摘要

A new and effective optimization approach to the inverse design problems of complex long-period fiber grating (LPG) filters was developed in the present study. The proposed synthesis method was based on the overlapped Gaussian-apodization method and the evolutionary optimization algorithm which can efficiently search for optimal solutions and simultaneously take into account various experimental requirements for the fabrication of the designed filters. To verify the effectiveness of the proposed method, a LPG filter for flattening EDFA gain spectra was designed. Compared to the existing results from discrete layer-peeling (DLP) inverse scattering algorithms, an LPG filter with adaptive grating lengths and much simpler, smoother, and less complicated coupling coefficient profiles for taking the practical fabrication conditions for gain flattening into account, was used in the proposed method. Simulation results confirm that optimal solutions of an LPG filter design are suitable for practical fabrication.

  • 出版日期2006-6-15