Practical boundary condition for finite difference time domain simulations

Abstract

We propose a practical finite difference time domain (FDTD) truncation boundary condition that is simple to implement and requires less computing resources as compared to standard absorbing boundary conditions. Its performance in terms of speed and reflection error is compared to the uniaxial perfectly matched layer (UPML) implementation. Although less effective than the UPML in suppressing reflections, simulations with our boundary conditions are up to 3 times faster and can save up to half the memory.