A physically unclonable function circuit of a reconfigurable ring oscillator and its excitation generation method

Abstract

The present invention proposes a reconfigurable ring oscillator physically unclonable function circuit and its excitation generation method. The reconfigurable ring oscillator physically unclonable function circuit uses a control module to randomly select delay elements according to the excitation signal to form a ring oscillation The device increases the flexibility and randomness of the ring oscillator, improves the unpredictability of the PUF response, and can increase the difficulty for attackers to model attacks through machine learning; the incentive generation method can fully avoid the same incentive from being used multiple times. Second, further improve the unpredictability of PUF response and improve the effect of PUF security.

Description

technical field [0001] The invention relates to a physically unclonable function circuit of a reconfigurable ring oscillator, in particular to a physically unclonable function circuit of a reconfigurable ring oscillator and an excitation generation method thereof. Background technique [0002] The Internet of Things has revolutionized our lives through telemedicine, self-driving cars, smart homes, and more. However, it also poses security and privacy concerns, as it provides criminal hackers with new attack vectors. The Internet of Things is expected to have a significant impact on a wide range of markets, from wearable medical devices to embedded systems in smart cars, many of which will be underpinned by devices with limited computing and power consumption. Traditional security methods based on complex cryptographic algorithms are usually too resource-intensive to be implemented on these resource-constrained devices. A Physically Unclonable Function (PUF) is a security p...

AI Technical Summary

Problems solved by technology

Using the random error introduced by the PUF circuit manufacturing process, the response to the input stimulus is unpredictable

However, under certain circumstances, if the attacker knows a certain number of incentive-response pairs (CRPs), it is possible to conduct machine learning modeling attacks on PUFs, thereby predicting its response to certain incentives and destroying the security of PUFs.

Images