Design of SIW Filter Based on the Equivalent De-embedding Technique and Inverse Neural Network

In this paper, we propose an equivalent de-embedding technique which is simpler than the thru-reflect-line (TRL) calibration technique to accurately extract the generalized scattering matrix (GSM) of the SIW discontinuities. Instead of modeling the corresponding equivalent rectangular waveguide, we can synthesize the SIW filter directly using the equivalent de-embedding technique. On the basis of K-impedance inverter model, we build an inverse neural network that can synthesize the dimensions of a SIW filter. The effectiveness of the proposed method verified through the design of an H-plane 5th order SIW iris filter.