Modeling Attack Resistant Arbiter PUF based on Dynamic Finite Field Matrix Multiplication scheme

The physical unclonable function (PUF) is widely applied in the field of information security. However, the strong PUFs are vulnerable to modeling attacks. In this paper, we propose an arbiter PUF based on dynamic finite field matrix multiplication scheme (DFM-APUF), which uses an obfuscation matrix multiplied with challenges in dynamic varying finite fields to achieve high resistance to modeling attacks. The proposed DFM-APUF is implemented on Xilinx Artix-7 FPGAs. The experiment results show that the prediction accuracies of LR, ES, SVM, and ANN are below 51%, with negligible effects on randomness, reliability, and uniqueness of APUF.