Optimizing Edge Topologies: A Review of Predictive Network Routing Algorithms
Authors
Kirill Andreevich Melnikov
Share
Annotation
The literature on machine-learning routing rests on an assumption that is rarely stated and almost never tested: network topology is treated as a fixed input over which a routing model operates. At the edge, where reconfiguration of links and nodes is the ordinary operating condition, this assumption does not hold. Packet-level evaluation of reinforcement-learning routing policies indicates that models trained under simplified network dynamics degrade once deployed against realistic traffic, and the degradation traces to the training regime. Drawing on fault-tolerance research from multi-agent coordination architectures outside the networking literature, this review separates topology-level disruptions, which require renegotiation across subgraphs and carry lasting downstream consequences, from resource-level disruptions, which recover within a fixed and comparatively short interval. The same body of research documents a general tendency of learned policies toward overconfidence under distributional shift, of which the routing degradation examined here is a specific instance. The review concludes that the outstanding problem is one of remedy transfer: whether a predicate-based safety architecture, proposed for this overconfidence problem in bounded robotic settings, can be constructed for network topologies whose state space is open-ended and subject to continuous reconfiguration.
Keywords
Authors
Kirill Andreevich Melnikov
Share
References:
Aktas, F., Shayea, I., Ergen, M., Saoud, B., Yahya, A. E., & Aldasheva, L. (2025). AI-enabled routing in next generation networks: A survey. Alexandria Engineering Journal, 120, 449–474. https://doi.org/10.1016/j.aej.2025.01.095
Almasan, P., Suárez-Varela, J., Wu, B., Xiao, S., Barlet-Ros, P., & Cabellos-Aparicio, A. (2021). Towards real-time routing optimization with deep reinforcement learning: Open challenges. In 2021 IEEE 22nd International Conference on High Performance Switching and Routing (HPSR) (pp. 1–6). IEEE. https://doi.org/10.1109/HPSR52026.2021.9481864
Bhavanasi, S. S., Pappone, L., & Esposito, F. (2023). Dealing with changes: Resilient routing via graph neural networks and multi-agent deep reinforcement learning. IEEE Transactions on Network and Service Management, 20(3), 2283–2294. https://doi.org/10.1109/TNSM.2023.3287936
Bogle, J., Bhatia, N., Ghobadi, M., Menache, I., Bjørner, N., Valadarsky, A., & Schapira, M. (2019). TeaVaR: Striking the right utilization-availability balance in WAN traffic engineering. In Proceedings of the ACM Special Interest Group on Data Communication (SIGCOMM ’19) (pp. 29–43). Association for Computing Machinery. https://doi.org/10.1145/3341302.3342069
Boltres, A., Freymuth, N., Jahnke, P., Karl, H., & Neumann, G. (2024). Learning sub-second routing optimization in computer networks requires packet-level dynamics. Transactions on Machine Learning Research.
Cao, L., Huo, T., Li, S., Zhang, X., Chen, Y., Lin, G., Wu, F., Ling, Y., Zhou, Y., & Xie, Q. (2024). Cost optimization in edge computing: A survey. Artificial Intelligence Review, 57(11). https://doi.org/10.1007/s10462-024-10947-4
He, Q., Wang, Y., Wang, X., Xu, W., Li, F., Yang, K., & Ma, L. (2023). Routing optimization with deep reinforcement learning in knowledge defined networking. IEEE Transactions on Mobile Computing, 23(2), 1444–1455. https://doi.org/10.1109/TMC.2023.3235446
Jiang, W., Han, H., Zhang, Y., Wang, J., He, M., Gu, W., Mu, J., & Cheng, X. (2024). Graph neural networks for routing optimization: Challenges and opportunities. Sustainability, 16(21), 9239. https://doi.org/10.3390/su16219239
Kim, G., Kim, Y., & Lim, H. (2022). Deep reinforcement learning-based routing on software-defined networks. IEEE Access, 10, 18121–18133. https://doi.org/10.1109/ACCESS.2022.3151081
Li, T., Zhu, K., Luong, N. C., Niyato, D., Wu, Q., Zhang, Y., & Chen, B. (2022). Applications of multi-agent reinforcement learning in future internet: A comprehensive survey. IEEE Communications Surveys & Tutorials, 24(2), 1240–1279.
Mikhailiuk, M. (2026). Predictive technologies and autonomous decision making. LAP LAMBERT Academic Publishing.
Pooyandeh, M., & Sohn, I. (2021). Edge network optimization based on AI techniques: A survey. Electronics, 10(22), 2830. https://doi.org/10.3390/electronics10222830
Rusek, K., Suárez-Varela, J., Almasan, P., Barlet-Ros, P., & Cabellos-Aparicio, A. (2020). RouteNet: Leveraging graph neural networks for network modeling and optimization in SDN. IEEE Journal on Selected Areas in Communications, 38(10), 2260–2270.
Suárez-Varela, J., Almasan, P., Ferriol-Galmés, M., Rusek, K., Geyer, F., Cheng, X., Shi, X., Xiao, S., Scarselli, F., Cabellos-Aparicio, A., & Barlet-Ros, P. (2022). Graph neural networks for communication networks: Context, use cases and opportunities. IEEE Network, 37(3), 146–153. https://doi.org/10.1109/MNET.001.2200269
Wu, J., Dong, F., Leung, H., Zhu, Z., Zhou, J., & Drew, S. (2024). Topology-aware federated learning in edge computing: A comprehensive survey. ACM Computing Surveys, 56(10), 1–41. https://doi.org/10.1145/3659205
Xiao, Y., Liu, J., Wu, J., & Ansari, N. (2021). Leveraging deep reinforcement learning for traffic engineering: A survey. IEEE Communications Surveys & Tutorials, 23(4), 2064–2097. https://doi.org/10.1109/COMST.2021.3102580
Xu, Z., Yan, F. Y., Singh, R., Chiu, J. T., Rush, A. M., & Yu, M. (2023). Teal: Learning-accelerated optimization of WAN traffic engineering. In Proceedings of the ACM SIGCOMM 2023 Conference (pp. 378–393). Association for Computing Machinery.
Yang, S., Tan, C., Madsen, D. Ø., Xiang, H., Li, Y., Khan, I., & Choi, B. J. (2022). Comparative analysis of routing schemes based on machine learning. Mobile Information Systems, 2022, Article 4560072. https://doi.org/10.1155/2022/4560072
Other articles of the issue
