Attacking Arbiter PUFs Using Various Modeling Attack Algorithms: A Comparative Study
Yue Fang, Chenghua Wang, Qingqing Ma, Chongyan Gu, Maire OrNeill, Weiqiang Liu · 2018
In this paper, we investigate the effectiveness of four different modeling attack algorithms, including Logistic Regression (LR), Naïve Bayes, AdaBoost and Covariance Matrix Adaptation Evolutionary Strategies (CMA-ES), on attacking arbiter physical unclonable functions (APUFs). A comparison of experimental results using theses algorithms is presented. The results show that the performance of the algorithms is related to the number of training data, the noise level involved in the APUF design and the number of stages in the generation of each bit response. It is found that the mainstream LR and CMA-ES are worse for a small number of data compared with Naïve Bayes and AdaBoost.