Deep Learning Models for Cyber Attack Detection

Internet of Things (IoT) has made Cyber Physical Systems (CPS) more secure. This has created unique security challenges as the traditional security solutions are meant for general IT or OT systems, which do not perform well in a CPS enabled environment. With an emphasis on Industrial Control Systems...

Full description

Saved in:

Bibliographic Details
Published in:	2024 8th International Conference on Inventive Systems and Control (ICISC) pp. 544 - 549
Main Authors:	Brintha, N. C., Hrudayesh, B., Anil, M., Reddy, A. Goutham, Guru Nandan, V.
Format:	Conference Proceeding
Language:	English
Published:	IEEE 29-07-2024
Subjects:	Computational modeling Control systems Cyber Physical Systems (CPS) Data models Deep learning Deep Neural Network(DNN) Feature extraction Industrial control Industrial Control System (ICS) Internet of Things (IoT) Vectors
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Internet of Things (IoT) has made Cyber Physical Systems (CPS) more secure. This has created unique security challenges as the traditional security solutions are meant for general IT or OT systems, which do not perform well in a CPS enabled environment. With an emphasis on Industrial Control Systems (ICS), this research study offers a thorough two-level ensemble attack detection and attribution system designed especially for Cyber Physical Systems (CPS). An ensemble deep learning model, which is especially good at identifying attacks in unbalanced Industrial Control Systems (ICS) environments, is combined with a decision tree at the first level. Moving on to the next stage, an ensemble deep neural network is designed to deliver accurate attack attribution. Real-time datasets are used to confirm that the proposed model works as intended. The results show better performance than other approaches with comparable computational complexity.
DOI:	10.1109/ICISC62624.2024.00096