The Optimization of Dispersion Properties of Photonic Crystal Fibers Using a Real-Coded Genetic Algorithm

A real-coded genetic algorithm (GA) combined with a fully vectorial effective index method (FVEIM) is employed to design structures of photonic crystal fibers (PCFs) with user defined dispersion properties theoretically. The structures of PCFs whose solid cores are doped GeO₂ with zero−dispersions at 0.7–3.9 μm are optimized and the flat dispersion ranges through the R+L+C band and the negative dispersion is −1576.26 ps⋅km⁻¹⋅nm⁻¹ at 1.55 μm. Analyses show that the zero−dispersion wavelength (ZDW) could be one of many ZDWs for the same fiber structure; PCFs could alter the dispersion to be flattened through the R+L+C band with a single air-hole diameter; and negative dispersion requires high air filling rate at 1.55 μm. The method is proved to be elegant for solving this inverse problem.