volume | PIER Journals

Abstract

Reliable modelling of rainfall-induced attenuation is crucial for designing and operating high-frequency communication systems, particularly those operating above 10 GHz, in regions with severe rainfall conditions such as subtropical climates. This study offers a comparison of supervised machine learning (ML) models - k-nearest neighbours (KNN), decision trees (DT), and random forests (RF) - against traditional statistical methods such as the lognormal and gamma distributions for estimating raindrop size distribution (DSD) and specific attenuation. Rainfall measurements taken between 2018 and 2019 were obtained from 1-minute disdrometer at the measurement location in Durban, South Africa (29.8651°S, 30.9734°E). The investigated models were then tested across four different rainfall regimes processed from the dataset: drizzle, widespread rain, shower, and thunderstorm. An adaptive tuning method for selecting the best k-value in KNN was introduced to enhance prediction accuracy across various rainfall intensities. Model performances are evaluated using error metrics of root mean square error (RMSE), mean absolute error (MAE), and the coefficient of determination (R²). The results show that KNN outperforms RF and DT, providing the highest accuracy and lowest prediction errors across all rainfall regimes. The confidence interval (CI) analysis confirms that KNN delivers more precise and stable estimates, while RF and DT exhibit greater variability and uncertainty in performance. Additionally, specific attenuation estimations from these ML models are compared at different rain rates with the ITU-R P.838-8 estimations for frequencies up to 100 GHz. These findings highlight the superiority of data-driven model, particularly the adaptive KNN, in capturing complex rainfall microstructures and improving attenuation predictions. This has direct implications for planning and deploying rain-resilient wireless networks in variable climatic regions.

1. Odedina, M. O. and T. J. Afullo, "Determination of rain attenuation from electromagnetic scattering by spherical raindrops: Theory and experiment," Radio Science, Vol. 45, No. 01, 1-15, 2010.
doi:10.1029/2009rs004192 Google Scholar

2. Alonge, Akintunde Ayodeji and Thomas Joachim Odhiambo Afullo, "Seasonal analysis and prediction of rainfall effects in Eastern South Africa at microwave frequencies," Progress In Electromagnetics Research B, Vol. 40, 279-303, 2012.
doi:10.2528/pierb12020305 Google Scholar

3. Adimula, I. A. and G. O. Ajayi, "Variations in raindrop size distribution and specific attenuation due to rain in Nigeria," Annals of Telecommunications, Vol. 51, No. 1, 87-93, 1996. Google Scholar

4. Timothy, K. I., Jin Teong Ong, and E. B. L. Choo, "Raindrop size distribution using method of moments for terrestrial and satellite communication applications in Singapore," IEEE Transactions on Antennas and Propagation, Vol. 50, No. 10, 1420-1424, Oct. 2002.
doi:10.1109/tap.2002.802091 Google Scholar

5. Ulbrich, Carlton W., "Natural variations in the analytical form of the raindrop size distribution," Journal of Climate and Applied Meteorology, Vol. 22, No. 10, 1764-1775, 1983. Google Scholar

6. Ajayi, G. Olalere and Roderic L. Olsen, "Modeling of a tropical raindrop size distribution for microwave and millimeter wave applications," Radio Science, Vol. 20, No. 02, 193-202, 1985.
doi:10.1029/rs020i002p00193 Google Scholar

7. Owolawi, P., "Raindrop size distribution model for the prediction of rain attenuation in Durban," Progress In Electromagnetics Research Symposium Proceedings, 1068-1074, Suzhou, China, 2011.

8. Alonge, Akintunde A. and Thomas J. Afullo, "Rainfall microstructures for microwave and millimeter wave link budget at tropical and subtropical sites," 2013 Africon, 1-5, Pointe aux Piments, Mauritius, 2013.
doi:10.1109/AFRCON.2013.6757852

9. Alonge, Akintunde A., "Semi-empirical characteristics of modified lognormal DSD inputs using rain rate distributions for radio links over the African continent," Advances in Space Research, Vol. 67, No. 1, 179-197, 2021.
doi:10.1016/j.asr.2020.09.017 Google Scholar

10. Sarker, Iqbal H., "Machine learning: Algorithms, real-world applications and research directions," SN Computer Science, Vol. 2, No. 3, 160, 2021.
doi:10.1007/s42979-021-00592-x Google Scholar

11. Kozu, Toshiaki and Kenji Nakamura, "Rainfall parameter estimation from dual-radar measurements combining reflectivity profile and path-integrated attenuation," Journal of Atmospheric and Oceanic Technology, Vol. 8, No. 2, 259-270, 1991. Google Scholar

12. Tokay, Ali and David A. Short, "Evidence from tropical raindrop spectra of the origin of rain from stratiform versus convective clouds," Journal of Applied Meteorology and Climatology, Vol. 35, No. 3, 355-371, 1996. Google Scholar

13. Testud, Jacques, Stéphane Oury, Robert A. Black, Paul Amayenc, and Xiankang Dou, "The concept of “normalized” distribution to describe raindrop spectra: A tool for cloud physics and cloud remote sensing," Journal of Applied Meteorology and Climatology, Vol. 40, No. 6, 1118-1140, 2001. Google Scholar

14. Altman, N. S., "An introduction to kernel and nearest-neighbor nonparametric regression," The American Statistician, Vol. 46, No. 3, 175-185, 1992.
doi:10.1080/00031305.1992.10475879 Google Scholar

15. Breiman, Leo, Jerome Friedman, R. A. Olshen, and Charles J. Stone, Classification and Regression Trees, Chapman and Hall/CRC, 2017.
doi:10.1201/9781315139470

16. Ramatladi, Tsietsi and Akintunde Alonge, "Machine learning techniques for evaluating disdrometer-derived raindrop measurements over radio links," 2023 IEEE AFRICON, 1-6, Nairobi, Kenya, 2023.
doi:10.1109/AFRICON55910.2023.10293561

17. Afullo, Thomas Joachim Odhiambo, "Raindrop size distribution modeling for radio link design along the eastern coast of South Africa," Progress In Electromagnetics Research B, Vol. 34, 345-366, 2011.
doi:10.2528/pierb11082005 Google Scholar

18. Marshall, J. S. and W. M. K. Palmer, "The distribution of raindrops with size," Journal of Meteorology, Vol. 5, No. 4, 165-166, 1948. Google Scholar

19. Malinga, Senzo Jerome and Pius Adewale Owolawi, "Obtaining raindrop size model using method of moment and its applications for South Africa radio systems," Progress In Electromagnetics Research B, Vol. 46, 119-138, 2013.
doi:10.2528/pierb12070806 Google Scholar

20. Alonge, Akintunde Ayodeji, "Correlation of rain dropsize distribution with rain rate derived from disdrometers and rain gauge networks in Southern Africa," MSc dissertation, Masters Dissertation submitted to the University of KwaZulu-Natal, Durban, 2011.

21. Cover, T. and P. Hart, "Nearest neighbor pattern classification," IEEE Transactions on Information Theory, Vol. 13, No. 1, 21-27, 1967.
doi:10.1109/tit.1967.1053964 Google Scholar

22. Cho, Poongjin, Minhyuk Lee, and Woojin Chang, "Instance-based entropy fuzzy support vector machine for imbalanced data," Pattern Analysis and Applications, Vol. 23, No. 3, 1183-1202, 2020.
doi:10.1007/s10044-019-00851-x Google Scholar

23. Beyer, Kevin, Jonathan Goldstein, Raghu Ramakrishnan, and Uri Shaft, "When is “nearest neighbor” meaningful?," International Conference on Database Theory, 217-235, 1999.

24. Quinlan, J. R., "Induction of decision trees," Machine Learning, Vol. 1, No. 1, 81-106, 1986.
doi:10.1023/a:1022643204877 Google Scholar

25. Breiman, Leo, "Random forests," Machine Learning, Vol. 45, No. 1, 5-32, 2001.
doi:10.1023/a:1010933404324 Google Scholar

26. Mallala, Balasubbareddy, Azka Ihtesham Uddin Ahmed, Sastry V. Pamidi, Md. Omar Faruque, et al. "Forecasting global sustainable energy from renewable sources using random forest algorithm," Results in Engineering, Vol. 25, 103789, 2025.
doi:10.1016/j.rineng.2024.103789 Google Scholar

27. Louppe, Gilles, Understanding random forests: From theory to practice, Universite de Liege (Belgium) , 2004.

28. Bartholomew, Mary Jane, Laser Disdrometer (LDIS) Instrument Handbook, Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States). Atmospheric Radiation Measurement (ARM) Data Center; Brookhaven National Laboratory (BNL), Upton, NY (United States), 2020.
doi:10.2172/1226796

29. Alonge, A. A. and T. J. Afullo, "Regime analysis of rainfall drop-size distribution models for microwave terrestrial networks," IET Microwaves, Antennas & Propagation, Vol. 6, No. 4, 393-403, 2012.
doi:10.1049/iet-map.2011.0411 Google Scholar

30. Tokay, Ali, Anton Kruger, Witold F. Krajewski, Paul A. Kucera, and Augusto Jose Pereira Filho, "Measurements of drop size distribution in the southwestern Amazon basin," Journal of Geophysical Research: Atmospheres, Vol. 107, No. D20, LBA 19-1-LBA 19-15, 2002.
doi:10.1029/2001jd000355 Google Scholar

31. Liaw, Andy and Matthew Wiener, "Classification and regression by randomForest," R News, Vol. 2, No. 3, 18-22, 2002. Google Scholar

32. Hastie, Trevor, Robert Tibshirani, and Jerome Friedman, The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Springer, 2009.
doi:1007/978-0-387-84858-7

33. Loh, Wei-Yin, "Classification and regression trees," Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, Vol. 1, No. 1, 14-23, 2011.
doi:10.1002/widm.8 Google Scholar

34. Chai, T. and R. R. Draxler, "Root mean square error (RMSE) or mean absolute error (MAE)? – Arguments against avoiding RMSE in the literature," Geoscientific Model Development, Vol. 7, No. 3, 1247-1250, 2014.
doi:10.5194/gmd-7-1247-2014 Google Scholar

35. Willmott, Cort J. and Kenji Matsuura, "Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance," Climate Research, Vol. 30, No. 1, 79-82, 2005.
doi:10.3354/cr030079 Google Scholar

36. Chicco, Davide, Matthijs J. Warrens, and Giuseppe Jurman, "The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation," PeerJ Computer Science, Vol. 7, e623, 2021.
doi:10.7717/peerj-cs.623 Google Scholar

37. Altman, Douglas, David Machin, Trevor Bryant, and Martin Gardner, Statistics with Confidence: Confidence Intervals and Statistical Guidelines, John Wiley & Sons, 2013.

38. Cumming, G. and S. Finch, "Inference by eye: Confidence intervals and how to read pictures of data," American Psychologist, Vol. 60, No. 2, 170-180, 2005.
doi:10.1037/0003-066x.60.2.170 Google Scholar

39. Crane, Robert K., Propagation Handbook for Wireless Communication System Design, CRC Press, 2003.
doi:10.1201/9780203506776

40. Hulst, Hendrik Christoffel and Hendrik C. van de Hulst, Light Scattering by Small Particles, Courier Corporation, 1981.

41. Mie, Gustav, "Beiträge zur optik trüber medien, speziell kolloidaler metallösungen," Annalen Der Physik, Vol. 330, No. 3, 377-445, 1908.
doi:10.1002/andp.19083300302 Google Scholar

42. ITU, "Specific attenuation model for rain for use in prediction methods," Rec. ITU-R P.838, 1992. Google Scholar

Dr. Tsietsi Condry Ramatladi

Dr. Akintunde Ayodeji Alonge