Deep Neural Networks for Rapid Simulation of Planar Microwave Circuits Based on Their Layouts

This article demonstrates a deep learning (DL)-based methodology for the rapid simulation of planar microwave circuits based on their layouts. We train convolutional neural networks (CNNs) to compute scattering parameters general, two-port consisting metallization layer printed grounded dielectric substrate, by processing pattern along with thickness and permittivity substrate. approach harnesses efficiency CNNs recognition tasks extends previous efforts employ parameterized circuit geometries. Furthermore, we integrate this CNN in hybrid network long-short term memory (LSTM) module that uses coarse mesh finite-difference time-domain (FDTD) data as an additional input. show is computationally efficient generalizable, accurately modeling geometries well beyond those has previously seen.