Study Of Rough Sea Surface Electromagnetic Scattering Method

In recent years, reseaches on electromagnetic scattering of rough surfaces have rapidly developed. It has widely application in radar detecting, remote sensing, and communication and so on. Research methods of electromagnetic scattering of rough surfaces include numerical method and approximation method. Approximation methods such as Kirchhoff approximation, small perturbation model (SPM), small slope approximation (SSA) and so on. These methods are simple and quick but low precision. Numerical methods such as method of moments (MoM), FDTD, Multilevel-Fast Multipole Algorithm (MLFMA) and so on, they are precise and the simulations are more exact, but these kinds of methods usually have large calculation.In this paper, we introduce the electromagnetic scattering theory such as SPM, SSA mainly and focus on the electromagnetic scattering application of the fist-order Small Perturbation Model (SPM) and the fist-order Small Slope Approximation (SSA) on small rough sea surfaces. Comparing the two models with the ensemble averaged scattering coefficient (the theoretical value), the results show that the SPM has better performance at slightly rough surfaces. Then, MLMFA which is usually can be regarded as reference in EM scattering is applied to validate the effectiveness of SPM. It shows that at the medium incident angles, the SPM results agree very well with that of MLFMA. This proves that SPM can be applied in small rough sea surface scattering at the medium incident angle.A Markov random field (MRF) is employed to leran the statistical properties of an equivalent Huygens current situated above a rough sea surface in this paper, the Huygens current reprsents near-field electromagnetic scattered fields, from which the MRF parameters are learned. Once the MRF parameters are learned, the MRF may be used to synthesize equivalent planar surface currents of arbitrary size. Through the free-space Green's function, the Huygens current can be used to compute the far-field scattered fields.The advantages of this approach are manifested in the fact that once the MRF parameters are learned, a forward model no longer needs to be applied to generate random instantiations of the far-field scattered fields, the forward model need only be capable of handling surfaces of sufficient size for learning the MRF parameters. Moreover, the MRF may be applied to generate equivalent surfaces of size beyond the capability of existing numerical models.