An Application of Sampling Theorem to Moment Method Simulation in Surface Scattering
The method of moment (MoM) is a numerical procedure for solving the equation Lf = g, where L is a continuous linear operator, f is the unknown function to be determined and g is a known excitation. In essence, the procedure converts the linear equation into a matrix equation and determines the unkno...
Saved in:
Published in | Journal of electromagnetic waves and applications Vol. 20; no. 4; pp. 531 - 546 |
---|---|
Main Authors | , |
Format | Journal Article |
Language | English |
Published |
Zeist
Taylor & Francis Group
01.01.2006
VSP |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The method of moment (MoM) is a numerical procedure for solving the equation Lf = g, where L is a continuous linear operator, f is the unknown function to be determined and g is a known excitation. In essence, the procedure converts the linear equation
into a matrix equation and determines the unknown function in the form of a set of samples of the function through matrix inversion. It is well known that the sampling theorem states that to completely represent a function, one requires a minimum of two samples per wavelength, when the basis
function is the Sinc function. Thus, by choosing the Sinc function as the testing and basis function in MoM implementation, we should have the smallest possible size for the matrix. To test the above stated approach, we calculate the scattering coefficient from a 2-D rough surface with a Gaussian
distribution. The standard point matching approach is also used so that comparisons can be made between the two results. By comparing these numerical calculations with the IEM model, it is found that to achieve the same degree of accuracy, the number of unknowns used in the point matching
method is an order of magnitude more than when the Sinc function is used as the basis function. This reduction in matrix size is specially significant in a three dimensional surface scattering problem, because we shall be able to reduce the matrix size by two orders of magnitude |
---|---|
ISSN: | 0920-5071 1569-3937 |
DOI: | 10.1163/156939306776117063 |