SPECTRAL-ITERATIVE ANALYSIS OF ELECTROMAGNETIC RADIATION AND SCATTERING PROBLEMS (RADAR-CROSS-SECTION, INTEGRAL EQUATION)

Integral equations in electromagnetics typically are convolutional in nature. When the geometry of the problem permits, the unknown function can be expanded in a set of identical, evenly spaced basis functions. If the integral equation is tested in the same way, the continuous convolutions reduce to discrete convolutions. These operations can then be readily computed by the Fast-Fourier Transform algorithm resulting in a significant reduction in computer time. This spectral domain method of calculating convolutions is incorporated into an iterative algorithm to provide a numerically efficient means of solving integral equations. This thesis develops and applies two such iterative techniques.;The second iterative technique used in this thesis is the method of conjugate gradients. It is applied to resistively edge-loaded strips and plates. The scattering properties of these structures are studied in detail and their application to radar-cross-section reduction is demonstrated. The iterative nature of the conjugate gradient technique, combined with the spectral domain calculation of convolutions, allows a large number of unknowns to be handled conveniently. Results are presented which involve over 3000 unknowns.;The first technique discussed is the Spectral-Iterative Technique. This method is applied to scattering from resistive strips and from metallic strips with resistive edge loading. The scattering behavior of these structures is of importance in low radar-cross-section applications where resistive and other materials are used to reduce the scattered field. An extension of the technique is also presented which is applicable to open waveguide discontinuity problems.