Detecting Cryptographic Hash Functions through Electromagnetic Side-Channel Analysis

Abstract

the era of Industry 4.0, the Internet of Things (IoT) has emerged as a transformative force, with the proliferation of IoT devices permeating various aspects of our daily lives. However, this rapid adoption of IoT technology has also given rise to an alarming increase in cyberattacks targeting these devices. Among many avenues of cybersecurity, Electromagnetic Side Channel Analysis (EM-SCA) stands as a crucial branch of information security that enables attackers to eavesdrop on and exfiltrate s ensitive i nformation, m aking i t a c ritical concern for IoT security. Among various security measures taken on IoT platforms, data integrity is ensured through cryptographic hash functions. This work explores the potential of utilising EM-SCA to detect the presence of cryptographic hash functions on IoT devices, which would play an important role at the surveillance stage of an attack. In pursuit of this objective, this study employs a set of supervised Machine Learning (ML) algorithms that are intricately crafted to automatically identify distinct patterns of EM radiation emissions associated with different hash algorithms. The results of this investigation demonstrate that the proposed methods can achieve classification accuracy rates exceeding 80%. The findings o f t his w ork h ighlights that an attacker can inspect an IoT device in a non-invasive manner to identify its critical data integrity mechanisms before a suitable subsequent action is taken to compromise it.