ArticlePublisher preview available

OPUF: Obfuscated PUF against Machine Learning Modeling Attacks

World Scientific
Journal of Circuits, Systems and Computers
Authors:
Mostafa Moghaddas at Iran University of Science and Technology
Mostafa Moghaddas
Marziye Pandi at Iran University of Science and Technology
Marziye Pandi
Hakem Beitollahi at Iran University of Science and Technology
Hakem Beitollahi
Read publisher preview
Download citation
To read the full-text of this research, you can request a copy directly from the authors.

Abstract and Figures

Physical Unclonable Functions (PUFs) are vulnerable to machine learning modeling attacks that can predict their responses. To counter this threat, we introduce Obfuscated PUF (OPUF), a novel PUF design that employs advanced obfuscation techniques to enhance security. OPUF significantly outperforms existing PUFs in terms of resistance to machine learning attacks, including logistic regression (LR) and multilayer perceptron (MLP). It achieves a 50.69% and 49.54% accuracy reduction in LR and MLP attacks, respectively, compared to the best-performing existing PUFs. Moreover, OPUF maintains excellent quality metrics, with 50% uniqueness and 50.15% uniformity, demonstrating its inherent randomness and unpredictability. OPUF achieves its enhanced security through a multi-layered architecture that incorporates obfuscation techniques and nonlinear properties. By introducing internal challenges that are difficult for attackers to access, OPUF effectively thwarts modeling attempts. Our experimental results demonstrate OPUF’s superior performance in resisting various machine learning attacks, even when faced with extensive training data. OPUF offers a promising solution for secure hardware implementations, addressing the critical need for PUFs that can withstand the challenges posed by modern machine learning attacks.
Figures - available from: Journal of Circuits, Systems and Computers
This content is subject to copyright.
A preview of this full-text is provided by World Scientific.
Content available from Journal of Circuits, Systems and Computers 
This content is subject to copyright.
OPUF: Obfuscated PUF against Machine Learning
Modeling Attacks
Mostafa Moghaddas
*
and Marziye Pandi
†
School of Computer Engineering,
Iran University of Science and Technology,
Tehran, Iran
*
m_moghaddas@alumni.iust.ac.ir
†
m_pandi@alumni.iust.ac.ir
Hakem Beitollahi
‡
School of Computer Engineering,
Iran University of Science and Technology,
Tehran, Iran
Department of Computer Science, Soran University,
Soran, Kurdistan Region, Iraq
hakem.beitollahi@soran.edu.iq
Received 16 April 2024
Accepted 27 January 2025
Published 29 April 2025
Physical Unclonable Functions (PUFs) are vulnerable to machine learning modeling attacks
that can predict their responses. To counter this threat, we introduce Obfuscated PUF (OPUF),
a novel PUF design that employs advanced obfuscation techniques to enhance security. OPUF
signi¯cantly outperforms existing PUFs in terms of resistance to machine learning attacks,
including logistic regression (LR) and multilayer perceptron (MLP). It achieves a 50.69% and
49.54% accuracy reduction in LR and MLP attacks, respectively, compared to the best-per-
forming existing PUFs. Moreover, OPUF maintains excellent quality metrics, with 50%
uniqueness and 50.15% uniformity, demonstrating its inherent randomness and unpredict-
ability. OPUF achieves its enhanced security through a multi-layered architecture that incor-
porates obfuscation techniques and nonlinear properties. By introducing internal challenges
that are di±cult for attackers to access, OPUF e®ectively thwarts modeling attempts. Our
experimental results demonstrate OPUF's superior performance in resisting various machine
learning attacks, even when faced with extensive training data. OPUF o®ers a promising
solution for secure hardware implementations, addressing the critical need for PUFs that can
withstand the challenges posed by modern machine learning attacks.
Keywords: Hardware security; physical unclonable function; machine learning; lightweight
encryption; PUF obfuscation.
‡
Corresponding author.
Journal of Circuits, Systems, and Computers
(2025) 2550245 (20 pages)
#
.
cWorld Scienti¯c Publishing Company
DOI: 10.1142/S0218126625502457
2550245-1
ResearchGate has not been able to resolve any citations for this publication.
FPGA-based Physical Unclonable Functions: A comprehensive overview of theory and architectures
Article
Full-text available
  • Jul 2021
  • INTEGRATION
Physically Unclonable Functions (PUFs) are a promising technology and have been proposed as central building blocks in many cryptographic protocols and security architectures. Among other uses, PUFs enable chip identifier/authentication, secret key generation/storage, seed for a random number generator and Intellectual Property (IP) protection. Field Programmable Gate Arrays (FPGAs) are re-configurable hardware systems which have emerged as an interesting trade-off between the versatility of standard microprocessors and the efficiency of Application Specific Integrated Circuits (ASICs). In FPGA devices, PUFs may be instantiated directly from FPGA fabric components in order to exploit the propagation delay differences of signals caused by manufacturing process variations. PUF technology can protect the individual FPGA IP cores with less overhead. In this article, we first provide an extensive survey on the current state-of-the-art of FPGA based PUFs. Then, we provide a detailed performance evaluation result for several FPGA based PUF designs and their comparisons. Subsequently, we briefly report on some of the known attacks on FPGA based PUFs and the corresponding countermeasures. Finally, we conclude with a brief overview of the FPGA based PUF application scenarios and future research directions.
View
Combining Optimization Objectives: New Modeling Attacks on Strong PUFs
Article
Full-text available
  • Feb 2021
Strong Physical Unclonable Functions (PUFs), as a promising security primitive, are supposed to be a lightweight alternative to classical cryptography for purposes such as device authentication. Most of the proposed candidates, however, have been plagued by modeling attacks breaking their security claims. The Interpose PUF (iPUF), which has been introduced at CHES 2019, was explicitly designed with state-of-the-art modeling attacks in mind and is supposed to be impossible to break by classical and reliability attacks. In this paper, we analyze its vulnerability to reliability attacks. Despite the increased difficulty, these attacks are still feasible, against the original authors’ claim. We explain how adding constraints to the modeling objective streamlines reliability attacks and allows us to model all individual components of an iPUF successfully. In order to build a practical attack, we give several novel contributions. First, we demonstrate that reliability attacks can be performed not only with covariance matrix adaptation evolution strategy (CMA-ES) but also with gradient-based optimization. Second, we show that the switch to gradient-based reliability attacks makes it possible to combine reliability attacks, weight constraints, and Logistic Regression (LR) into a single optimization objective. This framework makes modeling attacks more efficient, as it exploits knowledge of responses and reliability information at the same time. Third, we show that a differentiable model of the iPUF exists and how it can be utilized in a combined reliability attack. We confirm that iPUFs are harder to break than regular XOR Arbiter PUFs. However, we are still able to break (1,10)-iPUF instances, which were originally assumed to be secure, with less than 107 PUF response queries.
View
Splitting the Interpose PUF: A Novel Modeling Attack Strategy
Article
Full-text available
  • Jun 2020
We demonstrate that the Interpose PUF proposed at CHES 2019, an Arbiter PUF-based design for so-called Strong Physical Unclonable Functions (PUFs), can be modeled by novel machine learning strategies up to very substantial sizes and complexities. Our attacks require in the most difficult cases considerable, but realistic, numbers of CRPs, while consuming only moderate computation times, ranging from few seconds to few days. The attacks build on a new divide-and-conquer approach that allows us to model the two building blocks of the Interpose PUF separately. For non-reliability based Machine Learning (ML) attacks, this eventually leads to attack times on (kup, kdown)-Interpose PUFs that are comparable to the ones against max{kup, kdown}-XOR Arbiter PUFs, refuting the original claim that Interpose PUFs could provide security similar to (kdown + kup/2)-XOR Arbiter PUFs (CHES 2019). On the technical side, our novel divide-and-conquer technique might also be useful in analyzing other designs, where XOR Arbiter PUF challenge bits are unknown to the attacker.
View
The Interpose PUF: Secure PUF Design against State-of-the-art Machine Learning Attacks
Article
Full-text available
  • Aug 2019
The design of a silicon Strong Physical Unclonable Function (PUF) that is lightweight and stable, and which possesses a rigorous security argument, has been a fundamental problem in PUF research since its very beginnings in 2002. Various effective PUF modeling attacks, for example at CCS 2010 and CHES 2015, have shown that currently, no existing silicon PUF design can meet these requirements. In this paper, we introduce the novel Interpose PUF (iPUF) design, and rigorously prove its security against all known machine learning (ML) attacks, including any currently known reliability-based strategies that exploit the stability of single CRPs (we are the first to provide a detailed analysis of when the reliability based CMA-ES attack is successful and when it is not applicable). Furthermore, we provide simulations and confirm these in experiments with FPGA implementations of the iPUF, demonstrating its practicality. Our new iPUF architecture so solves the currently open problem of constructing practical, silicon Strong PUFs that are secure against state-of-the-art ML attacks.
View
A survey on physical unclonable function (PUF)-based security solutions for Internet of Things
Article
Full-text available
  • Oct 2020
  • COMPUT NETW
The vast areas of applications for IoTs in future smart cities, smart transportation systems, and so on represent a thriving surface for several security attacks with economic, environmental and societal impacts. This survey paper presents a review of the security challenges of emerging IoT networks and discusses some of the attacks and their countermeasures based on different domains in IoT networks. Most conventional solutions for IoT networks are adopted from communication networks while noting the particular characteristics of IoT networks such as the nodes quantity, heterogeneity, and the limited resources of the nodes, these conventional security methods are not adequate. One challenge towards utilizing common secret key-based cryptographic methods in large-scale IoTs is the problem of secret key generation, distribution, and storage and protecting these secret keys from physical attacks. Physically unclonable functions (PUFs) can be utilized as a possible hardware remedy for identification and authentication in IoTs. Since PUFs extract the unique hardware characteristics, they potentially offer an affordable and practical solution for secret key generation. However, several barriers limit the PUFs’ applications for key generation purposes. We discuss the advantages of PUF-based key generation methods, and we present a survey of state-of-the-art techniques in this domain. We also present a proof-of-concept PUF-based solution for secret key generation using resistive random-access memories (ReRAM) embedded in IoTs.
View
A Fast Deep Learning Method for Security Vulnerability Study of XOR PUFs
Article
Full-text available
  • Oct 2020
Physical unclonable functions (PUF) are emerging as a promising alternative to traditional cryptographic protocols for IoT authentication. XOR Arbiter PUFs (XPUFs), a group of well-studied PUFs, are found to be secure against machine learning (ML) attacks if the XOR gate is large enough, as both the number of CRPs and the computational time required for modeling n-XPUF increases fast with respect to n, the number of component arbiter PUFs. In this paper, we present a neural network-based method that can successfully attack XPUFs with significantly fewer CRPs and shorter learning time when compared with existing ML attack methods. Specifically, the experimental study in this paper shows that our new method can break the 64-bit 9-XPUF within ten minutes of learning time for all of the tested samples and runs, with magnitudes faster than the fastest existing ML attack method, which takes over 1.5 days of parallel computing time on 16 cores.
View
View
View
Hardware Supply Chain Security Threat Modelling, Emerging Attacks and Countermeasures: Threat Modelling, Emerging Attacks and Countermeasures
Book
  • Jan 2021
This book presents a new threat modelling approach that specifically targets the hardware supply chain, covering security risks throughout the lifecycle of an electronic system. The authors present a case study on a new type of security attack, which combines two forms of attack mechanisms from two different stages of the IC supply chain. More specifically, this attack targets the newly developed, light cipher (Ascon) and demonstrates how it can be broken easily, when its implementation is compromised with a hardware Trojan. This book also discusses emerging countermeasures, including anti-counterfeit design techniques for resources constrained devices and anomaly detection methods for embedded systems.
View
A Computationally Efficient Tensor Regression Network-Based Modeling Attack on XOR Arbiter PUF and Its Variants
Article
  • Oct 2020
XOR Arbiter PUF (XOR APUF), where the outputs of multiple APUFs are XOR-ed, has proven to be more robust to machine learning based modeling attacks. The reported successful modeling attacks for XOR APUF either employ auxiliary side-channel or reliability information, or require enormous computational effort. This robustness is primarily due to the difficulty in learning the unknown internal delay parameter terms in the mathematical model of a XOR APUF, and the robustness increases as the number of APUFs being XOR-ed increases. In this paper, we employ a novel machine learning based modeling technique called efficient CANDECOMP/PARAFAC-Tensor Regression Network (CP-TRN), a variant of CP-decomposition based tensor regression network, to reduce the computational resource requirement of model building attacks on XOR APUF. We theoretically prove the reduction in computational complexity, as well as give supporting experimental results. In addition, our proposed technique does not require any auxiliary information, and is robust to noisy training data. The proposed technique allowed us to successfully model 64-bit 8-XOR APUF and 128-bit 7-XOR APUF on a single desktop workstation, with high prediction accuracy. Further, we extend the proposed modeling attack technique to XOR APUF variants, e.g. Lightweight Secure PUF (LSPUF), which rely on input challenge transformation. The modeling accuracy results obtained by us for the LSPUF are comparable with those obtained by applying other state-of-the-art techniques, while requiring less training data.
View