Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters

Hakki M. Torun; Ahmet C. Durgun; Kemal Aygun; Madhavan Swaminathan

doi:10.1109/EPEPS47316.2019.193234

Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters

Hakki M. Torun
, Ahmet C. Durgun
, Kemal Aygun
, Madhavan Swaminathan

Electrical Engineering

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

11 Scopus citations

Abstract

This paper proposes a method to ensure that S-Parameters generated using neural network (NN) models are physically consistent and can be safely used in subsequent time-domain simulations. This is achieved by introducing causality and passivity enforcement layers as the last two layers of the NN, while minimizing their computational overhead to the training and inference of the NN model. Proposed technique is demonstrated on learning the mapping from 13 dimensional geometrical parameters of a differential plated through hole (PTH) in package core to its corresponding broadband S-Parameters up to 100 GHz.

Original language	English (US)
Title of host publication	2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728145853
DOIs	https://doi.org/10.1109/EPEPS47316.2019.193234
State	Published - Oct 2019
Event	28th IEEE Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019 - Montreal, Canada Duration: Oct 6 2019 → Oct 9 2019

Publication series

Name	2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019

Conference

Conference	28th IEEE Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019
Country/Territory	Canada
City	Montreal
Period	10/6/19 → 10/9/19

All Science Journal Classification (ASJC) codes

Hardware and Architecture
Energy Engineering and Power Technology
Electrical and Electronic Engineering
Safety, Risk, Reliability and Quality
Electronic, Optical and Magnetic Materials

Access to Document

10.1109/EPEPS47316.2019.193234

Cite this

Torun, H. M., Durgun, A. C., Aygun, K., & Swaminathan, M. (2019). Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters. In 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019 Article 9073310 (2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EPEPS47316.2019.193234

Torun, Hakki M. ; Durgun, Ahmet C. ; Aygun, Kemal et al. / Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters. 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019. Institute of Electrical and Electronics Engineers Inc., 2019. (2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019).

@inproceedings{e181ec5865be444396c1e0a54ec198fa,

title = "Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters",

abstract = "This paper proposes a method to ensure that S-Parameters generated using neural network (NN) models are physically consistent and can be safely used in subsequent time-domain simulations. This is achieved by introducing causality and passivity enforcement layers as the last two layers of the NN, while minimizing their computational overhead to the training and inference of the NN model. Proposed technique is demonstrated on learning the mapping from 13 dimensional geometrical parameters of a differential plated through hole (PTH) in package core to its corresponding broadband S-Parameters up to 100 GHz.",

author = "Torun, \{Hakki M.\} and Durgun, \{Ahmet C.\} and Kemal Aygun and Madhavan Swaminathan",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 28th IEEE Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019 ; Conference date: 06-10-2019 Through 09-10-2019",

year = "2019",

month = oct,

doi = "10.1109/EPEPS47316.2019.193234",

language = "English (US)",

series = "2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019",

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

booktitle = "2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019",

address = "United States",

}

Torun, HM, Durgun, AC, Aygun, K & Swaminathan, M 2019, Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters. in 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019., 9073310, 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019, Institute of Electrical and Electronics Engineers Inc., 28th IEEE Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019, Montreal, Canada, 10/6/19. https://doi.org/10.1109/EPEPS47316.2019.193234

Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters. / Torun, Hakki M.; Durgun, Ahmet C.; Aygun, Kemal et al.
2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019. Institute of Electrical and Electronics Engineers Inc., 2019. 9073310 (2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019).

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

TY - GEN

T1 - Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters

AU - Torun, Hakki M.

AU - Durgun, Ahmet C.

AU - Aygun, Kemal

AU - Swaminathan, Madhavan

PY - 2019/10

Y1 - 2019/10

N2 - This paper proposes a method to ensure that S-Parameters generated using neural network (NN) models are physically consistent and can be safely used in subsequent time-domain simulations. This is achieved by introducing causality and passivity enforcement layers as the last two layers of the NN, while minimizing their computational overhead to the training and inference of the NN model. Proposed technique is demonstrated on learning the mapping from 13 dimensional geometrical parameters of a differential plated through hole (PTH) in package core to its corresponding broadband S-Parameters up to 100 GHz.

AB - This paper proposes a method to ensure that S-Parameters generated using neural network (NN) models are physically consistent and can be safely used in subsequent time-domain simulations. This is achieved by introducing causality and passivity enforcement layers as the last two layers of the NN, while minimizing their computational overhead to the training and inference of the NN model. Proposed technique is demonstrated on learning the mapping from 13 dimensional geometrical parameters of a differential plated through hole (PTH) in package core to its corresponding broadband S-Parameters up to 100 GHz.

UR - https://www.scopus.com/pages/publications/85084533313

UR - https://www.scopus.com/pages/publications/85084533313#tab=citedBy

U2 - 10.1109/EPEPS47316.2019.193234

DO - 10.1109/EPEPS47316.2019.193234

M3 - Conference contribution

AN - SCOPUS:85084533313

T3 - 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019

BT - 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 28th IEEE Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019

Y2 - 6 October 2019 through 9 October 2019

ER -

Torun HM, Durgun AC, Aygun K, Swaminathan M. Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters. In 2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019. Institute of Electrical and Electronics Engineers Inc. 2019. 9073310. (2019 IEEE 28th Conference on Electrical Performance of Electronic Packaging and Systems, EPEPS 2019). doi: 10.1109/EPEPS47316.2019.193234

Enforcing Causality and Passivity of Neural Network Models of Broadband S-Parameters

Abstract

Publication series

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this