Design of Multilayer Polarization Beam Splitters Using Multiobjective Evolutionary and Genetic Algorithms

Abstract

Multilayer polarization beam splitters (MPBSs) are important optical components in optical communication systems and optical devices. The components of transverse electric (TE) and transverse magnetic (TM) light plane wave in MPBSs are separated in different ways. In this paper, the Pareto optimization (multiobjective evolutionary and genetic algorithms) is used to design MPBSs on C-band frequency. The reflectances of TM light plane wave for Pareto optimization are the lowest and zero at the wavelength of 1.535 μm. The degree of polarization for Pareto optimization is the highest at wavelength from 1.528 - 1.54 μm and reaches 100% at wavelength of 1.535 μm.