RATIONAL APPROXIMATION AND CONTINUED-FRACTION APPROXIMATION APPROACHES FOR COMPUTATION EFFICIENCY OF DIFFRACTION SIGNALS

摘要

Methods and apparatuses for improving computation efficiency for diffraction signals in optical metrology are described. The method includes simulating a set of diffraction orders for a structure. A set of diffraction efficiencies is determined for the set of diffraction orders. A rational approximation or a continued-fraction approximation is applied to the set of diffraction efficiencies to obtain a rationally approximated set of diffraction efficiencies or a continued-fraction approximated set of diffraction efficiencies, respectively. A simulated spectrum is then provided.