Deep Reinforcement Learning-Driven Dynamic Resource Allocation Framework for QoS-Aware 5G Network Slicing
Keywords:
5G network slicing, quality of service, deep reinforcement learning, resource allocation, orchestration, governance, fairnessAbstract
The proliferation of diverse 5G services with stringent and heterogeneous quality of service (QoS) requirements has elevated network slicing to a foundational architectural principle. Realizing the full potential of network slicing demands dynamic, automated resource allocation mechanisms that can adapt to rapidly fluctuating traffic and service-level demands across shared physical infrastructures. This paper presents a comprehensive framework for QoS-aware 5G network slicing driven by deep reinforcement learning, specifically leveraging the proximity policy optimization (PPO) algorithm as a central decision-making engine. Departing from conventional optimization or heuristic-based approaches, the framework treats slice resource allocation as a sequential decision-making problem under uncertainty, enabling continuous learning and adaptation in real time. The discussion is oriented toward system-level analysis, examining the architectural decomposition of the framework into monitoring, policy, and orchestration components, and evaluating structural trade-offs among centralization, scalability, and response latency. Critical dimensions such as multi-tenancy governance, sustainability through energy-aware policy shaping, robustness against non-stationary network conditions, and fairness among competing slices are analyzed in depth. Particular attention is given to deployment pragmatics, including the integration with existing software-defined networking and network function virtualization control planes, the role of digital twins for pre-deployment policy validation, and the challenge of maintaining slice isolation under dynamic resource fluctuations. Policy implications are explored with respect to regulatory frameworks for infrastructure sharing and net neutrality, arguing that intelligent resource brokers must be designed to balance commercial incentives with equitable access. The analysis draws on cross-domain insights from cloud orchestration, autonomous systems, and industrial control, positioning the proposed framework as a pathway toward self-optimizing, resilient, and ethically aligned 5G networks. By synthesizing advances in deep reinforcement learning with architectural and governance considerations, the paper contributes a holistic perspective on future-ready slice orchestration.
References
1. Foukas, X., Patounas, E., Elmokashfi, A., & Marina, M. K. (2017). Network slicing in 5G: Survey and challenges. IEEE Communications Surveys & Tutorials, 19(2), 791–822.
2. Shafi, M., Molisch, A. F., Smith, P. J., Haustein, T., Zhu, P., De Silva, P., … & Wunder, G. (2017). 5G: A tutorial overview of standards, trials, challenges, deployment, and practice. IEEE Journal on Selected Areas in Communications, 35(6), 1201–1221.
3. 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.
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. Mao, Q., Hu, F., & Hao, Q. (2018). Deep learning for intelligent wireless networks: A comprehensive survey. IEEE Communications Surveys & Tutorials, 20(4), 2595–2621.
6. Schulman, J., Wolski, F., Dhariwal, P., Radford, A., & Klimov, O. (2017). Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347.
7. Taleb, T., Mada, B., Corradi, A., & Flinck, H. (2017). PERMIT: Network slicing for personalized 5G mobile telecommunications. IEEE Communications Magazine, 55(5), 88–93.
8. Sallent, O., Perez-Romero, J., Ferrus, R., & Agusti, R. (2019). On radio access network slicing from a radio resource management perspective. IEEE Communications Magazine, 57(5), 45–51.
9. Xiang, H., Zhou, W., Daneshmand, M., & Yu, P. (2020). Deep reinforcement learning based resource allocation in heterogeneous network slicing for 5G. IEEE Transactions on Cognitive Communications and Networking, 6(4), 1225–1236.
10. Li, R., Zhao, Z., Sun, Q., I, C. L., Yang, C., Chen, X., … & Zhang, H. (2018). Deep reinforcement learning for resource management in network slicing. IEEE Access, 6, 74464–74476.
11. Sun, W., Li, S., Zhang, N., & Liu, Y. (2020). PPO-based online learning for dynamic resource allocation in 5G network slicing. IEEE Internet of Things Journal, 7(10), 973–988.
12. Li, Q. (2026). QoS Assurance Mechanism for 5G Network Slicing Based on the Deep Reinforcement Learning PPO Algorithm. arXiv preprint arXiv:2605.03345.
13. Chen, M., Saad, W., & Yin, C. (2019). Deep reinforcement learning for fairness in resource allocation of network slicing. IEEE Transactions on Cognitive Communications and Networking, 5(2), 291–303.
14. Mhaisen, N., Marfoq, O., & Takrouni, A. (2020). Federated reinforcement learning for resource management in sliced networks. IEEE Access, 8, 133240–133253.
15. Tessler, C., Mankowitz, D. J., & Mannor, S. (2019). Reward constrained policy optimization. In International Conference on Learning Representations (ICLR).
16. Ksentini, A., & Nikaein, N. (2017). Toward enforcing network slicing on RAN: Flexibility and resources abstraction. IEEE Communications Magazine, 55(6), 102–108.
17. Xiao, L., Li, Y., Liu, J., & Du, X. (2020). Deep reinforcement learning for energy-efficient 5G network slicing. IEEE Internet of Things Journal, 7(7), 6322–6333.
18. Pinto, L., Davidson, J., Sukthankar, R., & Gupta, A. (2017). Robust adversarial reinforcement learning. In International Conference on Machine Learning (pp. 2817–2826).
19. Mayer, J. (2019). The role of transparency in autonomous network management under EU net neutrality rules. Telecommunications Policy, 43(8), 101–112.
20. Martini, B., Gharbaoui, M., Castoldi, P., & Valcarenghi, L. (2020). Programmable and adaptable service function chain embedding with deep reinforcement learning and integer linear programming. IEEE Transactions on Network and Service Management, 17(2), 749–762.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 International Journal of Artificial Intelligence Engineering and 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.
