Towards Closed-Loop Management in Private 5G Networks with P4/eBPF Telemetry and Reinforcement Learning

• Vinícius S. Simão IFPB https://orcid.org/0009-0003-7334-9561
• Lucas V. Monteiro IFPB https://orcid.org/0009-0005-7127-8354
• Pedro H. Rodrigues Alves IFPB https://orcid.org/0009-0006-0177-5985
• Ruan D. Gomes IFPB https://orcid.org/0000-0003-4700-7843
• Leandro C. de Almeida IFPB https://orcid.org/0000-0003-4342-3030
• Paulo D. Maciel Jr. IFPB https://orcid.org/0000-0002-0732-752X

DOI: https://doi.org/10.5753/wgrs.2026.23155

Resumo

The growing demand for latency-critical services in private 5G deployments drives the need for autonomous, closed-loop network management that can continuously adapt to changing conditions. An important challenge is to obtain a unified view of both the transport-network behavior and the use of compute resources without incurring excessive overhead. In this context, we introduce a closed-loop design that integrates P4-based out-of-band network telemetry (ONT) with CPU KPIs obtained from the User Plane Functions (UPFs) using the Extended Berkeley Packet Filter (eBPF). The metrics obtained are used as input to a reinforcement learning (RL) agent that adjusts the priority of the switch queue and redirects traffic through different paths in the 5G network, dynamically selecting the UPF through the Network Exposure Function (NEF) API. In the scenario that combines UPF monitoring with RL-based traffic prioritization, a 46.48% reduction in latency and a 57.95% reduction in jitter were obtained, demonstrating the benefit of joint transport-and-core control.

Referências

Abranches, M., Michel, O., Keller, E., and Schmid, S. (2021). Efficient Network Monitoring Applications in the Kernel with eBPF and XDP. In 2021 IEEE Conference on Network Function Virtualization and Software Defined Networks (NFV-SDN), pages 28–34.

Eswaran, S. and Honnavalli, P. (2022). Private 5G networks: a survey on enabling technologies, deployment models, use cases and research directions: Private 5G networks: a survey on enabling technologies, deployment models, use cases and research directions. Telecommun. Syst., 82(1):3–26.

ETSI (2026). 5G; System architecture for the 5G System (5GS). Technical Specification ETSI TS 123 501, European Telecommunications Standards Institute (ETSI). 3GPP TS 23.501 version 19.6.0 Release 19. Accessed: 2026-01-31.

Goshi, E., Jarschel, M., Pries, R., He, M., and Kellerer, W. (2021). Investigating Inter-NF Dependencies in Cloud-Native 5G Core Networks. In 2021 17th International Conference on Network and Service Management, pages 370–374.

Hoa, L. C., Dang, T. C., and Vo, V. M. N. (2025). An integrated model of threshold-based scaling and fractional admission controlling to improve resource utilization efficiency in 5G core networks. PLOS ONE, 20(8):1–21.

Hyun, J., Van Tu, N., and Hong, J. W.-K. (2018). Towards knowledge-defined networking using in-band network telemetry. In NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Symposium, pages 1–7.

Pizzutti, M. and Schaeffer-Filho, A. E. (2019). Adaptive Multipath Routing based on Hybrid Data and Control Plane Operation. In IEEE Conference on Computer Communications, pages 730–738.

Puttlitz, C., Parizotto, R., and Schaeffer-Filho, A. (2024). P4NetIntel: End-to-End Network Telemetry with eBPF and XDP. In 2024 IEEE Conference on Network Function Virtualization and Software Defined Networks, pages 1–6.

Puttlitz, C. and Schaeffer-Filho, A. (2025). P4eBalancer: Leveraging P4 and eBPF for Optimized Load Balancing with Network and Host Insights. In Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos (SBRC), pages 392–405. SBC.

Simão, V. S., Monteiro, L. V., Gomes, R. D., de Almeida, L. C., and Maciel, P. D. (2025). Closed-Loop Network Control for Industrial Edge Computing: A Telemetry-Driven and AI-Based Approach for Latency-Critical Applications in Private 5G. IEEE Networking Letters, pages 1–1.

Tajbakhsh, H., Parizotto, R., Schaeffer-Filho, A., and Haque, I. (2026). Reinforcement Learning-Based In-Network Load Balancing. IEEE Transactions on Network and Service Management, 23:1100–1111.

Wu, D., Li, J., Ferini, A., Xu, Y. T., Jenkin, M., Jang, S., Liu, X., and Dudek, G. (2023). Reinforcement learning for communication load balancing: approaches and challenges. Frontiers in Computer Science, Volume 5 - 2023.

Zheng, C., Rienecker, B., and Zilberman, N. (2023). QCMP: Load Balancing via In-Network Reinforcement Learning. In Proceedings of the 2nd ACM SIGCOMM Workshop on Future of Internet Routing & Addressing, FIRA ’23, page 35–40, New York, NY, USA. Association for Computing Machinery.