DART: Distribution-Aware Hardware Trojan Detection

Luke Chen; Youssef Gamal; Yanda  Li; Shih-Yuan  Yu; Ihsen  Alouani; Mohammad  Al Faruque

doi:10.1109/TIFS.2025.3607240

DART: Distribution-Aware Hardware Trojan Detection

Luke Chen
, Youssef Gamal
, Yanda Li
, Shih-Yuan Yu
, Ihsen Alouani
, Mohammad Al Faruque

School of Electronics, Electrical Engineering and Computer Science

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

86 Downloads (Pure)

Abstract

Machine Learning (ML) has proven effective in Integrated Circuits (IC) security, particularly in Hardware Trojan (HT) detection. However, a model’s generalization potential depends on its ability to address distribution shifts (DS) in unseen data. Mitigating DS enhances a model’s adaptability to novel variations and threats within the dynamic realm of IC designs and HTs. We formulate HT detection as a DS problem, introducing DART, a novel Distribution-Aware HT detection framework, to enhance model generalization. Applying DART on state-of-the-art Graph Neural Network architecture yields up to 22.96% and 17.37% F1-score improvements for unseen IC designs diverging significantly from the training data.

Original language	English
Pages (from-to)	9600 - 9609
Number of pages	10
Journal	IEEE Transactions on Information Forensics and Security
Volume	20
DOIs	https://doi.org/10.1109/TIFS.2025.3607240
Publication status	Published - 08 Sept 2025

Publications and Copyright Policy

This work is licensed under Queen’s Research Publications and Copyright Policy

Access to Document

10.1109/TIFS.2025.3607240Licence: Unspecified

DART: Distribution-Aware Hardware Trojan Detection
Copyright 2025 the authors. This is an accepted manuscript distributed under a Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the author and source are cited.
Accepted author manuscript, 2.99 MBLicence: CC BY

Cite this

@article{6c765fbd83754a74b4af395a3209bb30,

title = "DART: Distribution-Aware Hardware Trojan Detection",

abstract = "Machine Learning (ML) has proven effective in Integrated Circuits (IC) security, particularly in Hardware Trojan (HT) detection. However, a model{\textquoteright}s generalization potential depends on its ability to address distribution shifts (DS) in unseen data. Mitigating DS enhances a model{\textquoteright}s adaptability to novel variations and threats within the dynamic realm of IC designs and HTs. We formulate HT detection as a DS problem, introducing DART, a novel Distribution-Aware HT detection framework, to enhance model generalization. Applying DART on state-of-the-art Graph Neural Network architecture yields up to 22.96\% and 17.37\% F1-score improvements for unseen IC designs diverging significantly from the training data.",

author = "Luke Chen and Youssef Gamal and Yanda Li and Shih-Yuan Yu and Ihsen Alouani and \{Al Faruque\}, Mohammad",

year = "2025",

month = sep,

day = "8",

doi = "10.1109/TIFS.2025.3607240",

language = "English",

volume = "20",

pages = " 9600 -- 9609",

journal = "IEEE Transactions on Information Forensics and Security",

issn = "1556-6021",

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

}

TY - JOUR

T1 - DART: Distribution-Aware Hardware Trojan Detection

AU - Chen, Luke

AU - Gamal, Youssef

AU - Li, Yanda

AU - Yu, Shih-Yuan

AU - Alouani, Ihsen

AU - Al Faruque, Mohammad

PY - 2025/9/8

Y1 - 2025/9/8

N2 - Machine Learning (ML) has proven effective in Integrated Circuits (IC) security, particularly in Hardware Trojan (HT) detection. However, a model’s generalization potential depends on its ability to address distribution shifts (DS) in unseen data. Mitigating DS enhances a model’s adaptability to novel variations and threats within the dynamic realm of IC designs and HTs. We formulate HT detection as a DS problem, introducing DART, a novel Distribution-Aware HT detection framework, to enhance model generalization. Applying DART on state-of-the-art Graph Neural Network architecture yields up to 22.96% and 17.37% F1-score improvements for unseen IC designs diverging significantly from the training data.

AB - Machine Learning (ML) has proven effective in Integrated Circuits (IC) security, particularly in Hardware Trojan (HT) detection. However, a model’s generalization potential depends on its ability to address distribution shifts (DS) in unseen data. Mitigating DS enhances a model’s adaptability to novel variations and threats within the dynamic realm of IC designs and HTs. We formulate HT detection as a DS problem, introducing DART, a novel Distribution-Aware HT detection framework, to enhance model generalization. Applying DART on state-of-the-art Graph Neural Network architecture yields up to 22.96% and 17.37% F1-score improvements for unseen IC designs diverging significantly from the training data.

U2 - 10.1109/TIFS.2025.3607240

DO - 10.1109/TIFS.2025.3607240

M3 - Article

SN - 1556-6021

VL - 20

SP - 9600

EP - 9609

JO - IEEE Transactions on Information Forensics and Security

JF - IEEE Transactions on Information Forensics and Security

ER -

DART: Distribution-Aware Hardware Trojan Detection

Abstract

Publications and Copyright Policy

Access to Document

Fingerprint

Cite this