The widespread use of the internet, combined with the prevalence of cybersecurity threats such as botnets, has resulted in significant economic losses for manufacturing enterprises. An AI-powered network intrusion detection system is required to address the growing number of botnet attacks caused by increased machine-to-machine communication. Several Machine Learning (ML) and Deep Learning (DL) algorithms were used in this study to detect botnet attacks on seven IoT devices, with the goal of developing highly secure and accurate models for detecting security threats. After comparing its performance to other models, the proposed model was found to be highly performant, accurate, and robust for threat detection. After implementing different models, 2-CNN model demonstrated the highest accuracy level of 99.95% in DANMINI Video Door Phone Doorbell Hands-free Wireless Intercom. Furthermore, it was noted that the performance of the Deep Belief Network (DBN) model was inferior to that of other Deep Learning (DL) models in identifying Gafgyt botnet attacks.
CITATION STYLE
Shahin, M., Chen, F. F., Hosseinzadeh, A., Lopez, E. C., Bouzary, H., & Koodiani, H. K. (2024). An AI-Powered Network Intrusion Detection System in Industrial IoT Devices via Deep Learning. In Lecture Notes in Mechanical Engineering (pp. 1149–1156). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-38165-2_131
Mendeley helps you to discover research relevant for your work.