Multiresolution compound-Gaussian models for sea clutter with independent identically distributed textures

Range resolution conversion (RRC) is a promising technology for maritime radar detection of targets of different size, and sea clutter modeling and parameter estimation are important in maritime radar signal processing. In this paper, two kinds of multiresolution compound-Gaussian models are proposed to model the clutter after RRC, which are multiresolution K (MRK) distribution and multiresolution CGIG (MRCGIG) model. As another main contribution, across-resolution methods of moments (ARMoM) are proposed to estimate the parameters of MRK and MRCGIG distributions, which have higher estimation accuracy than that of methods of moments. Moreover, the fitting performance of proposed models are verified by the clutter after RRC which is generated by simulating clutter data before RRC, and the comparisons of estimation performance base on simulated data demonstrate the superiority of ARMoM.