A BMIA/FFT algorithm for the Monte Carlo simulations of large scale random rough surface scattering: application to grazing incidence

A banded matrix iterative approach (BMIA) has been developed to study electromagnetic scattering from one-dimensional random rough surfaces. This new method not only provides a faster solution time, but it also allows one to analyze much longer surfaces that could not be considered before using the conventional integral equation method. The capability of analyzing random surfaces of large surface length is essential in the study of backscattering at near grazing incident angle. We report our refinement of the BMIA. We first identify the inefficient part of the algorithm and then provide a remedy to alleviate the problem which results in further reduction in solution time as well as computer memory requirement. We then apply the new improved BMIA to investigate backscattering from random rough surface near grazing incidence.