Analysis of Dielectric Post-Wall Waveguide-based Passive Circuits using Recurrent Neural Network

The dielectric post-wall waveguide-based passive circuits are analyzed using a surrogate model represented by an attention-based recurrent neural network. The predicted results from this much simpler, trained model show a good agreement with those obtained by a corresponding full-wave computational electromagnetics analysis with a fully independent EM commercial software package.