Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications

Osama Waqar Bhatti; Oluwaseyi Akinwande; Madhavan Swaminathan

doi:10.1109/NEMO51452.2022.10038959

Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications

Osama Waqar Bhatti, Oluwaseyi Akinwande, Madhavan Swaminathan

Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We present a machine learning based tool to quantify uncertainty for prediction problems regarding signal integrity. Harnessing invertible neural networks, we convert the inverse posterior distribution given by the network to address uncertainty in frequency responses as a function of design space parameters. As an example, we consider a differential plated-through-hole via in package core and predict S-parameters from its geometrical properties. Results show 3.3 % normalized mean squared error when compared with responses from a fullwave EM simulator.

Original language	English (US)
Title of host publication	2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781665486330
DOIs	https://doi.org/10.1109/NEMO51452.2022.10038959
State	Published - 2022
Event	2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022 - Limoges, France Duration: Jul 6 2022 → Jul 8 2022

Publication series

Name	2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022

Conference

Conference	2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022
Country/Territory	France
City	Limoges
Period	7/6/22 → 7/8/22

All Science Journal Classification (ASJC) codes

Fluid Flow and Transfer Processes
Computational Mechanics
Electrical and Electronic Engineering
Electronic, Optical and Magnetic Materials
Numerical Analysis
Instrumentation
Modeling and Simulation

Access to Document

10.1109/NEMO51452.2022.10038959

Cite this

Bhatti, O. W., Akinwande, O., & Swaminathan, M. (2022). Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications. In 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022 (2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/NEMO51452.2022.10038959

Bhatti, Osama Waqar ; Akinwande, Oluwaseyi ; Swaminathan, Madhavan. / Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications. 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022. Institute of Electrical and Electronics Engineers Inc., 2022. (2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022).

@inproceedings{e27ae99e6f08408cba0092fb2af7ef72,

title = "Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications",

abstract = "We present a machine learning based tool to quantify uncertainty for prediction problems regarding signal integrity. Harnessing invertible neural networks, we convert the inverse posterior distribution given by the network to address uncertainty in frequency responses as a function of design space parameters. As an example, we consider a differential plated-through-hole via in package core and predict S-parameters from its geometrical properties. Results show 3.3 % normalized mean squared error when compared with responses from a fullwave EM simulator.",

author = "Bhatti, {Osama Waqar} and Oluwaseyi Akinwande and Madhavan Swaminathan",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022 ; Conference date: 06-07-2022 Through 08-07-2022",

year = "2022",

doi = "10.1109/NEMO51452.2022.10038959",

language = "English (US)",

series = "2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022",

address = "United States",

}

Bhatti, OW, Akinwande, O & Swaminathan, M 2022, Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications. in 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022. 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022, Institute of Electrical and Electronics Engineers Inc., 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022, Limoges, France, 7/6/22. https://doi.org/10.1109/NEMO51452.2022.10038959

Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications. / Bhatti, Osama Waqar; Akinwande, Oluwaseyi; Swaminathan, Madhavan.
2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022. Institute of Electrical and Electronics Engineers Inc., 2022. (2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications

AU - Bhatti, Osama Waqar

AU - Akinwande, Oluwaseyi

AU - Swaminathan, Madhavan

PY - 2022

Y1 - 2022

N2 - We present a machine learning based tool to quantify uncertainty for prediction problems regarding signal integrity. Harnessing invertible neural networks, we convert the inverse posterior distribution given by the network to address uncertainty in frequency responses as a function of design space parameters. As an example, we consider a differential plated-through-hole via in package core and predict S-parameters from its geometrical properties. Results show 3.3 % normalized mean squared error when compared with responses from a fullwave EM simulator.

AB - We present a machine learning based tool to quantify uncertainty for prediction problems regarding signal integrity. Harnessing invertible neural networks, we convert the inverse posterior distribution given by the network to address uncertainty in frequency responses as a function of design space parameters. As an example, we consider a differential plated-through-hole via in package core and predict S-parameters from its geometrical properties. Results show 3.3 % normalized mean squared error when compared with responses from a fullwave EM simulator.

UR - http://www.scopus.com/inward/record.url?scp=85149328679&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85149328679&partnerID=8YFLogxK

U2 - 10.1109/NEMO51452.2022.10038959

DO - 10.1109/NEMO51452.2022.10038959

M3 - Conference contribution

AN - SCOPUS:85149328679

T3 - 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022

BT - 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022

Y2 - 6 July 2022 through 8 July 2022

ER -

Bhatti OW, Akinwande O, Swaminathan M. Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications. In 2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022. Institute of Electrical and Electronics Engineers Inc. 2022. (2022 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO 2022). doi: 10.1109/NEMO51452.2022.10038959

Uncertainty Quantification with Invertible Neural Networks for Signal Integrity Applications

Abstract

Publication series

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Cite this