ENHANCING THE CYBERSECURITY OF INDUSTRIAL INTERNET OF THINGS INFRASTRUCTURE THROUGH AN INTELLIGENT HYBRID DEEP LEARNING BASED BOTNET DETECTION AND PREVENTION FRAMEWORK

Abstract

The rapid developmental pace of the Industrial Internet of Things (IIoT) has significantly aided automation processes in the industries, efficiency in the operations, and networking within the system. However, this shift has also given rise to new security threats and more so with the increasing sophistication and scale of cyberattacks by means of botnet on vital industrial systems. The study project will provide an effective Hybrid Deep Learning-based security architecture, which is solely designed to safeguard IIoT systems against advanced botnet attacks. The proposed methodology relies on the synergistic theory of integrating the Convolutional Neural Networks (CNNs) to obtain spatial feature at the higher level with Recurrent Neural Networks (RNNs) to acquire the temporal relation among the traffic and the serial nature of the attacks. This hybrid design enables the characterization of fixed and dynamically varying network patterns, as well as an evaluation to a large extent further since the model is trained on heterogeneous and different IIoT traffic patterns. This further enhancement by adding anomaly detection mechanisms and deep feature learning is to bolster the abilities of the system to distinguish between malicious botnet entries and normal industrial traffic. In turn, the hybrid deep learning model offered outperforms the existing state-of-the-art approaches in the areas of detection accuracy, robustness, scalability, and computational cost which predetermines its applicability in resource-based IIoT systems and enables the use of the model in real-time.Substantial experimental evaluations attest to the superiority of the hybrid deep learning model to the existing state-of-the-art solutions both in the area of detection accuracy and low false-positive rates and computation cost. This study presents both scalability and resilience in the defense of IIoT-enabled critical infrastructure through the provision of real-time detection of the threats and preventive actions to contain the dynamic world of botnets that, despite its much larger size, is no different than the threat posed on the infrastructure.