Design and Implementation of Multiplexed and Obfuscated Physical Unclonable Function

Model building attack on Physical Unclonable Functions (PUFs) by using machine learning (ML) techniques has been a focus in the PUF research area. is hardware security primitive which can extract unique characteristics (i.e., device-specific) exploiting intrinsic manufacturing process variations during integrated circuit (IC) fabrication. The nature of random and complex makes realistically physically impossible to clone atom-by-atom. Nevertheless, its function vulnerable model-building attacks ML techniques. Arbiter-PUF one earliest proposed delay-based PUFs ML-attack. In past, several have increase resiliency, but often sacrifice reproducibility response. this paper, we propose new derivative called Mixed (MA-PUF). Four Arbiter-PUFs are combined their outputs multiplexed generate final We show that MA-PUF good properties uniqueness, reliability, uniformity. Moreover, resilient against ML-attack 15% better than conventional Arbiter-PUF. predictability close 65% could be achieved when combining with challenge permutation technique.