Side-Channel Analysis for Malicious Activity Detection Using Deep Learning Techniques

Devajit Das, Manash Pratim Lahkar, Abhijit Gogor, Debojit Boro

2024

Abstract

The continued existence of malicious software constitutes a significant threat to network systems. This necessitates the urgent need for the development of robust detection mechanisms. The rapid development of malware variants frequently makes it challenging for traditional signature-based detection techniques. It has been found that side-channel analysis of physical systems can disclose sensitive data, including the secret key used for encryption, software activities, cryptographic operations, time-based features of the system, and data processing patterns. The side-channel analysis involves obtaining data via unintentional leakage channels from the physical system. The leakage channel may contain electromagnetic radiation, power consumption, timing information, acoustic emissions, cache access patterns, and other side-channel leakage from hardware. In this work, we employ deep-learning techniques with side-channel leakage from the hardware for identifying malicious activities within a system. We demonstrate the effectiveness of our technique by deploying deep neural networks for feature extraction and to recognize complicated correlations within data, specifically using Bidirectional Long Short-Term Memory (BiLSTM) networks. The findings of our experiments demonstrate the accuracy with which recurrent neural networks classify malware instances. We achieved 95.97%, 95.08%, and 92.46% accuracy, recall, and precision, respectively. Furthermore, we carried out real-time malware detection experiments to test our strategy for protecting systems from cyber threats.

Paper Citation

in Harvard Style

Das D., Lahkar M., Gogor A. and Boro D. (2024). Side-Channel Analysis for Malicious Activity Detection Using Deep Learning Techniques. In Proceedings of the 1st International Conference on Cognitive & Cloud Computing - Volume 1: IC3Com; ISBN 978-989-758-739-9, SciTePress, pages 123-129. DOI: 10.5220/0013345500004646

in Bibtex Style

@conference{ic3com24,
author={Devajit Das and Manash Lahkar and Abhijit Gogor and Debojit Boro},
title={Side-Channel Analysis for Malicious Activity Detection Using Deep Learning Techniques},
booktitle={Proceedings of the 1st International Conference on Cognitive & Cloud Computing - Volume 1: IC3Com},
year={2024},
pages={123-129},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013345500004646},
isbn={978-989-758-739-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Cognitive & Cloud Computing - Volume 1: IC3Com
TI - Side-Channel Analysis for Malicious Activity Detection Using Deep Learning Techniques
SN - 978-989-758-739-9
AU - Das D.
AU - Lahkar M.
AU - Gogor A.
AU - Boro D.
PY - 2024
SP - 123
EP - 129
DO - 10.5220/0013345500004646
PB - SciTePress