PI PUF: A Processor-Intrinsic PUF for IoT. (January 2023)

• Record Type: Journal Article
• Authors: Ni, Li; Wang, Pengjun; Zhang, Yuejun; Li, Gang; Ding, Lin; Zhang, Jiliang
• Is Part Of: Computers & electrical engineering. Volume 105(2023)
• Publication Date: 2023-01
• Subjects: Internet of Things (IoT) -- Processor-Intrinsic PUF (PI PUF) -- RISC-V processors -- Device authentication -- Hardware security; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2022.108540
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: Most IoT devices cannot offer sufficient resources for traditional cryptography technologies, posing a severe security challenge. Physical unclonable function (PUF) is a new lightweight hardware security primitive that extracts the entropy from process variations of circuits. However, the existing PUFs commonly are ASICs that require extra overhead on the area and extended instruction. To address such issues, this paper proposes a novel Processor-Intrinsic PUF (PI PUF). It exploits the pipeline design of processors and the timing dependence of data paths to generate abnormal information which carries the unclonable hardware characteristics. Moreover, a multi-round instruction obfuscation (MRIO) mechanism and the Gray code encoding method are conducted in the abnormal information to improve the uniqueness and randomness of our PUF. Finally, we design a lightweight authentication protocol based on the proposed PUF, whose security is further enhanced by shared key encryption and XOR random number encryption. Meanwhile, we also provide detailed security analysis under various attack scenarios. The Monte Carlo simulation results show that the proposed PUF achieves 50.93% of uniqueness, 50.62% of randomness, and 5.02% of the worst bit error rate, which outperforms the same type of state-of-the-art works. It also passes the NIST test.