Multi-Agent Reinforcement Learning for Cross-Slice Resource Competition and Service-Level Agreement Optimization
Keywords:
network slicing, multi-agent reinforcement learning, service-level agreement, resource allocation, fairness, orchestration, 5G, decentralized coordinationAbstract
Network slicing enables operators to partition a shared physical infrastructure into multiple isolated logical networks, each tailored to distinct service requirements. As the number and heterogeneity of slices increase, guaranteeing service-level agreements under dynamic demand and resource scarcity becomes a challenging cross-slice coordination problem. This paper presents a system-level investigation of multi-agent reinforcement learning as a governance, orchestration, and optimization framework for cross-slice resource competition. Departing from single-agent formulations, we conceptualize each slice as an autonomous learning agent that must negotiate compute, radio, and transport resources while jointly preserving end-to-end latency, throughput, and reliability commitments. The paper elaborates on architectural configurations, ranging from fully decentralized independent learners to centralized training with decentralized execution, and examines the inherent trade-offs between policy coherence, scalability, and information exchange overhead. A central concern is the structural tension between competing operational objectives: maximizing resource utilization, maintaining slice isolation, and ensuring inter-slice fairness. We analyze how multi-agent learning can internalize fairness constraints and operator-defined priority policies through reward shaping, hierarchical critics, and coalitional coordination mechanisms. Governance implications are discussed, including the role of the infrastructure provider as a meta-regulator, auditability of learned allocation policies, and integration with open radio access network architectures. The paper further explores deployment robustness in non-stationary traffic environments, the sustainability dimension via energy-aware reward design, and policy transfer across network domains. By framing resource competition as a continual multi-agent negotiation process, we highlight how multi-agent reinforcement learning can transform slice orchestration from reactive threshold-based control toward anticipatory, policy-compliant infrastructure sharing. The analysis is supported by an extensive review of current literature and provides forward-looking perspectives on standardization and regulatory alignment.
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. Zhang, H., Liu, N., Chu, X., Long, K., Aghvami, A.-H., & Leung, V. C. M. (2017). Network slicing based 5G and future mobile networks: Mobility, resource management, and challenges. IEEE Communications Magazine, 55(8), 138–145.
3. 3GPP. (2020). Management and orchestration; Concepts, use cases and requirements (3GPP TS 28.530 version 16.2.0). 3rd Generation Partnership Project.
4. O-RAN Alliance. (2021). O-RAN architecture description (O-RAN.WG1.OAD-R003-v07.00). O-RAN Alliance.
5. Li, R., Zhao, Z., Sun, Q., I, C.-L., Yang, C., Chen, X., Zhao, M., & Zhang, H. (2018). Deep reinforcement learning for resource management in network slicing. IEEE Access, 6, 74429–74441.
6. Liu, Y., Ding, J., & Liu, X. (2020). A constrained reinforcement learning based approach for network slicing. IEEE Transactions on Network and Service Management, 17(4), 2301–2314.
7. Van Le, D., & Tham, C.-K. (2019). A deep reinforcement learning based multi-resource management framework for 5G network slicing. IEEE Access, 7, 164664–164678.
8. Hua, Y., Li, R., Zhao, Z., Chen, X., & Zhang, H. (2020). GAN-powered deep distributional reinforcement learning for resource management in network slicing. IEEE Journal on Selected Areas in Communications, 38(2), 334–349.
9. Li, Q. (2026). QoS Assurance Mechanism for 5G Network Slicing Based on the Deep Reinforcement Learning PPO Algorithm. arXiv preprint arXiv:2605.03345.
10. Zhang, K., Yang, Z., & Başar, T. (2021). Multi-agent reinforcement learning: A selective overview of theories and algorithms. Studies in Systems, Decision and Control, 325, 321–384.
11. Feriani, A., & Hossain, E. (2021). Single and multi-agent deep reinforcement learning for AI-enabled wireless networks: A tutorial. IEEE Communications Surveys & Tutorials, 23(2), 1226–1252.
12. Liang, L., Ye, H., & Li, G. Y. (2019). Spectrum sharing in vehicular networks based on multi-agent reinforcement learning. IEEE Journal on Selected Areas in Communications, 37(10), 2282–2292.
13. Zhang, Z., Yang, Y., Ma, M., & Zhao, Y. (2020). Multi-agent deep reinforcement learning based resource allocation for network slicing in 5G and beyond. IEEE Transactions on Network and Service Management, 17(4), 2342–2356.
14. Samdanis, K., Costa-Perez, X., & Sciancalepore, V. (2016). From network sharing to multi-tenancy: The 5G network slice broker. IEEE Communications Magazine, 54(7), 32–39.
15. Bega, D., Gramaglia, M., Banchs, A., Sciancalepore, V., Samdanis, K., & Costa-Perez, X. (2017). Optimising 5G infrastructure markets: The business of network slicing. IEEE International Conference on Computer Communications (INFOCOM), 1–9.
16. Zhou, Y., Fadlullah, Z. M., Mao, B., & Kato, N. (2018). A deep-learning-based radio resource assignment technique for 5G ultra dense networks. IEEE Network, 32(6), 28–34.
17. Foerster, J., Assael, I. A., de Freitas, N., & Whiteson, S. (2016). Learning to communicate with deep multi-agent reinforcement learning. Advances in Neural Information Processing Systems, 29, 2137–2145.
18. Rashid, T., Samvelyan, M., de Witt, C. S., Farquhar, G., Foerster, J., & Whiteson, S. (2018). QMIX: Monotonic value function factorisation for deep multi-agent reinforcement learning. International Conference on Machine Learning, 4295–4304.
19. Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction (2nd ed.). MIT Press.
20. Mao, H., Alizadeh, M., Menache, I., & Kandula, S. (2016). Resource management with deep reinforcement learning. ACM Workshop on Hot Topics in Networks, 50–56.
21. Sun, Y., Peng, M., Mao, S., Yan, S., & Wang, C. (2019). Autonomous resource slicing for virtualized wireless networks with deep reinforcement learning. IEEE Transactions on Vehicular Technology, 68(7), 7025–7039.
22. Chen, M., Challita, U., Saad, W., Yin, C., & Debbah, M. (2019). Artificial neural networks-based machine learning for wireless networks: A tutorial. IEEE Communications Surveys & Tutorials, 21(4), 3039–3071.
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.
