Intrusion Detection and Obfuscation Mechanism for PUF-Based Authentication

• Published in: 2022 IEEE 31st Asian Test Symposium (ATS) pp. 90 - 95
• Main Authors: Wang, Sying-Jyan, Li, Katherine Shu-Min, Lin, Chen-Yeh, Chong, Song-Kong
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2022
• Subjects: Authentication; Intrusion detection; Machine learning; modeling attack; Physical unclonable function; Servers
• ISSN: 2377-5386
• DOI: 10.1109/ATS56056.2022.00028

Physical Unclonable Functions (PUF) provide a lightweight secure primitive. However, it is well known that PUFs are vulnerable to machine learning based modeling attacks (MA). A lot of efforts have been devoted to develop MA-resistant PUFs, while MA techniques are also evolving. In this paper, we propose to protect PUF-based authentication systems by providing an intrusion detection mechanism. By replacing some responses with the check information only known to the server and the PUF, any attacks will trigger alerts to the server. The proposed method provides an active protection mechanism so that the verifier can initiate an intrusion detection process to determine whether the prover is an attacker in disguise. In this way, the proposed method can provide an early alert so as to achieve another layer of protection. Experimental results show that the proposed mechanism can effectively provide alerts after PUFs being attacks.