Higher-Order Approximate CN-PML Theory for Magnetized Ferrite Simulations

Two unconditionally stable implementations of the higher-order perfectly matched layer are proposed for the modeling of magnetized ferrite in the finite-difference time-domain lattice. By incorporating the approximate Crank-Nicolson (CN) algorithm and the modified auxiliary differential equation (ADE) approach, the proposed implementations take full advantage of the CN methods in terms of reducing the computational time and improving the computational efficiency. Approximate CN algorithms including the Crank-Nicolson-Douglas-Gunn and the Crank-Nicolson Approximate-Decoupling schemes are implemented in 2D simulation. Furthermore, based on the ADE method, an alternative method at the integer time step is proposed to analyze the anisotropic magnetized ferrite structures. Its computational efficiency can be further enhanced compared to the shift operator method from the previous works. A full-filled ferrite model and a ridge waveguide structure are introduced to illustrate the effectiveness and efficiency of the proposed algorithms. The results show that the proposed algorithms can improve the computational efficiency, overcome the Courant-Friedrich-Levy limit, and obtain considerable absorbing performance.