Home > Published Issues > 2022 > Volume 13, No. 5, October 2022 >
JAIT 2022 Vol.13(5): 456-461
doi: 10.12720/jait.13.5.456-461

GAAINet: A Generative Adversarial Artificial Immune Network Model for Intrusion Detection in Industrial IoT Systems

Siphesihle P. Sithungu and Elizabeth M. Ehlers
University of Johannesburg, Johannesburg, South Africa

Abstract—The expansion of the Internet of Things (IoT) in various industrial sectors (also referred to as the Industrial Internet of Things or IIoT) promises increased economic productivity and quality of life. However, the expansion of IIoT also presents unprecedented security concerns due to increased connectivity between appliances and the cloud. Among the security concerns on IIoT is the threat of intrusions on IIoT networks, resulting in unauthorised access to sensitive data generated by IIoT devices or the compromise of the entire IIoT network. Current work proposes a novel Generative Adversarial Artificial Immune Network (GAAINet) model for intrusion detection in IIoT systems. GAAINet aims to improve the quality of an Artificial Immune Network (AIN)-based classifier by introducing a generator AIN responsible for generating fake intrusion samples from a latent space to fool the classifier (or discriminator) AIN. The adversarial training of the generator and discriminator AINs is expected to improve the intrusion detection capability of the discriminator such that it potentially surpasses traditional training methods that only use preexisting datasets. Current work proposes GAAINet, an immunologically inspired generative adversarial conceptual model, for intrusion detection in IIoT systems.
 
Index Terms—immunologically inspired computation, generative adversarial models, artificial immune networks, industrial internet of things, industry 4.0
 
Cite: Siphesihle P. Sithungu and Elizabeth M. Ehlers, "GAAINet: A Generative Adversarial Artificial Immune Network Model for Intrusion Detection in Industrial IoT Systems," Journal of Advances in Information Technology, Vol. 13, No. 5, pp. 456-461, October 2022.

Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.