A Self-Adaptive Zero-Trust Data Communication Framework Using Deep Reinforcement Learning in Software-Defined Networks

Osita Miracle Nwakeze1, Nwafor Anthony2, Obaze Caleb3
1 2 Department of Computer Science, Chukwuemeka Odumegwu Ojukwu University, Uli 3 Department of Computer science, Dennis Osadebay University Asaba, Delta State, Nigeria.

View / Download Full Article (PDF)

Abstract

Software-Defined Networking (SDN) is quickly expanding, which provides flexibility in managing the network but opens systems to highly dynamic and advanced cyber threats. Conventional defence systems such as rule-based firewalls and threshold-based intrusion detection systems can be inadequate because they are not adaptable and are also static. In this work, a self-adaptive zero-trust data communication framework is suggested and combines Deep Reinforcement Learning (DRL) to enforce security in SDN environments in an intelligent and real-time manner. The framework includes a zero-trust engine and a DRL agent that is implemented in the SDN controller to assess the level of trust constantly and implement access control policies dynamically.

Keywords

Software-Defined Networking (SDN), Zero Trust Architecture, Deep Reinforcement Learning (DRL), Network Security, Adaptive Access Control.

References

[1] Barach, J. (2025). Towards zero trust security in SDN: A multi-layered defence strategy. In Proceedings of the 26th International Conference on Distributed Computing and Networking (pp. 331–339). Association for Computing Machinery. https://doi.org/10.1145/3700838.3703671

[2] Bijlwan, V. P., & Kumar, N. (2026). Deep reinforcement learning in software defined networking: A survey, research challenges, and future perspectives. IEEE Communications Surveys & Tutorials, 28, 4624–4653.

[3] Chenghao, S., Wang, Y., Su, Y., & Wan, Y. (2022). Adaptive threat mitigation in SDN using improved D3QN. In Second International Conference on Cloud Computing and Mechatronic Engineering (I3CME 2022) (Proc. SPIE 12339, 1233911). https://doi.org/10.1117/12.2652679

[4] Eke, R. N., & Onyejegbu, L. N. (2024). A deep reinforcement learning deep Q network-based approach for systematic detection of distributed denial of service attacks in software-defined networking. SSRN. https://ssrn.com/abstract=5023193 or http://dx.doi.org/10.2139/ssrn.5023193

[5] Kalekar, V. (2026). Uncertainty-aware reinforcement learning for zero trust: An empirical evaluation of PPO, MC dropout, and Bayesian neural networks. In Proceedings of the 12th International Conference on Information Systems Security and Privacy (ICISSP 2026) (Vol. 1, pp. 505–514). https://doi.org/10.5220/0014245800004061

[6] Kanimozhi, R., & Ramesh, P. S. (2025). Deep reinforcement learning-based intrusion detection scheme for software-defined networking. Scientific Reports, 15, 38827. https://doi.org/10.1038/s41598-025-24869-w

[7] Ke, W., Li, Z., Chen, P., Chen, B., & Lv, J. et al. (2025). DRL-AMIR: Intelligent flow scheduling for software-defined zero trust networks. Computers, Materials & Continua, 84(2), 3305–3319. https://doi.org/10.32604/cmc.2025.065665

[8] Khan, R. I., Habib, M. Q. B., Hasan, M. M., & Shad, K. W. U. (2026). Zero trust architecture in cloud-native environments: A scalable framework for cybersecurity. International Journal of Science and Research Archive, 18(1), 296–305. https://doi.org/10.30574/ijsra.2026.18.1.0063

[9] Mahida, A. (2026). An intellectual zero trust security framework using deep reinforcement learning for predictive threat mitigation in AI-based fraud detection systems. IEEE Access, 14, 24602–24617.

[10] Muhammad, Z. (2026). Zero trust architecture (ZTA) design and implementation: A comprehensive review. American Journal of Innovation in Science and Engineering, 5(1), 18–25. https://doi.org/10.54536/ajise.v5i1.5905

[11] Niknami, N., & Wu, J. (2023). DeepIDPS: An adaptive DRL-based intrusion detection and prevention system for SDN. Center for Networked Computing, Temple University, USA.

[12] Nwakeze, O. M., & Mohammed, N. U. (2025). Development of a deep learning-based framework for security of IoT network protocol. International Journal of Innovative and Applied Research (IJIAR), 7(10), 1–12. International Journal of Innovative and Applied Research. https://doi.org/10.58538/IJIAR/2136

[13] Rose, S., Borchert, O., Mitchell, S., & Connelly, S. (2020). Zero trust architecture (NIST Special Publication 800-207). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.SP.800-207

[14] Umerah, A. T., Azaka, M., Nwakeze, O. M., Obaze, C. A., & Ibeh, S. C. (2025). AI-driven diagnostic imaging: Hybrid CNN-GNN models for early detection of cancer from pathological images. International Journal of Latest Technology in Engineering, Management & Applied Science (IJLTEMAS), 14(9), 579–588. International Journal of Latest Technology in Engineering, Management & Applied Science. https://EconPapers.repec.org/RePEc:bjb:journl:v:14:y:2025:i:9:p:579-588