A Three Index Profile Core Dataset for Designing Few Modes Fiber for Long Haul Optical Communications

S. P. Obaje, Z. M. Abdullahi, A. D. Usman, I. Yau, Agbon E. Ezeikel, A. S. Yaro

Abstract


Fiber is widely recognized as the medium for next-generation communication, offering secure transmission and efficient switching capabilities. Few-Mode Fiber (FMF), which accommodates multiple modes with weak coupling, is particularly attractive for Mode Division Multiplexing (MDM) systems. Nevertheless, the multi-parameter characteristics of FMF make its design highly complex, often inaccurate, and time-intensive due to its intricate structure and the presence of numerous higher-order modes. This research work presents a 3-index profile core dataset for FMF design with the aid of Machine Learning (ML) regression model for long haul optical communications applications. The generated dataset for the FMF design parameters was developed for long haul optical fiber length, which span 100 Km to 1000 Km. The FMF parameters used for generating the dataset consists of the effective refractive index of core, the radian distance of the 3-index core profile and the difference between the refractive indices of the 3-index core and that of the cladding. With additional design properties which consist of the fiber birefringence of the propagating modes, the coupling correlation lengths of the optical length, and the coupling coefficient of the propagating modes, the Differential Mode Group Delay (DMGD) of the fiber design and the IM-XT for the five guides modes for the FMF. The realized FMF design dataset was generated aid of numerical equations that characterize FMF parameters and MATLAB software. The dataset of the FMF design parameters consists of 10000 rows and 53 columns with step size of 1×10-4. The simulation analysis carried out for the 3-index FMF design dataset validation shows that, the correlation matrix analysis, for very high positive entries of 10.5%, for high positive entries of 12%, moderate positive entries of 15%, moderate negative entries of 6%, high negative entries of 3%, very high negative entries of 1% and for weak/no entries of 52.5%. Furthermore, the boxplot analysis for the percentage outlier and normal data for each design parameters was presented. They are, overall average percentage outlier of 2.80% and that of normal data of 97.20% was realized. Finally, the percentage cumulative variance for the first Principal Component (PC1) of 63.9795%, that of second the Principal Component (PC2) of 83.2868% and that of third the Principal Component (PC3) is 13.2516% respectively. With the aid of the developed dataset, machine learning models can be trained to predict different optimal design parameters for FMF, therefor minimizing design complexities and enhance precision in FMF designs.


Full Text:

PDF

References


Agrawal, G. P. (2010) Fiber-Optic Communication Systems, Wiley-Interscience, 4th Edition, 2010.

Arun, K. V. & John C. M. (2019). Fundamentals of Inorganic Glasses (Third Edition) pp 537-594.

Bhagyalaxmi, B., Gyana, R. P., Shailendra, K. V., & Mihir, N. M. (2023). Machine Learning-based Inverse Model for Few-Mode Fiber Designs. Computer Systems Science & Engineering. 45(1):311-328. DOI: 10.32604/csse.2023.029325.

Chandola, V., Banerjee, A., & Kumar, V. (2009). Anomaly detection: A survey. ACM Computing Surveys (CSUR).

Hall, M. A. (1999). Correlation-based Feature Selection for Machine Learning. researchcommons.waikato.ac.nz

Junling Hu, Hongwei Li, Hailiang Chen, Sa Zhang, Ruyue Shi, Yuxin Li, Meiyu Cai, and Shuguang Li (2024). Forward design method for the design of panda polarization-maintaining few-mode optical fiber based on artificial neural network. Vol. 32, No. 21 / 7 Oct 2024 / Optics Express 36848.

Ken-ichi K. & Nikolaos-Panteleimon, D., (2017). Few-Mode Optical Fibers: Original Motivation and Recent Progress. IEEE Communications Magazine. DOI: 10.1109/MCOM.2017.1600876

Linqi, W., Zhang, X., Liu, Y., et al. (2022). Mode coupling and differential mode group delay analysis in few-mode fibers for SDM systems. Journal of Lightwave Technology, 40(12), 3456–3465.

Liu, S. & Xiang, L. (2024). Inter-Mode Crosstalk Estimation between Cores for LPmn Modes in Weakly Coupled Few-Mode Multicore Fiber with Perturbations. Sensors 2024, 24(18), 5969; https://doi.org/10.3390/s24185969

Luca, P. (2025). Mode Coupling in Optical Fibers. May 2025. Journal of Lightwave Technology. PP(99):1-12. DOI: 10.1109/JLT.2025.3543623

Mizuno, T. & Miyamoto, Y. (2017). High-capacity dense space division multiplexing transmission. OpticalFiberTechnology35(2017)108–117.

Pierre, S., Marianne B-A., & Denis, M. (2014). Few-Mode Fibers for Mode-Division-Multiplexed Systems. Journal of Lightwave Technology. 32(16):2824-2829

Qiuyan, Y., Hui, Yang, R. Z., Ao, Y., Wei, B., Yuanlong, A., Jie, Z., & Hongyun, X. (2018). Core, Mode, and Spectrum Assignment Based on Machine Learning in Space Division Multiplexing Elastic Optical Networks. IEEE Access, 6, 15898-15907.

Rademacher, G., Luis, R. S., Puttnam, B., Ryf, R., van der Heide, S., Eriksson, T. A., Fontaine, N. K., Chen, H., Essiambre, R. J., Awaji, Y., & Furukawa, H. (2022). A Comparative Study of Few-Mode Fiber and Coupled-Core Multi-Core Fiber Transmission. Journal of Lightwave Technology, 40(6), 1590-1596.

Shreya M. (2023). Fiber Optics in Communication Networks: Trends,

Challenges, and Future Directions. International Journal of All Research Education and Scientific Methods (IJARESM), Volume 11, Issue 7

Snyder, A.W., & Love, J. D. (1983). Optical Waveguide Theory. 3(3), pp. 378-379

Xinyi, C., Jiangbing, D., Baining, Y., Ke, X., & Zuyuan, H. (2022). "Active learning aided four-mode fiber design with equalized zero dispersion for short-reach MDM optical communications," J. Opt. Soc. Am. B 39, 1958-1963

Zhiqin, H., Jiangbing, D., Xinyi, C., Weihong, S., Yuting, H., Chang W., Ke X., & Zuyuan, H. (2020). Machine learning aided inverse design for few-mode fiber weak-coupling optimization. Optics Express, 28(15).


Refbacks

  • There are currently no refbacks.