Learning to Design Analog Circuits to Meet Threshold Specifications

Dmitrii Krylov, Pooya Khajeh, Junhan Ouyang, Thomas Reeves, Tongkai Liu, Hiba Ajmal, Hamidreza Aghasi, Roy Fox
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:17858-17873, 2023.

Abstract

Automated design of analog and radio-frequency circuits using supervised or reinforcement learning from simulation data has recently been studied as an alternative to manual expert design. It is straightforward for a design agent to learn an inverse function from desired performance metrics to circuit parameters. However, it is more common for a user to have threshold performance criteria rather than an exact target vector of feasible performance measures. In this work, we propose a method for generating from simulation data a dataset on which a system can be trained via supervised learning to design circuits to meet threshold specifications. We moreover perform the to-date most extensive evaluation of automated analog circuit design, including experimenting in a significantly more diverse set of circuits than in prior work, covering linear, nonlinear, and autonomous circuit configurations, and show that our method consistently reaches success rate better than 90% at 5% error margin, while also improving data efficiency by upward of an order of magnitude.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-krylov23a, title = {Learning to Design Analog Circuits to Meet Threshold Specifications}, author = {Krylov, Dmitrii and Khajeh, Pooya and Ouyang, Junhan and Reeves, Thomas and Liu, Tongkai and Ajmal, Hiba and Aghasi, Hamidreza and Fox, Roy}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {17858--17873}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/krylov23a/krylov23a.pdf}, url = {https://proceedings.mlr.press/v202/krylov23a.html}, abstract = {Automated design of analog and radio-frequency circuits using supervised or reinforcement learning from simulation data has recently been studied as an alternative to manual expert design. It is straightforward for a design agent to learn an inverse function from desired performance metrics to circuit parameters. However, it is more common for a user to have threshold performance criteria rather than an exact target vector of feasible performance measures. In this work, we propose a method for generating from simulation data a dataset on which a system can be trained via supervised learning to design circuits to meet threshold specifications. We moreover perform the to-date most extensive evaluation of automated analog circuit design, including experimenting in a significantly more diverse set of circuits than in prior work, covering linear, nonlinear, and autonomous circuit configurations, and show that our method consistently reaches success rate better than 90% at 5% error margin, while also improving data efficiency by upward of an order of magnitude.} }
Endnote
%0 Conference Paper %T Learning to Design Analog Circuits to Meet Threshold Specifications %A Dmitrii Krylov %A Pooya Khajeh %A Junhan Ouyang %A Thomas Reeves %A Tongkai Liu %A Hiba Ajmal %A Hamidreza Aghasi %A Roy Fox %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-krylov23a %I PMLR %P 17858--17873 %U https://proceedings.mlr.press/v202/krylov23a.html %V 202 %X Automated design of analog and radio-frequency circuits using supervised or reinforcement learning from simulation data has recently been studied as an alternative to manual expert design. It is straightforward for a design agent to learn an inverse function from desired performance metrics to circuit parameters. However, it is more common for a user to have threshold performance criteria rather than an exact target vector of feasible performance measures. In this work, we propose a method for generating from simulation data a dataset on which a system can be trained via supervised learning to design circuits to meet threshold specifications. We moreover perform the to-date most extensive evaluation of automated analog circuit design, including experimenting in a significantly more diverse set of circuits than in prior work, covering linear, nonlinear, and autonomous circuit configurations, and show that our method consistently reaches success rate better than 90% at 5% error margin, while also improving data efficiency by upward of an order of magnitude.
APA
Krylov, D., Khajeh, P., Ouyang, J., Reeves, T., Liu, T., Ajmal, H., Aghasi, H. & Fox, R.. (2023). Learning to Design Analog Circuits to Meet Threshold Specifications. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:17858-17873 Available from https://proceedings.mlr.press/v202/krylov23a.html.

Related Material