Machine Learning-Based Prediction of Path Attenuation Coefficient for Terrestrial FSO Link: Northern Cape, South Africa
- DOI
- 10.2991/978-94-6463-644-4_20How to use a DOI?
- Keywords
- Bit Error Ratio; FSO Links; Machine Learning Techniques; Path attenuation; Visibility
- Abstract
Free Space Optics (FSO) has garnered a lot of attention in recent years for the transmission of optical signals due to its perceived advantages such as ease of deployment, license-free, and high throughput. However, its performance is impacted negatively by meteorological parameters such as haze, snow, mist, rain, fog, and scattering. These meteorological parameters affect visibility, which ultimately impacts negatively and leads to the outage of FSO link availability. The study applied Machine Learning Techniques (MLT), multiple linear regression, and Ridge and Lasso regression models to 10 years’ data (2010–2019) obtained from the South Africa Weather Service (SAWS). This is to develop predictive models for visibility and path attenuation coefficient for the FSO link with a specific focus on the Northern Cape Province of South Africa. The selected supervised learning techniques were adopted, and their performance was compared. The ridge model yielded the best performance when compared with the measured data based on the Mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2). The path attenuation coefficients are formulated as a function of visibility and wavelength at 1500 nm, 1250 nm, and 950 nm for fogs, which affect the signal link adversely. The selected model (Ridge regression) performed well in predicting the attenuation coefficient for the three wavelengths for a visibility value of 15–30 km but was found to underperform for visibility values less than 15 km in the studied location. The attenuation coefficient was calculated at three different wavelengths for the study location. This was used to carry out a performance analysis of a free space optical communication link based on quality factor, bit error ratio (BER) pattern, and eye height at the three wavelengths using OptiSystem Simulation Software. The best performance metrics were obtained at 950 nm for 1 km link distance while at 3 km and 5 km link distance, the best metrics were obtained at 1250 nm.
- Copyright
- © 2025 The Author(s)
- Open Access
- Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
Cite this article
TY - CONF AU - O. B. Aborisade AU - J. S. Ojo AU - P. Owolawi AU - K. D. Adedayo PY - 2025 DA - 2025/02/04 TI - Machine Learning-Based Prediction of Path Attenuation Coefficient for Terrestrial FSO Link: Northern Cape, South Africa BT - Proceedings of the 8th URSI-NG Annual Conference (URSI-NG 2024) PB - Atlantis Press SP - 199 EP - 212 SN - 2352-541X UR - https://doi.org/10.2991/978-94-6463-644-4_20 DO - 10.2991/978-94-6463-644-4_20 ID - Aborisade2025 ER -