Knowing Your AMS System’s Limits: System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Authors

  • Georg Gläser
  • Hyun Sek Lukas Lee
  • Markus Olbrich
  • Erich Barke

Research Organisations

External Research Organisations

  • IMMS Institut für Mikroelektronik- und Mechatronik-Systeme gemeinnützige GmbH (IMMS GmbH)
View graph of relations

Details

Original languageEnglish
Title of host publicationLanguages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016
EditorsRobert Wille, Franco Fummi
PublisherSpringer Verlag
Pages1-14
Number of pages14
ISBN (print)9783319629193
Publication statusPublished - 2018
EventForum on specification and Design Languages, FDL 2016 - Bremen, Germany
Duration: 14 Sept 201616 Sept 2016

Publication series

NameLecture Notes in Electrical Engineering
Volume454
ISSN (Print)1876-1100
ISSN (electronic)1876-1119

Abstract

Virtual prototyping of Analog/Mixed-Signal (AMS) systems is a key concern in modern SoC verification. Achieving first-time right designs is a challenging task: Every relevant functional and non-functional property has to be examined throughout the complete design process. Many faulty designs have been verified carefully before tape out but are still missing at least one low-level effect which arises from interaction between one or more system components. Since these extra-functional effects are often neglected on system level, the design cannot be rectified in early design stages or verified before fabrication. We introduce a method to determine system acceptance regions tackling this challenge: We include extra-functional effects into the system models, and we investigate their behavior with parallel simulations in combination with an accelerated analog simulation scheme. The accelerated simulation approach is based on local linearizations of nonlinear circuits, which result in piecewise-linear systems. High-level simulation speed-up is achieved by avoiding numerical integration and using parallel computing. This approach is fully automated requiring only a circuit netlist. To reduce the overall number of simulations, we use an adaptive sampling algorithm for exploring systems acceptance regions which indicate feasible and critical operating conditions of the AMS system.

Keywords

    Accelerated simulation, Acceptance region, Automated model refinement, Bordersearch, Design automation, Extra-functional properties, Mixed-signal, Modeling, Parameter space, Piece-wise linear, Simulation, System level, Verification, Virtual prototyping

ASJC Scopus subject areas

Cite this

Knowing Your AMS System’s Limits: System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation. / Gläser, Georg; Lee, Hyun Sek Lukas; Olbrich, Markus et al.
Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016. ed. / Robert Wille; Franco Fummi. Springer Verlag, 2018. p. 1-14 (Lecture Notes in Electrical Engineering; Vol. 454).

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Gläser, G, Lee, HSL, Olbrich, M & Barke, E 2018, Knowing Your AMS System’s Limits: System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation. in R Wille & F Fummi (eds), Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016. Lecture Notes in Electrical Engineering, vol. 454, Springer Verlag, pp. 1-14, Forum on specification and Design Languages, FDL 2016, Bremen, Germany, 14 Sept 2016. https://doi.org/10.1007/978-3-319-62920-9_1
Gläser, G., Lee, H. S. L., Olbrich, M., & Barke, E. (2018). Knowing Your AMS System’s Limits: System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation. In R. Wille, & F. Fummi (Eds.), Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016 (pp. 1-14). (Lecture Notes in Electrical Engineering; Vol. 454). Springer Verlag. https://doi.org/10.1007/978-3-319-62920-9_1
Gläser G, Lee HSL, Olbrich M, Barke E. Knowing Your AMS System’s Limits: System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation. In Wille R, Fummi F, editors, Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016. Springer Verlag. 2018. p. 1-14. (Lecture Notes in Electrical Engineering). doi: 10.1007/978-3-319-62920-9_1
Gläser, Georg ; Lee, Hyun Sek Lukas ; Olbrich, Markus et al. / Knowing Your AMS System’s Limits : System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation. Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016. editor / Robert Wille ; Franco Fummi. Springer Verlag, 2018. pp. 1-14 (Lecture Notes in Electrical Engineering).
Download
@inproceedings{39ce24719b97483bb8bb81d07d8458f8,
title = "Knowing Your AMS System{\textquoteright}s Limits: System Acceptance Region Exploration by Using Automated Model Refinement and Accelerated Simulation",
abstract = "Virtual prototyping of Analog/Mixed-Signal (AMS) systems is a key concern in modern SoC verification. Achieving first-time right designs is a challenging task: Every relevant functional and non-functional property has to be examined throughout the complete design process. Many faulty designs have been verified carefully before tape out but are still missing at least one low-level effect which arises from interaction between one or more system components. Since these extra-functional effects are often neglected on system level, the design cannot be rectified in early design stages or verified before fabrication. We introduce a method to determine system acceptance regions tackling this challenge: We include extra-functional effects into the system models, and we investigate their behavior with parallel simulations in combination with an accelerated analog simulation scheme. The accelerated simulation approach is based on local linearizations of nonlinear circuits, which result in piecewise-linear systems. High-level simulation speed-up is achieved by avoiding numerical integration and using parallel computing. This approach is fully automated requiring only a circuit netlist. To reduce the overall number of simulations, we use an adaptive sampling algorithm for exploring systems acceptance regions which indicate feasible and critical operating conditions of the AMS system.",
keywords = "Accelerated simulation, Acceptance region, Automated model refinement, Bordersearch, Design automation, Extra-functional properties, Mixed-signal, Modeling, Parameter space, Piece-wise linear, Simulation, System level, Verification, Virtual prototyping",
author = "Georg Gl{\"a}ser and Lee, {Hyun Sek Lukas} and Markus Olbrich and Erich Barke",
note = "Publisher Copyright: {\textcopyright} 2018, Springer International Publishing AG. Copyright: Copyright 2017 Elsevier B.V., All rights reserved.; Forum on specification and Design Languages, FDL 2016 ; Conference date: 14-09-2016 Through 16-09-2016",
year = "2018",
doi = "10.1007/978-3-319-62920-9_1",
language = "English",
isbn = "9783319629193",
series = "Lecture Notes in Electrical Engineering",
publisher = "Springer Verlag",
pages = "1--14",
editor = "Robert Wille and Franco Fummi",
booktitle = "Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016",
address = "Germany",

}

Download

TY - GEN

T1 - Knowing Your AMS System’s Limits

T2 - Forum on specification and Design Languages, FDL 2016

AU - Gläser, Georg

AU - Lee, Hyun Sek Lukas

AU - Olbrich, Markus

AU - Barke, Erich

N1 - Publisher Copyright: © 2018, Springer International Publishing AG. Copyright: Copyright 2017 Elsevier B.V., All rights reserved.

PY - 2018

Y1 - 2018

N2 - Virtual prototyping of Analog/Mixed-Signal (AMS) systems is a key concern in modern SoC verification. Achieving first-time right designs is a challenging task: Every relevant functional and non-functional property has to be examined throughout the complete design process. Many faulty designs have been verified carefully before tape out but are still missing at least one low-level effect which arises from interaction between one or more system components. Since these extra-functional effects are often neglected on system level, the design cannot be rectified in early design stages or verified before fabrication. We introduce a method to determine system acceptance regions tackling this challenge: We include extra-functional effects into the system models, and we investigate their behavior with parallel simulations in combination with an accelerated analog simulation scheme. The accelerated simulation approach is based on local linearizations of nonlinear circuits, which result in piecewise-linear systems. High-level simulation speed-up is achieved by avoiding numerical integration and using parallel computing. This approach is fully automated requiring only a circuit netlist. To reduce the overall number of simulations, we use an adaptive sampling algorithm for exploring systems acceptance regions which indicate feasible and critical operating conditions of the AMS system.

AB - Virtual prototyping of Analog/Mixed-Signal (AMS) systems is a key concern in modern SoC verification. Achieving first-time right designs is a challenging task: Every relevant functional and non-functional property has to be examined throughout the complete design process. Many faulty designs have been verified carefully before tape out but are still missing at least one low-level effect which arises from interaction between one or more system components. Since these extra-functional effects are often neglected on system level, the design cannot be rectified in early design stages or verified before fabrication. We introduce a method to determine system acceptance regions tackling this challenge: We include extra-functional effects into the system models, and we investigate their behavior with parallel simulations in combination with an accelerated analog simulation scheme. The accelerated simulation approach is based on local linearizations of nonlinear circuits, which result in piecewise-linear systems. High-level simulation speed-up is achieved by avoiding numerical integration and using parallel computing. This approach is fully automated requiring only a circuit netlist. To reduce the overall number of simulations, we use an adaptive sampling algorithm for exploring systems acceptance regions which indicate feasible and critical operating conditions of the AMS system.

KW - Accelerated simulation

KW - Acceptance region

KW - Automated model refinement

KW - Bordersearch

KW - Design automation

KW - Extra-functional properties

KW - Mixed-signal

KW - Modeling

KW - Parameter space

KW - Piece-wise linear

KW - Simulation

KW - System level

KW - Verification

KW - Virtual prototyping

UR - http://www.scopus.com/inward/record.url?scp=85034443751&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-62920-9_1

DO - 10.1007/978-3-319-62920-9_1

M3 - Conference contribution

AN - SCOPUS:85034443751

SN - 9783319629193

T3 - Lecture Notes in Electrical Engineering

SP - 1

EP - 14

BT - Languages, Design Methods, and Tools for Electronic System Design - Selected Contributions from FDL 2016

A2 - Wille, Robert

A2 - Fummi, Franco

PB - Springer Verlag

Y2 - 14 September 2016 through 16 September 2016

ER -