Abstract:Silicon Physical Unclonable Function (PUF) is arguably the most promising hardware security primitive. In particular, PUFs that are capable of generating a large amount of challenge response pairs (CRPs) can be used in many security applications. However, these CRPs can also be exploited by machine learning attacks to model the PUF and predict its response. In this paper, we first show that, based on data in the public domain, two popular PUFs that can generate CRPs (i.e., arbiter PUF and reconfigurable ring oscillator (RO) PUF) can be broken by simple logistic regression (LR) attack with about 99% accuracy. We then propose a feedback structure to XOR the PUF response with the challenge and challenge the PUF again to generate the response. Results show that this successfully reduces LR's learning accuracy to the lower 50%, but artificial neural network (ANN) learning attack still has an 80% success rate. Therefore, we propose a configurable ring oscillator based dual-mode PUF which works with both odd number of inverters (like the reconfigurable RO PUF) and even number of inverters (like a bistable ring (BR) PUF). Since currently there are no known attacks that can model both RO PUF and BR PUF, the dual-mode PUF will be resistant to modeling attacks as long as we can hide its working mode from the attackers, which we achieve with two practical methods. Finally, we implement the proposed dual-mode PUF on Nexys 4 FPGA boards and collect real measurement to show that it reduces the learning accuracy of LR and ANN to the mid-50% and low 60%, respectively. In addition, it meets the PUF requirements of uniqueness, randomness, and robustness.

A Reconfigurable and Machine Learning Attack Resistant Strong PUF Based on Arbiter Mechanism and SOT-MRAM.

A Reconfigurable Arbiter MPUF With High Resistance Against Machine Learning Attack

A Modeling Attack Resilient Physical Unclonable Function Based on STT-MRAM

A Novel Configurable RO-Obfuscated PUF Design with Machine Learning Immunity

A Reconfigurable Arbiter PUF Based on STT-MRAM

A Machine Learning Attack-Resilient Strong PUF Leveraging the Process Variation of MRAM

SCD-PUF: Shuffled Chaotic-dual-PUF with High Machine Learning Attack Resilience

DMOS-PUF: Dynamic Multi-key-selection Obfuscation for Strong PUFs against Machine Learning Attacks.

STT-MRAM-Based Reliable Weak PUF

A novel PUF against machine learning attack: Implementation on a 16 Mb RRAM chip

A Response-Feedback-Based Strong PUF with Improved Strict Avalanche Criterion and Reliability

A Machine Learning Attack Resistant Dual-mode PUF

A High Area-Efficiency RRAM-Based Strong PUF with Multi-Entropy Source and Configurable Double-Read Process

Memristive Crypto Primitive for Building Highly Secure Physical Unclonable Functions

A Dynamically Configurable LFSR-based PUF Design Against Machine Learning Attacks.

Reconfigurable and Dynamically Transformable In-Cache-MPUF System with True Randomness Based on the SOT-MRAM

A Novel RRAM Based PUF for Anti-Machine Learning Attack and High Reliability

Physical Unclonable Function Chip based on SOT-MRAM in 180 nm CMOS Technology

A Sequential Strong PUF Architecture Based on Reconfigurable Neural Networks (rnns) Against State-of-the-art Modeling Attacks.

High-Speed and Reconfigurable Physical Unclonable Functions Based on SOT-MTJ Array

Modeling Attack Resilient Reconfigurable Latent Obfuscation Technique for PUF based Lightweight Authentication