An Improved Paralleling-in-Order Solving Scheme for AH-FDTD Method Using Eigenvalue Transformation

This paper describes a modification of the unconditionally stable finite difference time domain (FDTD) algorithm based on the Associated Hermite (AH) functions. The original method is involved with solving a banded nested matrix equation to calculate the AH expanding coefficients using lower-upper (LU) decomposition, which is the main cost for CPU time and memory storage. Here, the calculation of expansion coefficients is modified as a paralleling-in-order solving scheme by using the eigenvalue transformation, which greatly enhances the computational efficiency and reduces the memory storage. The numerical results demonstrate the performance of this improvement.