Evolutionary programming synthesis of optimal long-period fiber grating filters for EDFA gain flattening

Abstract

An innovative long-period fiber grating (LPG) synthesis method based on the stochastic evolutionary programming (EP) algorithms is demonstrated to be effective for designing optimal LPG filters. Synthesis of a single-stage LPG filter for the entire C-band erbium-doped fiber amplifier (EDFA) gain flattening is used as an example to show the feasibility and the effectiveness of the proposed algorithm. To the best of our knowledge, this is the first demonstration that good EDFA gain flattening for the entire band can be achieved with a single-stage LPG filter that is properly designed.