Design and Analysis of Reliable Delay based PUF Architectures for Authentication

Abstract

In the present day scenario as numerous numbers of electronic devices are increased security plays a prominent role across all transactions of day to day communication. Over the last decade Internet of Things (IoT) strongly influences all the day to day activities. Especially smart devices and sensors occupy major space in all the domains to speed up the process. Due to the ever increasing number of these IoT devices, security plays one of the critical aspects to perform the faithful transaction from sender to receiver. However security and privacy rise to several criticalities in the IoT distributed chain. newlineThe conventional cryptographic techniques are more complex to implement in the IoT end points and occupy a large area. Like this there are several anomalies raising the IoT security chain. Hence to overcome these issues to enhance the level of security Physical Unclonable Functions (PUFs) have emerged to be one of the light weight security primitives in the IoT community. PUFs are acting as hardware digital signatures in the cryptographic area. For each set of challenges it produces a unique response, these challenges are considered as a challenge response pair. By using this behaviour these several kinds of PUF architectures were proposed and embedded in the IoT modules to generate and identify the unique set of keys. Even though several kinds of architectures were proposed, the IoT supply chain suffers from generating the unique set of secret keys and stability across operating conditions. newlineHence to address this above issues in this research proposed a strong delay based PUF architectures using digital delay elements. To enhance randomness proposed decoder-mux (4*1) PUF delivers the uniqueness of 47.5% and randomness of 45% respectively. To further enhance the randomness proposed a decoder-mux (2*1).whose uniqueness and randomness are 48.50% and 48%.Compared to liner mux based PUF architecture whose uniqueness and randomness is greater than 4.3% and 15.7%. newline