The Effects of Rain on Terrestrial Links at K, Ka and E-Bands in South Korea: Based on Supervised Learning
Open Access
- 8 January 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
Abstract
At the rise of the fourth industrial revolution, artificial intelligence (AI), along with key enabling technologies such as millimeter waves (mm-waves) can be used to launch the fifth-generation (5G) and beyond communication links. However, the quality of radio links at higher frequency bands is limited by atmospheric elements. Among others, rainfall is the major propagation impairment at millimetric wave bands, which needs to be considered during the link budget planning. In this study, we investigated the rain attenuation results obtained from experimental data, existing models, and proposed supervised artificial neural network (SANN) at K, Ka, and E-bands, respectively, for terrestrial links in South Korea. The measurement campaigns were between Incheon, National Radio Research Agency (RRA) tower station, to the EMS Dongyoksang tower station operating at 75 GHz over a 100-m path length, and between Incheon, RRA tower station to Khumdang, Korea Telecom (KT) tower station, operating at 18 and 38 GHz over a 3.2-km path length. The three-year rainfall and received signal level data measurements over these paths were used to determine rain attenuation distributions at different percentages of exceedance time distribution. Additionally, three existing attenuation models, ITU-R 530.17, Lin, and Revised Silva Mello (RSM) models were compared with measured rain attenuation. Our results indicate that these models did not correspond with measured results. Therefore, in this research, we proposed a supervised learning-based attenuation prediction method, which provides better performance than existing models. Furthermore, we validated our proposed model with measured received-signal level and rainfall data at the above-mentioned operating frequencies.Funding Information
- research fund from Chosun University 2020
This publication has 34 references indexed in Scilit:
- Rain Attenuation Prediction Model for Satellite Communications in Tropical RegionsIEEE Transactions on Antennas and Propagation, 2014
- Millimeter Wave Mobile Communications for 5G Cellular: It Will Work!IEEE Access, 2013
- The SC EXCELL model for prediction of rain attenuation on terrestrial radio linksElectronics Letters, 2013
- Unified method for the prediction of rain attenuation in satellite and terrestrial linksJournal of Microwaves, Optoelectronics and Electromagnetic Applications, 2012
- RAIN ATTENUATION PREDICTIONS AT KU-BAND IN SOUTH EAST ASIA COUNTRIESProgress In Electromagnetics Research, 2007
- Prediction of rain attenuation in terrestrial links using full rainfall rate distributionElectronics Letters, 2007
- Improvement of a rain attenuation prediction method for terrestrial microwave linksIEEE Transactions on Antennas and Propagation, 1984
- The aRbrelation in the calculation of rain attenuationIEEE Transactions on Antennas and Propagation, 1978
- 11-GHz Radio: Nationwide Long-Term Rain Rate Statistics and Empirical Calculation of 11-GHz Microwave Rain AttenuationBell System Technical Journal, 1977
- A Method for Calculating Rain Attenuation Distributions on Microwave PathsBell System Technical Journal, 1975