Title :
Preconditioned iterative solution of scattering from rough surfaces
Author_Institution :
Sch. of Electr. & Comput. Eng., Oklahoma State Univ., Stillwater, OK, USA
fDate :
6/1/2000 12:00:00 AM
Abstract :
Extensions to the functionally identical forward-backward (FB) and method of ordered multiple interactions iterative techniques have been introduced that improve the convergence characteristics with specific scattering geometries. These extensions are shown to be mathematically equivalent to applying preconditioners to the discretized integral equation that is iteratively solved. The same preconditioners can be used with any iterative solution technique. Numerical examples show that the generalized minimal residual (GMRES) and bi-conjugate gradient-stable (BICGSTAB) algorithms give similarly rapid convergence when applied to a preconditioned discretized integral equation
Keywords :
conjugate gradient methods; electromagnetic wave scattering; integral equations; iterative methods; numerical stability; rough surfaces; EM wave scattering; bi-conjugate gradient-stable algorithm; convergence characteristics; forward-backward method; generalized minimal residual algorithm; iterative techniques; ordered multiple interactions; preconditioned discretized integral equation; preconditioned iterative solution; rough surfaces; scattering geometries; Convergence of numerical methods; Geometry; Integral equations; Iterative algorithms; Iterative methods; Jacobian matrices; Rough surfaces; Scattering; Surface roughness; Symmetric matrices;
Journal_Title :
Antennas and Propagation, IEEE Transactions on
