We propose using machine learning models for the direct synthesis of on-chip electromagnetic (EM) passive structures to enable rapid or even automated designs and optimizations of RF/mm-Wave circuits. As a proof of concept, we demonstrate the direct synthesis of a 1:1 transformer on a 45nm SOI process using our proposed neural network model. Using pre-existing transformer s-parameter files and their geometric design training samples, the model predicts target geometric designs.
@article{arxiv.2008.10755,
title = {Residual Network Based Direct Synthesis of EM Structures: A Study on One-to-One Transformers},
author = {David Munzer and Siawpeng Er and Minshuo Chen and Yan Li and Naga S. Mannem and Tuo Zhao and Hua Wang},
journal= {arXiv preprint arXiv:2008.10755},
year = {2020}
}
Comments
IEEE Radio Frequency Integrated Circuits Symposium (RFIC) 2020