Defense Against Chip Cloning Attacks Based on Fractional Hopfield Neural Networks

摘要: This paper presents a state-of-the-art application of fractional hopfield neural networks (FHNNs) to defend against chip cloning attacks, and provides insight into the reason that proposed method is superior physically unclonable functions (PUFs). In past decade, PUFs have been evolving as one best types hardware security. However, development has somewhat limited by its implementation cost, temperature variation effect, electromagnetic interference amount entropy in it, etc. Therefore, it imperative discover, through promising mathematical methods physical modules, some novel mechanisms overcome aforementioned weaknesses PUFs. Motivated this need, paper, we propose applying FHNNs attacks. At first, implement arbitrary-order fractor FHNN. Secondly, describe cost FHNNs. Thirdly, achievement constant-order performance FHNN when ambient varies. Fourthly, analyze electrical stability under disturbance conditions. Fifthly, study Lastly, perform experiments pass-band width an defense attacks capability particular, capabilities anti-electromagnetic interference, anti-temperature are illustrated experimentally detail. Some significant advantages their considerably lower than PUFs, much more stable different conditions, conditions robust significantly higher with same rank circuit scale.