A novel physical unclonable function (PUF) using 16 × 16 pure-HfOferroelectric tunnel junction array for security...

Area of Science:

• Materials Science
• Electrical Engineering
• Computer Science

Background:

• Edge computing necessitates enhanced security for resource-constrained devices.
• Conventional cryptography is often unsuitable for edge devices due to power and area limitations.
• Physical unclonable functions (PUFs) provide a power- and area-efficient security solution using inherent device variations.

Purpose of the Study:

• To demonstrate the physical unclonable function (PUF) operations of a novel ferroelectric tunnel junction (FTJ) array.
• To evaluate the security and robustness of the FTJ-based PUF against advanced attacks.
• To explore the potential of FTJs for low-power, small-area security applications in edge computing.

Main Methods:

• Fabrication and characterization of a 16x16 hafnium oxide (pure-HfO2)-based ferroelectric tunnel junction (FTJ) array.
• Implementation of PUF operations utilizing the inherent process and domain variations within the FTJ array.
• Evaluation of PUF randomness using established metrics and assessment of robustness against model-based machine learning attacks.

Main Results:

• Successful demonstration of PUF operations in a scaled 16x16 pure-HfO2 FTJ array.
• The FTJ array exhibits adequate randomness, suitable for cryptographic key generation.
• Array-level PUF operations proved robust against sophisticated model-based machine learning attacks.

Conclusions:

• The proposed FTJ array offers a viable, low-power, and small-area solution for physical unclonable functions.
• This technology addresses the security challenges posed by resource-limited edge computing environments.
• The demonstrated robustness against machine learning attacks highlights the potential of FTJ-based PUFs for secure applications.