OMLA: An Oracle-Less Machine Learning-Based Attack on Logic Locking

Lilas Alrahis; Satwik Patnaik; Muhammad Shafique; Ozgur Sinanoglu

doi:10.1109/TCSII.2021.3113035

OMLA: An Oracle-Less Machine Learning-Based Attack on Logic Locking

Lilas Alrahis, Satwik Patnaik, Muhammad Shafique, Ozgur Sinanoglu

Research output: Contribution to journal › Article › peer-review

Abstract

Hardware-based attacks on the semiconductor supply chain are emerging due to the globalization of the design flow. Logic locking is a design-for-trust scheme that promises protection throughout the supply chain. While attacks have heavily relied on an oracle to break logic locking, machine learning (ML)-based attacks demonstrate the daunting possibility of breaking locking even without an oracle. Although very potent, current ML-based attacks recover only a subset of the transformations introduced by locking. We aim to address this shortcoming by developing an oracle-less graph neural network-based attack called OMLA, questioning once again the security of logic locking. Our experiments on ISCAS-85 and ITC-99 benchmarks demonstrate that OMLA achieves a key-prediction accuracy up to 97.22% and outperforms state-of-the-art SnapShot and SAIL attacks for all evaluated benchmarks.

Original language	English (US)
Pages (from-to)	1602-1606
Number of pages	5
Journal	IEEE Transactions on Circuits and Systems II: Express Briefs
Volume	69
Issue number	3
DOIs	https://doi.org/10.1109/TCSII.2021.3113035
State	Published - Mar 1 2022

Keywords

Logic locking
graph neural networks
machine learning
oracle-less attack

ASJC Scopus subject areas

Electrical and Electronic Engineering

Access to Document

10.1109/TCSII.2021.3113035

Cite this

@article{67407a14d993434bb7b5ab3628e6b9c4,

title = "OMLA: An Oracle-Less Machine Learning-Based Attack on Logic Locking",

abstract = "Hardware-based attacks on the semiconductor supply chain are emerging due to the globalization of the design flow. Logic locking is a design-for-trust scheme that promises protection throughout the supply chain. While attacks have heavily relied on an oracle to break logic locking, machine learning (ML)-based attacks demonstrate the daunting possibility of breaking locking even without an oracle. Although very potent, current ML-based attacks recover only a subset of the transformations introduced by locking. We aim to address this shortcoming by developing an oracle-less graph neural network-based attack called OMLA, questioning once again the security of logic locking. Our experiments on ISCAS-85 and ITC-99 benchmarks demonstrate that OMLA achieves a key-prediction accuracy up to 97.22% and outperforms state-of-the-art SnapShot and SAIL attacks for all evaluated benchmarks.",

keywords = "Logic locking, graph neural networks, machine learning, oracle-less attack",

author = "Lilas Alrahis and Satwik Patnaik and Muhammad Shafique and Ozgur Sinanoglu",

note = "Publisher Copyright: {\textcopyright} 2004-2012 IEEE.",

year = "2022",

month = mar,

day = "1",

doi = "10.1109/TCSII.2021.3113035",

language = "English (US)",

volume = "69",

pages = "1602--1606",

journal = "IEEE Transactions on Circuits and Systems II: Express Briefs",

issn = "1549-7747",

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

number = "3",

}

TY - JOUR

T1 - OMLA

T2 - An Oracle-Less Machine Learning-Based Attack on Logic Locking

AU - Alrahis, Lilas

AU - Patnaik, Satwik

AU - Shafique, Muhammad

AU - Sinanoglu, Ozgur

PY - 2022/3/1

Y1 - 2022/3/1

N2 - Hardware-based attacks on the semiconductor supply chain are emerging due to the globalization of the design flow. Logic locking is a design-for-trust scheme that promises protection throughout the supply chain. While attacks have heavily relied on an oracle to break logic locking, machine learning (ML)-based attacks demonstrate the daunting possibility of breaking locking even without an oracle. Although very potent, current ML-based attacks recover only a subset of the transformations introduced by locking. We aim to address this shortcoming by developing an oracle-less graph neural network-based attack called OMLA, questioning once again the security of logic locking. Our experiments on ISCAS-85 and ITC-99 benchmarks demonstrate that OMLA achieves a key-prediction accuracy up to 97.22% and outperforms state-of-the-art SnapShot and SAIL attacks for all evaluated benchmarks.

AB - Hardware-based attacks on the semiconductor supply chain are emerging due to the globalization of the design flow. Logic locking is a design-for-trust scheme that promises protection throughout the supply chain. While attacks have heavily relied on an oracle to break logic locking, machine learning (ML)-based attacks demonstrate the daunting possibility of breaking locking even without an oracle. Although very potent, current ML-based attacks recover only a subset of the transformations introduced by locking. We aim to address this shortcoming by developing an oracle-less graph neural network-based attack called OMLA, questioning once again the security of logic locking. Our experiments on ISCAS-85 and ITC-99 benchmarks demonstrate that OMLA achieves a key-prediction accuracy up to 97.22% and outperforms state-of-the-art SnapShot and SAIL attacks for all evaluated benchmarks.

KW - Logic locking

KW - graph neural networks

KW - machine learning

KW - oracle-less attack

UR - http://www.scopus.com/inward/record.url?scp=85115120467&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85115120467&partnerID=8YFLogxK

U2 - 10.1109/TCSII.2021.3113035

DO - 10.1109/TCSII.2021.3113035

M3 - Article

AN - SCOPUS:85115120467

SN - 1549-7747

VL - 69

SP - 1602

EP - 1606

JO - IEEE Transactions on Circuits and Systems II: Express Briefs

JF - IEEE Transactions on Circuits and Systems II: Express Briefs

IS - 3

ER -

OMLA: An Oracle-Less Machine Learning-Based Attack on Logic Locking

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this