Journal of AI-Driven Communication Engineering

Special Issue: Artificial Intelligence Across the Communication Stack: Engineering, Human Interaction, and Governance in the 6G Era

Vol. 2 (2026)

Large Language Models as Cognitive Network Agents: Intent-Based Management and Self-Optimization in 6G Communication Systems

DOI:
https://doi.org/10.31875/2979-1081.2026.02.04
Submitted
July 4, 2026
Published
2026-07-04

Abstract

The management of sixth-generation (6G) communication networks demands a level of autonomy, contextual reasoning, and cross-layer decision-making that exceeds the capacity of conventional rule-based network management systems. This paper positions large language models (LLMs) not merely as text-generating systems or benchmark-oriented assistants, but as supervisory cognitive agents within AI-driven communication engineering. The core contribution is a system-level analysis of how LLM agents can translate operator intent, telemetry, alarms, and standards knowledge into verifiable network-control actions affecting network slices, QoS profiles, RAN Intelligent Controller (RIC) xApps/rApps, scheduling policies, and fault-management workflows. The paper reviews telecom-specific LLM adaptation methods, intent-based networking, zero-touch configuration, resource management, and security risks, while evaluating practical relevance through communication engineering metrics such as latency, throughput, spectral efficiency, SLA violation rate, control overhead, and closed-loop stability. Deployment constraints are discussed with explicit attention to O-RAN and 6G control-loop timescales, showing that near-term feasibility lies in management-plane and policy-supervision roles rather than direct sub-millisecond PHY/MAC control.

References

  1. ITU-R: IMT-2030 (6G) Framework Recommendation. ITU-R M.2160; 2023.
  2. Dang S, Amin O, Shihada B, Alouini MS: What should 6G be? Nature Electronics 2020, 3: 20-29. https://doi.org/10.1038/s41928-019-0355-6
  3. ETSI: Zero-Touch Network and Service Management (ZSM): Reference Architecture. ETSI GS ZSM 002; 2019.
  4. Mwanje SS, Decarreau G, Mannweiler C, Naseer ul Islam M, Schotten HD: Towards cognitive autonomous networks. IEEE Network 2021, 35(6): 248-256.
  5. Brown TB, Mann B, Ryder N, et al.: Language models are few-shot learners. Advances in Neural Information Processing Systems 2020, 33: 1877-1901.
  6. Zhou S, Gan L, Bengio S, Schiele B, Leskovec J: A survey of large language models for network management. IEEE Communications Surveys and Tutorials 2024, 26(3): 1894-1938.
  7. Hallucination survey in telecom LLMs. IEEE Open Journal of the Communications Society 2024 [in press].
  8. Vaswani A, Shazeer N, Parmar N, et al.: Attention is all you need. Advances in Neural Information Processing Systems 2017, 30: 5998-6008.
  9. Jiang A, Sablayrolles A, Mensch A, et al.: Mistral 7B. arXiv preprint arXiv:2310.06825; 2023.
  10. Kim S, Hooper C, Wattanawong T, et al.: Full stack optimization of transformer inference. In: MLSys 2023.
  11. Lewis P, Perez E, Piktus A, et al.: Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in Neural Information Processing Systems 2020, 33: 9459-9474.
  12. Zou Y, Zhao Q, Li N, et al.: TelecomBench: A benchmark for evaluating LLMs on telecommunications. arXiv:2406.08932; 2024.
  13. Hu EJ, Shen Y, Wallis P, et al.: LoRA: Low-rank adaptation of large language models. ICLR 2022.
  14. Jiang X, Wang H, Luo Y, et al.: TelecomGPT: A framework to build telecom-specific large language models. arXiv: 2407.09424; 2024.
  15. Zhu Y, Ouyang W, Shen J, et al.: MobileLLM: Optimizing sub-billion parameter language models for on-device use cases. arXiv: 2402.14905; 2024.
  16. Hinton G, Vinyals O, Dean J: Distilling the knowledge in a neural network. arXiv: 1503.02531; 2015.
  17. Leivadeas A, Falkner M: A survey on intent-based networking. IEEE Communications Surveys and Tutorials 2023, 25(4): 2215-2264. https://doi.org/10.1109/COMST.2022.3215919
  18. Wang Z, Ye X, Luo H, et al.: NetLLM: Adapting large language models for networking. In: ACM SIGCOMM 2024.
  19. Roy S, Saha D, Mitra R: Large language model for autonomous fault management in 5G networks. IEEE Transactions on Network and Service Management 2024, 21(5): 4821-4835.
  20. Giordani M, Polese M, Mezzavilla M, Rangan S, Zorzi M: Toward 6G networks: Use cases and technologies. IEEE Communications Magazine 2020, 58(3): 55-61. https://doi.org/10.1109/MCOM.001.1900411
  21. Wei J, Wang X, Schuurmans D, et al.: Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems 2022, 35: 24824-24837. https://doi.org/10.52202/068431-1800
  22. Zhang Y, Li K, Liu Y, et al.: LLM-based dynamic spectrum management for 6G heterogeneous networks. IEEE Journal on Selected Areas in Communications 2025, 43(2): 412-428.
  23. Chen X, Zhu H, Lian Z, et al.: Retrieval-augmented LLM agent for O-RAN slice management. IEEE Wireless Communications Letters 2024, 13(8): 2195-2199.
  24. Perez F, Ribeiro I: Ignore previous prompt: Attack techniques for language models. arXiv: 2211.09527; 2022.
  25. Liu Y, Deng G, Li Y, et al.: Prompt injection attack against LLM-integrated applications. arXiv: 2306.05499; 2023.
  26. Zou A, Wang Z, Kolter JZ, Fredrikson M: Universal and transferable adversarial attacks on aligned language models. arXiv: 2307.15043; 2023.
  27. Xue S, Huang C, Shi J, et al.: Byzantine-robust RAG for network management. In: IEEE ICC 2025 [in press].
  28. Carlini N, Liu C, Erlingsson U, Kos J, Song D: The secret sharer: Evaluating and testing unintended memorization in neural networks. USENIX Security 2019.
  29. Yu D, Naik S, Backurs A, et al.: Differentially private fine-tuning of language models. ICLR 2022.
  30. Kaur K, Bhatt U, Weller A: Trustworthy explainability of network management AI. IEEE Network 2024, 38(1): 112-119.
  31. European Parliament: Artificial Intelligence Act. Regulation (EU) 2024/1689; 2024.
  32. Shen X, Guo J, Li A, et al.: Physics-informed LLM fine-tuning for wireless channel configuration. IEEE Communications Letters 2025, 29(4): 789-793.
  33. Park J, Samarakoon S, Bennis M, Debbah M: Wireless network intelligence at the edge. Proceedings of the IEEE 2019, 107(11): 2204-2239. https://doi.org/10.1109/JPROC.2019.2941458
  34. McCloskey M, Cohen NJ: Catastrophic interference in connectionist networks. Psychology of Learning and Motivation 1989, 24: 109-165. https://doi.org/10.1016/S0079-7421(08)60536-8
  35. Wang L, Zhang X, Su H, Zhu J: Comprehensive study of catastrophic forgetting in large language models. arXiv: 2402.18240; 2024.