Adversarial Attack against Modeling Attack on PUFs

Chen, Yu-Shen; Li, Katherine Shu-Min

doi:10.1145/3316781.3317761

Cited by 25 publications

(14 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This pulls the accuracy down and forces it to a value below τ , as confirmed by the results. As will be discussed below if the CRPs are poisoned in a periodic fashion (similar to [18]), the PUF can be modeled. Accordingly, to ensure that the proposed scheme does not follow a periodic poisoning pattern, Fig.…”

Section: A Experimental Resultsmentioning

confidence: 99%

“…The next set of results compares the efficacy of the proposed method with the AML-based scheme presented in [18] that introduced different poisoning schemes. In [18], the response toggling decision is made either based on a combinational function applied on each challenge bit-stream (when satisfied, the PUF response is poisoned) or the poisoning is performed in a periodic fashion. Although claimed to be resilient against modeling, we were able to attack these schemes with a high success.…”

Section: A Experimental Resultsmentioning

confidence: 99%

“…Wang [18] used adversarial models to fool an attacker who eavesdrops CRPs. The approach changes the response of some challenges based on a functions of the challenge bits or in a periodic manner.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Countering PUF Modeling Attacks through Adversarial Machine Learning

Ebrahimabadi

Lalouani

Younis

et al. 2021

2021 IEEE Computer Society Annual Symposium on VLSI (ISVLSI)

View full text Add to dashboard Cite

A Physically Unclonable Function (PUF) is an effective option for device authentication, especially for IoT frameworks with resource-constrained devices. However, PUFs are vulnerable to modeling attacks which build a PUF model using a small subset of its Challenge-Response Pairs (CRPs). We propose an effective countermeasure against such an attack by employing adversarial machine learning techniques that introduce errors (poison) to the adversary's model. The approach intermittently provides wrong response for the fed challenges. Coordination among the communicating parties is pursued to prevent the poisoned CRPs from causing the device authentication to fail. The experimental results extracted for a PUF implemented on FPGA demonstrate the efficacy of the proposed approach in thwarting modeling attacks. We also discuss the resiliency of the proposed scheme against impersonation and Sybil attacks.

show abstract

Section: A Experimental Resultsmentioning

confidence: 99%

Section: A Experimental Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Countering PUF Modeling Attacks through Adversarial Machine Learning

Ebrahimabadi

Lalouani

Younis

et al. 2021

2021 IEEE Computer Society Annual Symposium on VLSI (ISVLSI)

View full text Add to dashboard Cite

show abstract

“…Tackling PUF modeling attacks via PUF response poisoning can also be realized via adversarial ML. The approach proposed in [117] sends the toggled PUF response (instead of the genuine response) whenever the challenge bit-stream follows a specific bit pattern or in a periodic manner. In the latter, the PUF response is toggled for every other N received challenges.…”

Section: B Modeling Attacks On Pufs and Robustness Enhancementsmentioning

confidence: 99%

Two Sides of the Same Coin: Boons and Banes of Machine Learning in Hardware Security

Liu

Chang

Wang

et al. 2021

IEEE J. Emerg. Sel. Topics Circuits Syst.

View full text Add to dashboard Cite

The last decade has witnessed remarkable research advances at the intersection of machine learning (ML) and hardware security. The confluence of the two technologies has created many interesting and unique opportunities, but also left some issues in their wake. ML schemes have been extensively used to enhance the security and trust of embedded systems like hardware Trojans and malware detection. On the other hand, ML-based approaches have also been adopted by adversaries to assist side-channel attacks, reverse engineer integrated circuits and break hardware security primitives like Physically Unclonable Functions (PUFs). Deep learning is a subfield of ML. It can continuously learn from a large amount of labeled data with a layered structure. Despite the impressive outcomes demonstrated by deep learning in many application scenarios, the dark side of it has not been fully exposed yet. The inability to fully understand and explain what has been done within the super-intelligence can turn an inherently benevolent system into malevolent. Recent research has revealed that the outputs of Deep Neural Networks (DNNs) can be easily corrupted by imperceptibly small input perturbations. As computations are brought nearer to the source of data creation, the attack surface of DNN has also been extended from the input data to the edge devices. Accordingly, due to the opportunities of MLassisted security and the vulnerabilities of ML implementation, in this paper, we will survey the applications, vulnerabilities and fortification of ML from the perspective of hardware security. We will discuss the possible future research directions, and thereby, sharing a roadmap for the hardware security community in general.

show abstract

“…For that reason, it is almost impossible to fabricate two identical chips, even with full knowledge of the original chip's design [13]. Thus, physical unclonability defines that two PUF instances should never have identical challenge-response pair (CRPs) behavior [26], making them ideal for security purposes. PUFs have been applied as an effective, lightweight mechanism for secure authentication [19] and secret key generation [24].…”

Section: Introductionmentioning

confidence: 99%

Noise-Resilient Neural Network-Based Adversarial Attack Modeling for XOR Physical Unclonable Functions

Aseeri

2020

JCSANDM

View full text Add to dashboard Cite

Authentication plays an essential role in preventing unauthorized access to data and resources on the Internet of Things. Classical security mechanisms fall short of achieving security requirements against many physical attacks due to the resource-constraint nature of IoT devices. Physical Unclonable Functions (PUFs) have been successfully used for lightweight security applications such as devices authentication and secret key generation. PUFs utilize the inevitable variation of integrated circuits during the fabrication process to produce unique responses for individual PUF, hence not reproducible even by the manufacturer itself. However, PUFs are mathematically cloneable by machine learning-based methods. XOR arbiter PUFs are one group of PUFs that can withstand existing attack methods unless exceedingly long training time and large dataset size are applied. In this paper, large-sized XOR PUFs with 64-bit and 128-bit challenges were efficiently and effectively attacked using a carefully engineered neural network-based method. Our fine-tuned neural network-based adversarial models achieve 99% prediction accuracy on noise-free datasets and as low as 96% prediction accuracy on noisy datasets, using up to 55% smaller dataset size compared to existing works known to us. Revealing such vulnerabilities is essential for PUF developers to re-evaluate existing PUF designs, hence avoiding potential risks for IoT devices.

show abstract

Adversarial Attack against Modeling Attack on PUFs

Cited by 25 publications

References 22 publications

Countering PUF Modeling Attacks through Adversarial Machine Learning

Countering PUF Modeling Attacks through Adversarial Machine Learning

Two Sides of the Same Coin: Boons and Banes of Machine Learning in Hardware Security

Noise-Resilient Neural Network-Based Adversarial Attack Modeling for XOR Physical Unclonable Functions

Contact Info

Product

Resources

About