Fault-Tolerant Reversible-Logic Based RO-PUF for Secure Device Authentication
Protecting data and hardware is vital, driving the adoption of Physically Unclonable Functions (PUFs) for generating unique circuit signatures. This paper introduces a fault-tolerant system featuring a ring-oscillator (RO) based PUF, utilizing a reversible logic (RL) design. The proposed system comp...
Gespeichert in:
Veröffentlicht in: | IEEE transactions on circuits and systems. I, Regular papers Regular papers, 2024-12, Vol.71 (12), p.5828-5837 |
---|---|
Hauptverfasser: | , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Protecting data and hardware is vital, driving the adoption of Physically Unclonable Functions (PUFs) for generating unique circuit signatures. This paper introduces a fault-tolerant system featuring a ring-oscillator (RO) based PUF, utilizing a reversible logic (RL) design. The proposed system comprises various sub-systems such as Fault-Tolerant RL-based inverter design, Reversible-Logic designing, Fault-Detection module, Fault-free path selection module, and the Reversible RO-PUF module. The proposed design is implemented on a Basys-3 FPGA board for calculating various PUF parameters. It is observed that the uniqueness, uniformity, and bit-aliasing of the proposed design at 27°C are 49.40%, 51.20%, and 48.30%, respectively. Further, bit-error-rate (BER), reliability, and key error rate (KER) are determined at three different temperatures, and the best results obtained are 0.003%, 99.7%, and 0.092 at 40°C, respectively. Compared to conventional PUFs, the proposed design showcases higher reliability (0.002% to 0.11%) and significantly reduced BER and KER ( 1.67\times to 22.67\times , and 1.6\times to 8.02\times respectively). The proposed design also passed 15 NIST tests against conventional RO-PUF, which could pass only 11 NIST tests. We have also tested the resilience of different PUF designs against three machine-learning models with the best accuracy of 58.9% against the Logistic Regression model. |
---|---|
ISSN: | 1549-8328 1558-0806 |
DOI: | 10.1109/TCSI.2024.3425957 |