Non-Coherent CFAR Detector Using Compound Gaussian Clutter

In practical, the problem of radar signal detection is to automatically detect a target embedded in clutter. For high resolution radars, the modeling of sea clutter showed that compound Gaussian distributions are appropriate to describe the clutter returns. In this paper, we introduced a novel Constant False Alarm Rate (CFAR) detector in a non-coherent context, where the clutter follows a non-Gaussian distribution. The simulations via Monte Carlo showed that this new detector is robust for three Compound Gaussian (CG) clutter models; namely the K distribution, Compound Gaussian with inverse gamma texture (Generalized Pareto model, GP) and Compound Inverse Gaussian (CIG) distribution. The false alarm regulation was then examined within the presence of interfering targets. Finally, the performance of the proposed algorithm was validated using real data sea clutter