A New Approach to Device Identification Using ICMetric and Chaotic Maps

Abstract

The rapid expansion of the Internet has increased security risks for connected devices. Traditional security systems rely on stored templates or cryptographic keys; if these are breached, the entire system becomes vulnerable. In addition, electronic systems have become increasingly vulnerable to identity theft, spoofing, and impersonation, compromising the system's overall integrity. To address and improve the security of the mentioned issues, various techniques and approaches have been explored, one of which is ICMetrics. This approach presents a more secure alternative by generating cryptographic keys from unique hardware or software behaviors, thereby eliminating the need to store sensitive data. However, ICMetrics can face limitations when device behaviors are too similar, reducing key uniqueness. To overcome this, the paper introduces the use of a logistic map chaotic system to enhance the randomness and strength of ICMetric-generated keys. Due to its sensitivity to initial conditions, the logistic map produces highly unpredictable sequences suitable for cryptography. Entropy analysis of keys derived from different decimal positions of the chaotic signal shows consistently high randomness, especially from the second digit onward. Experiments across key lengths from 128 to 2048 bits demonstrate near-maximal entropy values (~8 bits/byte) and improved key generation efficiency. The new approach, ICMetric-Chaotic is based on the ICM-RSA framework and achieves faster key generation without compromising security, offering a scalable and reliable method for enhancing ICMetric-based cryptographic systems, particularly for securing embedded devices, and gives a solution for the ICMetrics limitation.

Keywords