Physical unclonable function using photonic spin Hall effect.

Area of Science:

• Optoelectronics
• Quantum Optics
• Hardware Security

Background:

• Physical unclonable functions (PUFs) are essential for hardware security, relying on unique device variations for authentication.
• Existing PUF designs face challenges in achieving high uniqueness and robustness against sophisticated attacks.
• The photonic spin Hall effect (PSHE) has emerged as a promising phenomenon for advanced sensing and computing applications.

Purpose of the Study:

• To develop a novel physical unclonable function (PUF) by leveraging the surface wave-assisted photonic spin Hall effect (PSHE).
• To evaluate the performance and security of the proposed PSHE-based PUF token against noise and machine learning attacks.
• To explore the potential of PSHE in expanding the design space for hardware security solutions.

Main Methods:

• Constructing PUF devices utilizing the photonic spin Hall effect (PSHE) induced by surface waves.
• Characterizing PUF performance using metrics such as inter-Hamming distance and proportion of unique responses.
• Assessing PUF security against a multi-layer perceptron (MLP) regression model employing a randomized search attack strategy.

Main Results:

• The proposed PUF token achieved an inter-Hamming distance of 47.50% and an average proportion of unique responses of 62.5%.
• The PUF demonstrated robustness against simulated noise and a low average successful attack prediction accuracy of 9.70%.
• The PSHE effect imparts high non-linearity and sensitivity to process variations, enhancing security and unclonability.

Conclusions:

• The novel PSHE-based PUF token offers a promising approach for secure hardware authentication.
• The design exhibits excellent resilience to noise and advanced machine learning-based attacks.
• This work highlights the potential of photonic phenomena in advancing the field of hardware security.