Reconfigurable Physical Unclonable Function-Based Cryptographic Scheme Based on Memory-Enabled GFET Modulators

Area of Science:

• Materials Science
• Electrical Engineering
• Cybersecurity

Background:

• The Internet of Things (IoT) generates vast amounts of data, increasing vulnerability to cyberattacks.
• Existing hardware security primitives face challenges in providing robust protection for sensitive information.

Purpose of the Study:

• To introduce a novel hardware security primitive: a strong radio-frequency (RF) physically unclonable function (PUF).
• To leverage graphene field-effect transistors (GFETs) for enhanced security applications in the IoT era.

Main Methods:

• Development of a GFET-based harmonic transponder for the RF PUF.
• Demonstration of PUF reconfigurability through gate voltage, frequency tuning, and doping state adjustments.
• Implementation of an AI-assisted classification framework using a 1D-CNN for device authentication.

Main Results:

• The GFET PUF demonstrated excellent randomness, uniqueness, and reliability for encryption.
• Reconfigurability enhanced PUF robustness against environmental factors.
• AI framework achieved robust device recognition across multiple operating conditions via contrastive learning and cosine similarity.

Conclusions:

• The proposed GFET PUF offers a significant advancement in hardware security for IoT devices.
• This technology presents a new pathway for analog PUFs and cryptographic applications.
• The GFET PUF enhances data security and device authentication in connected systems.