Novel Shuffling Countermeasure for Advanced Encryption Standard (AES) against Profiled Attack in Mobile Multimedia Services

• Published in: Wireless communications and mobile computing Vol. 2022; pp. 1 - 12
• Main Authors: Lee, JongHyeok, Kim, Jiyoon, Han, Dong-Guk
• Format: Journal Article
• Language: English
• Published: Oxford Hindawi 13.07.2022; Hindawi Limited
• Subjects: Algorithms; Artificial neural networks; Cryptography; Cybersecurity; Deep learning; Electronic devices; Insertion; Machine learning; Multimedia; Neural networks; Threat evaluation; Vibration; Wireless communications
• ISSN: 1530-8669; 1530-8677
• DOI: 10.1155/2022/6495546

Mobile multimedia services are gaining popularity among many users by developing wireless communication and mobile devices. Mobile multimedia has alleviated conventional multimedia’s time and space limits, making it easier for consumers to access services and meet content demands. However, cyber risks lie in the shadows of the expansion of mobile multimedia services, threatening to continue wreaking havoc. Although various methods exist to defend against these cyber threats, side-channel analysis has remained a critical challenge in the current approaches that rely on cryptographic algorithms. Nowadays, research on deep learning-based side-channel analysis is receiving much attention. Attacks are constantly performed against implementations, to which existing countermeasures against traditional side-channel analysis are applied, using various artificial neural network structures. However, while studies on the implementations to which masking and simple hiding schemes using jitter are active, studies on the implementations to which the shuffling scheme or the random insertion of dummy operations scheme are applied have been relatively less attention. In a previous study, Lee and Han has used deep learning to distinguish between real and dummy operations in an implementation that combined shuffling scheme and random insertion of dummy operations scheme. They also proposed countermeasures against their attacks. However, they did not choose an appropriate environment that is as close to noise-free as possible, and their countermeasure still has flaws. Therefore, in this study, we analyze the causes of vulnerability of the previous countermeasure and propose a novel countermeasure that can completely solve them. The novel countermeasure is a method of uniformly applying shuffling schemes and random insertion of dummy operation schemes to byte-independent and byte-dependent operations of an advanced encryption standard, respectively. It was confirmed that our countermeasure is safe from attackers who perform profiled attacks even in an experimental environment with almost no noise.