A recent advance in the use of the finite-difference time-domain (FDTD
) method has been the introduction of the perfectly matched layer (PML
) to act as the absorbing boundary condition, This letter suggests usi
ng fictitious magnetic and electric displacement (as opposed to electr
ic field) conductivities in order to better isolate the PML from the r
est of the FDTD problem, It further describes the implementation for t
he PML directly from the FDTD formulation itself, This results in an a
nalysis that is much easier to understand and to program.