Stochastic optimization of low-loss optical negative-index metamaterial

Optical metamaterial consisting of metal-dielectric composites creates a complicated system that is not amenable to analytical solutions. This presents a challenge in optimizing these intricate systems. We present the application of three nature-inspired stochastic optimization techniques in conjunction with fast numerical electromagnetic solvers to yield a metamaterial that satisfies a required design criterion. In particular, three stochastic optimization tools (genetic algorithm, particle swarm optimization, and simulated annealing) are used for designing a low-loss optical negative index metamaterial. A negative refractive index around -0.8+0.2i is obtained at a wavelength of 770 nm. The particle swarm optimization algorithm is found to be the most efficient in this case.