Integrated Optical Devices Design by Genetic Algorithm

Abstract

In this work we use multiple scattering in conjunction with a genetic algorithm to reliably determine the optimized photonic-crystal-based structure able to perform a specific optical task. The genetic algorithm operates on a population of candidate structures to produce new candidates with better performance in an iterative process. The potential of this approach is illustrated by designing a spot size converter that has a very low F-number (F=0.47) and a conversion ratio of 11:1. Also, we have designed a coupler device that introduces the light from the optical fiber into a photonic-crystal-based waveguide with a coupling efficiency over 87% for a wavelength that can be tuned to 1.5 microns.