Neural Network-Based Performance Prediction for High-Capacity Fiber-Optic Links
Keywords:
neural networks; fiber-optic links; performance prediction; optical networks; quality of transmission; socio-technical infrastructure; deployment governance; sustainabilityAbstract
High-capacity fiber-optic links have become foundational to cloud computing, financial settlement, scientific collaboration, public administration, and real-time digital services. As link capacities approach nonlinear Shannon limits and as network operators confront heterogeneous equipment, aging infrastructure, dynamic traffic, and stringent latency expectations, performance prediction has shifted from a narrow engineering calculation to a socio-technical governance problem. This paper examines neural network-based performance prediction for high-capacity fiber-optic links from a system-level perspective. It argues that predictive models should not be evaluated only by regression accuracy on simulated data but also by their architectural fit, operational interpretability, uncertainty behavior, robustness under distribution shift, sustainability consequences, and fairness implications. The discussion first reviews the physical and operational determinants of link performance, including attenuation, dispersion, nonlinear interference, amplifier noise, and quality-of-transmission estimation. It then analyzes neural network architectures for link-level prediction, the role of training data and generalization, and the challenges of integrating learned models into software-defined and elastic optical infrastructures. The paper further considers deployment governance, model lifecycle management, redundancy, and fallback mechanisms, as well as energy use, embodied carbon, rural equity, and policy accountability. It concludes that neural network performance prediction can improve capacity utilization and resilience only if it is embedded within transparent, auditable, and institutionally accountable network operations. The central claim is that predictive accuracy is necessary but insufficient; sustainable high-capacity optical infrastructure requires models that are robust, explainable, and aligned with public-interest governance.
References
1. Agrawal, G. P. (2021). Fiber-optic communication systems (5th ed.). Wiley.
2. Cisco. (2020). Cisco annual internet report (2018–2023) white paper. Cisco Systems.
3. Essiambre, R.-J., Kramer, G., Winzer, P. J., Foschini, G. J., & Goebel, B. (2010). Capacity limits of optical fiber networks. Journal of Lightwave Technology, 28(4), 662–701.
4. Winzer, P. J., Neilson, D. T., & Chraplyvy, A. R. (2018). Fiber-optic transmission and networking: The previous 20 and the next 20 years. Optics Express, 26(18), 24190–24239.
5. Kobyakov, A., Sauer, M., & Chowdhury, D. (2010). Stimulated Brillouin scattering in optical fibers. Advances in Optics and Photonics, 2(1), 1–59.
6. Savory, S. J. (2010). Digital coherent optical receivers: Algorithms and subsystems. IEEE Journal of Selected Topics in Quantum Electronics, 16(5), 1164–1179.
7. Poggiolini, P. (2012). The GN model of non-linear propagation in uncompensated coherent optical systems. Journal of Lightwave Technology, 30(24), 3857–3879.
8. Rottondi, C., Chrapko, D., Valvo, M., Savory, S. J., & Tornatore, M. (2019). Machine learning for quality of transmission estimation in optical networks. IEEE Communications Surveys & Tutorials, 21(1), 233–260.
9. Mata, J., de Miguel, I., Durán, R. J., Merayo, N., Singh, S. K., Jukan, A., & Chamania, M. (2018). Artificial intelligence (AI) methods in optical networks: A comprehensive survey. Optical Switching and Networking, 28, 43–57.
10. Jiménez, T., Aguado, J. C., de Miguel, I., Durán, R. J., Lorenzo, R. M., & Abril, E. J. (2014). A cognitive quality of transmission estimator for core optical networks. Journal of Lightwave Technology, 32(5), 942–951.
11. Caballero, A., Aguado, J. C., de Miguel, I., Durán, R. J., Lorenzo, R. M., & Abril, E. J. (2018). Machine learning based linear and nonlinear noise estimation. Journal of Optical Communications and Networking, 10(9), D42–D51.
12. LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444.
13. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
14. Zhou, Z., Wei, J., Luo, Y., Clark, K. A., Sillekens, E., Deakin, C., ... & Liu, Z. (2023). Communications with guaranteed bandwidth and low latency using frequency-referenced multiplexing. Nature Electronics, 6(9), 694-702.
15. Shalev-Shwartz, S., & Ben-David, S. (2014). Understanding machine learning: From theory to algorithms. Cambridge University Press.
16. Thyagaturu, A. S., Mercian, A., McGarry, M. P., Reisslein, M., & Kellerer, W. (2016). Software defined optical networks (SDONs): A comprehensive survey. IEEE Communications Surveys & Tutorials, 18(4), 2738–2786.
17. Napoli, A., Bohn, M., Rafique, D., Stavropoulos, T., & Tomkos, I. (2015). Next generation elastic optical networks: The vision of the European research project IDEALIST. IEEE Communications Magazine, 53(2), 152–162.
18. Zhang, G., De Leenheer, M., Morea, A., & Mukherjee, B. (2013). A survey on OFDM-based elastic core optical networking. IEEE Communications Surveys & Tutorials, 15(1), 65–87.
19. Kani, J., Bourgart, F., Cui, A., Rafel, A., Campbell, M., Davey, R., & Rodrigues, S. (2009). Next-generation PON-part I: Technology roadmap and general requirements. IEEE Communications Magazine, 47(11), 43–49.
20. Boutaba, R., Salahuddin, M. A., Limam, N., Ayoubi, S., Shahriar, N., Estrada-Solano, F., & Caicedo, O. M. (2018). A comprehensive survey on machine learning for networking: Evolution, applications and research opportunities. Journal of Internet Services and Applications, 9(1), 16.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Journal of Advanced Artificial Intelligence Research
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
