Logo PTI Logo FedCSIS

Proceedings of the 21st Conference on Computer Science and Intelligence Systems (FedCSIS)

Annals of Computer Science and Information Systems, Volume 47

Container-based Machine Learning-Driven Routing Framework for Adaptive Traffic Management in Real-World SDN Environments

, ,

DOI: http://dx.doi.org/10.15439/2026F6690

Citation: Mariem Bouguerra, ,

Full text

Abstract. Software-Defined Networking (SDN) decouples the control plane from the data plane, enabling centralized programmability and paving the way for more intelligent, adaptive traffic management strategies. This paper presents MLRouter, a machine learning-based routing framework that enhances path selection by jointly considering traffic classification outputs and real-time network conditions. Unlike the majority of existing approaches, which are validated exclusively in simulation environments, the proposed solution is implemented and evaluated on a realistic experimental SDN testbed comprising virtual machines (VMs) running under KVM/QEMU, Open vSwitch (OVS) instances, and the Faucet SDN controller, with observability provided by Gauge and Prometheus. The system is built incrementally across three phases: (i) a REST API-driven static routing layer that allows programmatic manipulation of Faucet routes without manual YAML editing; (ii) a classical dynamic routing layer based on OSPF implemented through FRRouting (FRR) running in Docker containers; and (iii) a Random Forest classifier that replaces OSPF path decisions by leveraging real-time telemetry. The ML module interacts with the Faucet controller through a RESTful API to apply routing updates dynamically, with an automatic fallback to OSPF in case of model failure or low prediction confidence. Experimental results demonstrate that MLRouter reduces packet loss by approximately 50--60\%, decreases end-to-end latency by roughly 30\%, and increases throughput by 20--25\% relative to OSPF, across varying traffic intensities and congestion scenarios. These gains confirm that integrating machine learning intelligence with SDN programmability constitutes a viable and effective pathway toward adaptive, efficient, and real-time routing in realistic networking environments.

References

  1. L. L. Peterson and B. S. Davie, Computer Networks: A Systems Approach. Elsevier, 2007.
  2. D. Kreutz, F. M. V. Ramos, P. E. Verissimo, C. E. Rothenberg, S. Azodolmolky, and S. Uhlig, “Software-defined networking: A comprehensive survey,” Proceedings of the IEEE, vol. 103, no. 1, pp. 14–76, 2015.
  3. B. A. A. Nunes, M. Mendonca, X.-N. Nguyen, K. Obraczka, and T. Turletti, “A survey of software-defined networking: Past, present, and future of programmable networks,” IEEE Communications Surveys & Tutorials, vol. 16, no. 3, pp. 1617–1634, 2014.
  4. V. Srivastava and R. S. Pandey, “Load balancing for software-defined network: A review,” International Journal of Computers and Applications, vol. 44, no. 8, pp. 746–759, 2022.
  5. J. Moy, “OSPF Version 2,” RFC 2328, Internet Engineering Task Force (IETF), Apr. 1998. [Online]. Available: https://datatracker.ietf.org/doc/ html/rfc2328
  6. S. Faezi and A. Shirmarz, “A comprehensive survey on machine learning usage in software-defined networks (SDN),” Human-Centric Intelligent Systems, vol. 3, no. 3, pp. 312–343, 2023.
  7. M. R. Parsaei, M. J. Sobouti, R. Javidan et al., “Network traffic classification using machine learning techniques over software-defined networks,” International Journal of Advanced Computer Science and Applications, vol. 8, no. 7, 2017.
  8. S. Zhang, Y. Shen, M. Herlich et al., “Ryuo: Using a high-level northbound API for control messages in software-defined network,” in Proc. 17th Asia-Pacific Network Operations and Management Symposium (APNOMS), 2015, pp. 115–120.
  9. G. Pereira and P. Sousa, “A routing approach in SDN environments,” in Proc. 17th Iberian Conference on Information Systems and Technologies (CISTI), 2022, pp. 1–6.
  10. M. P. K. Perera Jayasuriya, K. Piamrat, and S. Hamma, “Network traffic classification using machine learning for software-defined networks,” in Machine Learning for Networking (MLN 2019), Springer, 2020, pp. 28– 39.
  11. M. Bouguerra, S. Khalfallah, and H. Ltifi, “MLRouter: A learning-based traffic classification within congestion-aware link-state routing protocol in SDN,” in Proc. International Conference on Innovation in Networking and Intelligent Systems (INISTA), 2025.
  12. C. Güemes-Palau, M. Ferriol-Galmés, J. Paillisse-Vilanova et al., “RouteNet-Gauss: Hardware-enhanced network modeling with machine learning,” IEEE Transactions on Networking, 2026.
  13. K. Danilchenko, N. Lazmi, and M. Segal, “Online learning framework for radio link failure prediction in FANETs,” in Proceedings of the 18th Conference on Computer Science and Intelligence Systems (FedCSIS), Warsaw, Poland, 2023, ser. Annals of Computer Science and Information Systems, vol. 35, pp. 41–48.
  14. M. Maliha, G. Habibi, and M. Atiquzzaman, “A survey on congestion control and scheduling for multipath TCP: Machine learning vs classical approaches,” in Proceedings of the 18th Conference on Computer Science and Intelligence Systems (FedCSIS), Warsaw, Poland, 2023, ser. Annals of Computer Science and Information Systems, vol. 35, pp. 49–61.