Authors: Chao Li and Raj Mittra
Source: FERMAT, Volume 19, Communication 2, Jan-Feb., 2017
Abstract: In this presentation, parallel multilevel characteristic basis function method (MLCBFM) has been hybridized with the randomized pseudo-skeleton approximation method (RSPA) and randomized singular value decomposition (rSVD) method, for the analysis of the scattering from electrically large rough surfaces at low grazing angles. MLCBFM defines the Characteristic Basis Functions (CBFs) on a larger domain and thereby achieves a higher compression rate. The reduced matrix has a much smaller size compared to that of the original system, which enables us to use a direct solver, rather than an iterative one. The RSPA algorithm accelerates the generation of the reduced matrix while the rSVD expedites the generation of CBFs. The hybrid method is found to be both accurate and efficient, and we use it in this work to investigate the problem of bistatic scattering from Gaussian and exponential rough surfaces at low grazing angles.
Index Terms: Characteristic basis Functions (CBFs), Rough Surface, Scattering, rSVD, RSPA.
View PDFSolution of Electrically Large Scattering problems using the Characteristic Basis Function Method