AI-Driven threat detection for securing 5G network slicing in hybrid cloud environments amid 2026 attacks
1 Department of Physics, University of Ilorin, P.M.B. 1515, Ilorin, Kwara State, Nigeria.
2 Levin College of Public Affairs and Education, Cleveland State University, 2121 Euclid Avenue, Cleveland, OH 44115, USA.
3 Ivan Hilton Center for Science Technology, Department of Computer Science, New Mexico Highlands University, Las Vegas, NM, USA.
4 McClure School of Emerging Communication Technologies, Ohio University, Athens, OH, USA.
5 College of Engineering, Northeastern University, Boston, MA, USA.
Research Article
World Journal of Advanced Research and Reviews, 2024, 24(03), 3648-3666
Publication history:
Received on 17 November 2024; revised on 23 December 2024; accepted on 29 December 2024
Abstract:
The adoption of the fifth-generation (5G) networks has come with revolutionary opportunities for network slicing, enabling various virtualized networks to coexist on the same infrastructure. Nevertheless, this proliferation of new technology has greatly increased the attack surface which exposes critical vulnerabilities including cross-slice attacks, resource exploitation, and unauthorized access. This paper focuses on AI-based threat detection systems that are specifically engineered to secure 5G network slicing architectures that are implemented in the hybrid cloud environments. The study uses a Transformer-based intrusion detection system with multi-head self-attention mechanisms to solve the upcoming security threats that are expected in the scenario of attack in 2026. The model was trained and tested on the 5G Network Intrusion Detection Dataset (5G-NIDD), achieving superior performance in the multi-class classification task, which involves both attack detection and the classification of the type. A comparison and analysis with the baseline models which comprise the Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM), ensemble Autoencoder-Support Vector Machines (AE-SVM), and Gradient Boosting discovered that the Transformer-based system had the highest detection accuracy of about 98.2% with an attack recall of 96.5%. The paper provides repeatable experimental procedures, complete performance indicators, and feasible suggestions on how AI-oriented security solutions can be incorporated into the functioning 5G networks. Future research directions include improving model interpretability through attention visualization and exploring federated learning to share threat intelligence among operators.
Keywords:
5G Security; Network Slicing; Intrusion Detection; Transformer Models; Hybrid Cloud; Threat Detection; Artificial Intelligence; Deep Learning; Cyber Attacks; Software-Defined Networking
Full text article in PDF:
Copyright information:
Copyright © 2024 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0