The Fourier transform method to design graded-index optical filters, that relates the desired reflection spectrum and the index profile through the use of a Q function, has two important drawbacks: (1) It relies on approximate Q
functions, and (2) it does not account for the dispersion of the index of refraction.
The former is usually addressed by an iterative correctionprocess.
We propose to address the latter by scaling the wavelength in the Fourier transform by the optical thickness of the filter and to multiply the Q
function by a wavelength-dependent correction factor. We demonstrate the high effectiveness of this approach by the performance of optical filters designed with such correction factors using the optical properties of
© 2007 Optical Society of America
Equations on this page are rendered with MathJax. Learn more.