Improving the ability of a distributed feedback laser transfer-matrix model to fit to spectra from distributed-feedback lasers

Fitting distributed-feedback (DFB) laser models to data is a difficult process that is time consuming and often inaccurate. Having a good initial estimate of the facet parameters greatly reduces the time required by a fitting routine. If the general effects of laser parameters are well known, fitting routines do not have to be overly complicated to obtain a best fit. It is also very important to fit to spectra from both facets simultaneously and over a wide bandwidth for quick, accurate, and nondegenerate results.