RCS computation of large inhomogeneous objects using a fast integral equation solver

In this paper, we apply a fast Fourier transform (FFT) accelerated volume integral equation solver to compute the radar cross section of large-scale inhomogeneous objects. This method is related to Bojarski's k-space method and the subsequent conjugate-(CG) and biconjugate-gradient (BCG) FFT algorithms. The method developed here combines the weak-form discretization with the BCG and stabilized BCG (BCGS) solvers. We show that this method has marked improvements over related algorithms. The numerical method has been validated by Mie series for multilayer spheres and applied to some practical problems. With this method we are currently able to solve a three-dimensional problem with a volume of size 3648/spl lambda//sup 3/ (21.23 million unknowns) on a single workstation.