Implementation of CPML for One-Step Leapfrog WCS-FDTD Method

In this letter, the convolutional perfectly matched layer (CPML) is modified and implemented for one-step leapfrog weakly conditionally stable finite-difference time-domain (WCS-FDTD) method. Since the WCS-FDTD method can be considered as an approximate factorization of the Crank—Nicolson scheme, the derived equations of CPML involve the calculation of auxiliary terms at two time steps which are different from that of conventional FDTD CPML method.

To verify the effectiveness of proposed CPML, numerical investigation is performed and it is shown that, even with large Courant-Friedrich-Levy number (CFLN) chosen, the reflection error is still small and the electromagnetic field is stable.