Deep neural network-based automatic metasurface design with a wide frequency range

Abstract Beyond the scope of conventional metasurface, which necessitates plenty computational resources and time, an inverse design approach using machine learning algorithms promises effective way for metasurface design. In this paper, benefiting from Deep Neural Network (DNN), procedure a in ultra-wide working frequency band is presented output unit cell structure can be directly computed by specified target. To reach highest training DNN, we consider 8 ring-shaped patterns to generate resonant notches at wide range frequencies 4 45 GHz. We propose two network architectures. one architecture, restrict so only input patterns. This drastically reduces while keeping network’s accuracy above 91%. show that our model based on DNN satisfactorily with average over 90% both Determination without time-consuming optimization procedures, frequency, high equip inspiring platform engineering projects need complex electromagnetic theory.