Digital twin-driven secured edge-private cloud Industrial Internet of Things (IIoT) framework.
In: Journal of Network & Computer Applications, Jg. 226 (2024-06-01), S. N.PAG
academicJournal
Zugriff:
With the growing popularity of Industrial Internet of Things (IIoT) technologies and the recent development of edge private cloud systems to fulfill the demands of industrial environments for high data rates, low latency, and powerful computing and storage resources, the security of these systems has become an increasingly important concern. Many existing machine learning-based attack detection models for the IIoT face challenges in the early identification of attacks. This is due to network traffic and physical process data's heterogeneous, high-dimensional, and unbalanced nature. Moreover, these models are typically trained offline and deployed in the cloud or embedded in devices, leading to resource strain and delayed attack detection. Thus, this paper proposes a real-time security framework for detecting attacks and mitigating their impact using Digital Twin and online ensemble machine learning. We explore various ensemble techniques and algorithms and evaluate their performance using gas pipeline and X-IIoTID datasets. The experimental results illustrate the effective performance of the proposed framework for detecting attacks, showcasing a comparable efficiency to offline ensemble techniques. [ABSTRACT FROM AUTHOR]
Titel: |
Digital twin-driven secured edge-private cloud Industrial Internet of Things (IIoT) framework.
|
---|---|
Autor/in / Beteiligte Person: | Al-Hawawreh, Muna ; Hossain, M. Shamim |
Zeitschrift: | Journal of Network & Computer Applications, Jg. 226 (2024-06-01), S. N.PAG |
Veröffentlichung: | 2024 |
Medientyp: | academicJournal |
ISSN: | 1084-8045 (print) |
DOI: | 10.1016/j.jnca.2024.103888 |
Schlagwort: |
|
Sonstiges: |
|