The acceptance of 5G technology depends on its security. 5G introduces new dynamics in terms of network softwareization, network function virtualization (NFV), dynamic network slicing, SDN, and Service customization [1]. While these provide the required functionalities to ensure 5G principles in terms of dynamic configuration, flexibility, scalability, and elasticity [1], they introduce increased attack space [2], causing additional security challenges and complexities in some cases [3, 4, 5 ]. Therefore, security services must be adapted to meet the resource-constrained nature of the IoT and new 5G dynamics. Thus, there is a need to develop an adaptive anomaly detection and prediction approach for security-related data collection, analytics, and prediction of incidents and the provision of response and mitigation measures autonomously to systematically understand, characterize, quantify, and manage cybersecurity in 5G -enabled IoT. It will apply closed-loop AI techniques in a privacy-preserving manner and adapt to security changes and contexts in the 5G-IoT dynamics and characteristics.
Tasks:
T1: To study related literature review and prepare a short report
T2: To test the AI-based adaptive anomaly detection for 5G-enabled IoT
T3: To compare and benchmark with already developed technique(s)
References
[1] H. Hellaoui, M. Koudil and A. Bouabdallah, "Energy Efficiency in Security of 5G-Based IoT: An End-to-End Adaptive Approach," in IEEE Internet of Things Journal, vol. 7, no. 7, pp. 6589-6602, July 2020, doi: 10.1109/JIOT.2020.2974618.
[2] ENISA Report NFV SECURITY IN 5G, February 2022
[3] Afaq, A., Haider, N., Baig, MZ, Khan, KS, Imran, M., & Razzak, I. (2021). Machine learning for 5G security: Architecture, recent advances, and challenges. Ad Hoc Networks, 123, 102667.
[4] Hussain, R., Hussain, F., Zeadally, S., & Lee, J. (2021). On the adequacy of 5G security for vehicular ad hoc networks. IEEE Communications Standards Magazine, 5(1), 32-39.
The project is in collaboration with Norsk Regnesentral (Norwegian Computing Center)
