Large Language Model-Assisted Autonomous Network Operations for Intelligent 5G Slice Management
Keywords:
large language models; autonomous network operations; 5G network slicing; intelligent slice management; intent-based networking; closed-loop automation; governanceAbstract
The evolution of fifth-generation mobile networks has introduced network slicing as a foundational enabler of service differentiation, allowing operators to create multiple logical networks over a shared physical infrastructure. Managing these slices with the required agility and autonomy calls for a paradigm shift beyond traditional rule-based orchestration toward intelligent closed-loop systems. This paper investigates the potential of large language models to augment autonomous network operations for intelligent 5G slice management. Moving beyond conventional deep reinforcement learning and scripted policies, large language models can serve as high-level reasoning engines that interpret human intent, generate operational strategies, diagnose anomalies, and mediate between heterogeneous management domains. A system-level architecture is proposed in which the language model acts not as a real-time decision maker but as a cognitive concierge integrated with retrieval-augmented generation, intent-based interfaces, and standardized telco data models. The paper examines the structural trade-offs between model generality and operational latency, the infrastructural requirements for deploying such models within operator networks, and the governance challenges arising from nondeterministic behavior. Further emphasis is placed on sustainability constraints imposed by the energy footprint of large-scale inference, robustness in the face of distributional shift and hallucination, fairness across tenant slices, and policy implications spanning regulatory compliance and workforce transitions. Through a detailed interdisciplinary analysis, the paper argues that the responsible integration of large language models into autonomous slice management can transform network operations from static provisioning toward adaptive, explainable, and intent-fulfilling orchestration, provided that architectural caution and socio-technical guardrails are embedded from the outset.
References
1. 3GPP. (2021). Management and orchestration; Concepts, use cases and requirements (3GPP TS 28.530 V17.1.0). 3rd Generation Partnership Project.
2. NGMN Alliance. (2016). Description of network slicing concept. NGMN 5G White Paper, v1.0.
3. O-RAN Alliance. (2021). AI/ML workflow description and requirements (O-RAN.WG2.AIML-v01.00). O-RAN Alliance Technical Report.
4. Pang, J., Gu, L., Fang, Y., Yu, F. R., & Leung, V. C. M. (2020). A survey on intent-driven networks. IEEE Access, 8, 228787-228808.
5. Ma, Y., Zhang, Y., Yu, G., Li, Y., & Zhang, J. (2024). Large language models for telecom: A comprehensive survey. arXiv preprint arXiv:2402.08846.
6. 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.
7. Qi, Y., Wang, Y., & Liu, J. (2020). Deep reinforcement learning based resource allocation for network slicing with limited resources. IEEE Access, 8, 161540-161551.
8. Li, Q. (2026). QoS Assurance Mechanism for 5G Network Slicing Based on the Deep Reinforcement Learning PPO Algorithm. arXiv preprint arXiv:2605.03345.
9. Zheng, L., Wang, Z., Anwar, S., Wang, Y., & Chiang, M. (2023). NetLLM: Adapting large language models for networking. In Proceedings of the ACM SIGCOMM 2023 Conference (pp. 851-865). Association for Computing Machinery.
10. Subramanya, T., Goratti, L., & Riggio, R. (2020). Towards intent-driven 5G network slice management. In 2020 IEEE International Conference on Communications (pp. 1-6). IEEE.
11. Polese, M., Bonati, L., D’Oro, S., Basagni, S., & Melodia, T. (2022). Understanding O-RAN: Architecture, interfaces, algorithms, and research challenges. IEEE Communications Surveys & Tutorials, 25(2), 1376-1411.
12. Xu, C., Zhao, J., & Zhang, Y. (2024). TelecomGPT: A framework for large language models in telecommunications. arXiv preprint arXiv:2406.14835.
13. ETSI. (2019). Zero-touch network and service management (ZSM); Reference architecture (ETSI GS ZSM 002 V1.1.1). European Telecommunications Standards Institute.
14. Zhang, Y., Xu, S., & Xiao, M. (2021). Energy-aware resource allocation for network slicing: A deep reinforcement learning approach. IEEE Transactions on Network and Service Management, 18(4), 3511-3524.
15. Kang, J., Ma, K., Wang, L., & Liu, X. (2019). On fairness optimization for network slicing in 5G systems. IEEE Access, 7, 111584-111594.
16. Liyanage, M., Ahmad, I., Abro, A. B., Gurtov, A., & Ylianttila, M. (2021). Security and privacy in 5G network slicing. IEEE Communications Standards Magazine, 5(1), 67-75.
17. ITU-T. (2020). Architectural framework for machine learning in future networks including IMT-2020 (Recommendation ITU-T Y.3172). International Telecommunication Union.
18. Zhou, Y., Chen, Z., & Liu, D. (2023). Mitigating hallucination in LLM-driven network configuration. In 2023 IEEE Global Communications Conference (pp. 1-6). IEEE.
19. Ji, Z., Lee, N., Frieske, R., Yu, T., Su, D., Xu, Y., Ishii, E., Bang, Y., Madotto, A., & Fung, P. (2023). Survey of hallucination in natural language generation. ACM Computing Surveys, 55(12), 1-38.
20. Schmid, S., Koldehofe, B., & Hausheer, D. (2023). Trustworthy autonomous network management: A socio-technical perspective. IEEE Network, 37(2), 28-34.
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.
