A coverage guided mining approach for automatic generation of succinct assertions

David Sheridan, Lingyi Liu, Hyungsul Kim, Shobha Vasudevan

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Machine learning techniques are widely employed for automatic assertion generation in hardware verification. Our previous method Goldmine uses a decision tree based approach for mining assertions and does not have design coverage related feedback. The assertions are unaware of the design, over-constrained and have low expressiveness. We introduce a coverage guided mining approach for mining assertions from simulation traces. Our approach combines association rule learning, greedy set covering and formal verification. It circumvents the exhaustive rule generation of association mining using coverage feedback. The algorithm has been implemented as one part of GoldMine The tool can be downloaded from the website: http://goldmine.csl.illinois.edu. Experiments using a variety of designs, including USB, PCI and OpenRisc, show that the assertions generated by coverage guided association mining cover an average of 6.14 times input space than those generated by decision tree based mining algorithms. We also show that the coverage guided association mining produces an average of 2.75 times fewer propositions per assertion than decision tree based mining. All these mean the assertions generated by coverage guided association mining are more succinct, and of higher value for hardware design verification.

Original languageEnglish (US)
Title of host publicationProceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design
Pages68-73
Number of pages6
DOIs
StatePublished - Mar 3 2014
Event27th International Conference on VLSI Design, VLSID 2014 - Held Concurrently with 13th International Conference on Embedded Systems Design - Mumbai, India
Duration: Jan 5 2014Jan 9 2014

Publication series

NameProceedings of the IEEE International Conference on VLSI Design
ISSN (Print)1063-9667

Other

Other27th International Conference on VLSI Design, VLSID 2014 - Held Concurrently with 13th International Conference on Embedded Systems Design
CountryIndia
CityMumbai
Period1/5/141/9/14

Fingerprint

Decision trees
Feedback
Hardware
Association rules
Learning systems
Experiments

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Cite this

Sheridan, D., Liu, L., Kim, H., & Vasudevan, S. (2014). A coverage guided mining approach for automatic generation of succinct assertions. In Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design (pp. 68-73). [6733108] (Proceedings of the IEEE International Conference on VLSI Design). https://doi.org/10.1109/VLSID.2014.19

A coverage guided mining approach for automatic generation of succinct assertions. / Sheridan, David; Liu, Lingyi; Kim, Hyungsul; Vasudevan, Shobha.

Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design. 2014. p. 68-73 6733108 (Proceedings of the IEEE International Conference on VLSI Design).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Sheridan, D, Liu, L, Kim, H & Vasudevan, S 2014, A coverage guided mining approach for automatic generation of succinct assertions. in Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design., 6733108, Proceedings of the IEEE International Conference on VLSI Design, pp. 68-73, 27th International Conference on VLSI Design, VLSID 2014 - Held Concurrently with 13th International Conference on Embedded Systems Design, Mumbai, India, 1/5/14. https://doi.org/10.1109/VLSID.2014.19
Sheridan D, Liu L, Kim H, Vasudevan S. A coverage guided mining approach for automatic generation of succinct assertions. In Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design. 2014. p. 68-73. 6733108. (Proceedings of the IEEE International Conference on VLSI Design). https://doi.org/10.1109/VLSID.2014.19
Sheridan, David ; Liu, Lingyi ; Kim, Hyungsul ; Vasudevan, Shobha. / A coverage guided mining approach for automatic generation of succinct assertions. Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design. 2014. pp. 68-73 (Proceedings of the IEEE International Conference on VLSI Design).
@inproceedings{9ce66150555640d2a60e95dcc81285c3,
title = "A coverage guided mining approach for automatic generation of succinct assertions",
abstract = "Machine learning techniques are widely employed for automatic assertion generation in hardware verification. Our previous method Goldmine uses a decision tree based approach for mining assertions and does not have design coverage related feedback. The assertions are unaware of the design, over-constrained and have low expressiveness. We introduce a coverage guided mining approach for mining assertions from simulation traces. Our approach combines association rule learning, greedy set covering and formal verification. It circumvents the exhaustive rule generation of association mining using coverage feedback. The algorithm has been implemented as one part of GoldMine The tool can be downloaded from the website: http://goldmine.csl.illinois.edu. Experiments using a variety of designs, including USB, PCI and OpenRisc, show that the assertions generated by coverage guided association mining cover an average of 6.14 times input space than those generated by decision tree based mining algorithms. We also show that the coverage guided association mining produces an average of 2.75 times fewer propositions per assertion than decision tree based mining. All these mean the assertions generated by coverage guided association mining are more succinct, and of higher value for hardware design verification.",
author = "David Sheridan and Lingyi Liu and Hyungsul Kim and Shobha Vasudevan",
year = "2014",
month = "3",
day = "3",
doi = "10.1109/VLSID.2014.19",
language = "English (US)",
isbn = "9781479925124",
series = "Proceedings of the IEEE International Conference on VLSI Design",
pages = "68--73",
booktitle = "Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design",

}

TY - GEN

T1 - A coverage guided mining approach for automatic generation of succinct assertions

AU - Sheridan, David

AU - Liu, Lingyi

AU - Kim, Hyungsul

AU - Vasudevan, Shobha

PY - 2014/3/3

Y1 - 2014/3/3

N2 - Machine learning techniques are widely employed for automatic assertion generation in hardware verification. Our previous method Goldmine uses a decision tree based approach for mining assertions and does not have design coverage related feedback. The assertions are unaware of the design, over-constrained and have low expressiveness. We introduce a coverage guided mining approach for mining assertions from simulation traces. Our approach combines association rule learning, greedy set covering and formal verification. It circumvents the exhaustive rule generation of association mining using coverage feedback. The algorithm has been implemented as one part of GoldMine The tool can be downloaded from the website: http://goldmine.csl.illinois.edu. Experiments using a variety of designs, including USB, PCI and OpenRisc, show that the assertions generated by coverage guided association mining cover an average of 6.14 times input space than those generated by decision tree based mining algorithms. We also show that the coverage guided association mining produces an average of 2.75 times fewer propositions per assertion than decision tree based mining. All these mean the assertions generated by coverage guided association mining are more succinct, and of higher value for hardware design verification.

AB - Machine learning techniques are widely employed for automatic assertion generation in hardware verification. Our previous method Goldmine uses a decision tree based approach for mining assertions and does not have design coverage related feedback. The assertions are unaware of the design, over-constrained and have low expressiveness. We introduce a coverage guided mining approach for mining assertions from simulation traces. Our approach combines association rule learning, greedy set covering and formal verification. It circumvents the exhaustive rule generation of association mining using coverage feedback. The algorithm has been implemented as one part of GoldMine The tool can be downloaded from the website: http://goldmine.csl.illinois.edu. Experiments using a variety of designs, including USB, PCI and OpenRisc, show that the assertions generated by coverage guided association mining cover an average of 6.14 times input space than those generated by decision tree based mining algorithms. We also show that the coverage guided association mining produces an average of 2.75 times fewer propositions per assertion than decision tree based mining. All these mean the assertions generated by coverage guided association mining are more succinct, and of higher value for hardware design verification.

UR - http://www.scopus.com/inward/record.url?scp=84894597433&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84894597433&partnerID=8YFLogxK

U2 - 10.1109/VLSID.2014.19

DO - 10.1109/VLSID.2014.19

M3 - Conference contribution

AN - SCOPUS:84894597433

SN - 9781479925124

T3 - Proceedings of the IEEE International Conference on VLSI Design

SP - 68

EP - 73

BT - Proceedings - 27th International Conference on VLSI Design, VLSID 2014; Held Concurrently with 13th International Conference on Embedded Systems Design

ER -