Secure Reinforcement Learning-Based Network Slice Allocation Against Adversarial Traffic Attacks
Keywords:
Network slicing, reinforcement learning, adversarial robustness, 5G security, resource allocation, traffic attacks, system resilienceAbstract
The proliferation of network slicing in 5G and beyond-5G infrastructures has introduced unprecedented flexibility in delivering differentiated services over shared physical resources, yet the growing reliance on reinforcement learning (RL) for dynamic slice allocation exposes these systems to adversarial traffic attacks that manipulate demand signals and degrade quality of service. This paper presents a system-level investigation into secure RL-based network slice allocation mechanisms that withstand adversarial perturbations while maintaining operational efficiency. A comprehensive adversarial threat model is developed, categorizing observation manipulation, reward poisoning, and black-box policy induction attacks that exploit the open nature of user-plane traffic. The proposed secure allocation framework integrates robust state sanitization, adversarially trained RL agents, behavioral anomaly monitoring, and distributed trust mechanisms to form a defense-in-depth architecture. The paper examines structural trade-offs between security overhead and resource utilization, response agility and resilience, and fairness under adversarial stress. Deployment considerations spanning hardware acceleration, multi-domain orchestration, and federated learning are discussed in the context of real-world 5G operations. Governance, regulatory compliance, and inter-operator trust are analyzed as essential socio-technical dimensions without which technical countermeasures remain incomplete. The study argues that secure RL-based slicing must be approached as a co-design challenge involving learning algorithms, infrastructure telemetry, and institutional safeguards, and it offers forward-looking perspectives on formal verification, zero-trust architectures, and certifiable robustness for next-generation autonomous network management.
References
1. Foukas, X., Patounas, G., Elmokashfi, A., & Marina, M. K. (2017). Network slicing in 5G: Survey and challenges. IEEE Communications Magazine, 55(5), 94–100.
2. Ordonez-Lucena, J., Ameigeiras, P., Lopez, D., Ramos-Munoz, J. J., Lorca, J., & Folgueira, J. (2017). Network slicing for 5G with SDN/NFV: Concepts, architectures, and challenges. IEEE Communications Magazine, 55(5), 80–87.
3. Mao, H., Alizadeh, M., Menache, I., & Kandula, S. (2016). Resource management with deep reinforcement learning. Proceedings of the 15th ACM Workshop on Hot Topics in Networks, 50–56.
4. Luong, N. C., Hoang, D. T., Gong, S., Niyato, D., Wang, P., Liang, Y. C., & Kim, D. I. (2019). Applications of deep reinforcement learning in communications and networking: A survey. IEEE Communications Surveys & Tutorials, 21(4), 3133–3174.
5. Ilahi, I., Usama, M., Qadir, J., Janbi, M., Al-Fuqaha, A., Hoang, D. T., & Niyato, D. (2022). Challenges and countermeasures for adversarial attacks on deep reinforcement learning. IEEE Transactions on Artificial Intelligence, 3(2), 160–175.
6. Khan, R., Kumar, P., Jayakody, D. N. K., & Liyanage, M. (2020). A survey on security and privacy of 5G technologies: Potential solutions, recent advances, and future directions. IEEE Communications Surveys & Tutorials, 22(1), 196–248.
7. Li, R., Zhao, Z., Sun, Q., Chih-Lin, I., Yang, C., Chen, X., Zhao, M., & Zhang, H. (2018). Deep reinforcement learning for resource management in network slicing. IEEE Access, 6, 74429–74441.
8. Sadeghi, A., & Liu, J. (2020). Adversarial attacks on deep learning based wireless signal classification. IEEE Transactions on Cognitive Communications and Networking, 6(2), 543–553.
9. Zhang, X., Ma, Y., Singla, A., & Zhu, X. (2020). Adaptive reward-poisoning attacks against reinforcement learning. Proceedings of the International Conference on Machine Learning (ICML), 11225–11234.
10. Pan, X., Wang, Z., He, W., Mundluru, N., & Zhu, J. (2020). Efficiently attacking deep reinforcement learning policies. Proceedings of the AAAI Conference on Artificial Intelligence, 34(04), 5557–5565.
11. Li, Q. (2026). QoS Assurance Mechanism for 5G Network Slicing Based on the Deep Reinforcement Learning PPO Algorithm. arXiv preprint arXiv:2605.03345.
12. Pinto, L., Davidson, J., Sukthankar, R., & Gupta, A. (2017). Robust adversarial reinforcement learning. Proceedings of the International Conference on Machine Learning (ICML), 2817–2826.
13. Tessler, C., Efroni, Y., & Mannor, S. (2019). Action robust reinforcement learning and applications in continuous control. Proceedings of the International Conference on Machine Learning (ICML), 6215–6224.
14. Yu, T., Shen, S., Yu, R., & Li, J. (2020). Multi-objective deep reinforcement learning for resource allocation in network slicing. IEEE Transactions on Vehicular Technology, 69(12), 15040–15053.
15. Cohen, J. M., Rosenfeld, E., & Kolter, J. Z. (2019). Certified adversarial robustness via randomized smoothing. Proceedings of the International Conference on Machine Learning (ICML), 1310–1320.
16. Ali, M., Nikanjam, A., & Znati, T. (2021). Blockchain-based secure network slicing orchestration. IEEE Transactions on Network and Service Management, 18(2), 2098–2112.
17. Rahman, A., & Chowdhury, N. M. M. K. (2020). Security-performance trade-off in network slicing: A game-theoretic approach. IEEE Transactions on Network and Service Management, 17(3), 1695–1709.
18. Chen, L., Ling, Q., & Sun, H. (2021). Meta-learning based resource allocation for dynamic network slicing. IEEE Transactions on Cognitive Communications and Networking, 7(3), 889–902.
19. Messaoudi, F., Bradai, A., & Ksentini, A. (2021). Multi-agent deep reinforcement learning for end-to-end network slicing orchestration. IEEE Transactions on Network and Service Management, 18(4), 4681–4694.
20. Xu, C., Wang, X., Li, F., & Hu, R. Q. (2022). Fairness-aware resource allocation for network slicing via constrained deep reinforcement learning. IEEE Internet of Things Journal, 9(4), 2924–2937.
21. Zanzi, L., Sciancalepore, V., Costa-Perez, X., & Samdanis, K. (2022). An AI-driven framework for network slice security management in 5G systems. IEEE Transactions on Network and Service Management, 19(1), 423–437.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Journal of Data Intelligence and AI Systems
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
