, carried out on simulated as well measured data comparison with suitable well-known competitors, highlights proposed can overcome CFAR competitors exhibits limited loss respect other non-CFAR detectors.