Magneto-sensitive photonic crystal optical filter with tunable response in 12–19 GHz; cross over from design to prediction of performance using machine learning

Abstract

Photonic crystal structure made of Si, SiO2 and a plasma defect layer is suggested for optical filter applications. Generation of stop band and resonant modes localizations at transmittance spectrum of the structure are studied by exploring the dependences of the response on the period number of the layers and the electron density of the plasma slab. Also, the role of the magnetically induced birefringence in the anisotropic defect layer and the light incident angle on the optical features of crystal are studied. Results show that the response of the proposed structure could be tuned remarkably in practice by the selected structural as well as external parameters. Furthermore, an artificial intelligence (AI) approach is proposed to predict the final response of the suggested crystal in various input circumstances. Very significant concordance between results of the AI and transfer matrix method (TMM) approaches is observed.