Investigation into Rain Attenuation Prediction Models at Locations in Lagos Using Remote Sensing

  • Olufunke Darley
  • Abayomi I.O. Yussuff
  • Adetokunbo A. Adenowo
Keywords: Path attenuation, Prediction models, Rainfall rate, Terrestrial microwave links, Tropical region

Abstract

This paper investigated the performances of some rain attenuation prediction models at some GSM network locations in Lagos, Nigeria, using remote sensing at Ku band. Remote sensing is a collection and interpretation of information about an object without physical contact with the object being measured. Three popular terrestrial prediction models were considered in this work. These are ITU-R P.530-17, Lin and Silva Mello Models. Ten years (2010-2019) annual rainfall data with hourly integration time were sourced from the Nigerian Meteorological Agency (NIMET) and link budgets for three microwave links (Tarzan Yard, Kofo Abayomi and GLO Shop) in Victoria Island at 18 GHz were obtained from Global Communications Limited (GLO), Nigeria. Data analysis and comparison of the microwave links rainfall estimates were carried out to identify the most suitable of the three models at the selected locations of interest. Measurement data obtained from both NIMET and GLO were used to validate the predicted attenuation data from the three selected models. The ITU-R P.530-17 prediction model overestimated the measurement at Tarzan Yard; closely followed by Silva Mello, while Lin underestimated the measured data. Again, at Kofo Abayomi station, the ITU-R model overestimated the measurement, while both Silva Mello and Lin models underestimated the measurement. At the GLO Shop, the Silva Mello overestimated the measured value, while ITU-R and Lin underestimated the measurement. At 0.01% of time exceeded, NIMET measurement was higher (at 48.2 dB) than that of Tarzan Yard, Kofo Abayomi and GLO shop (43.1, 46.3 and 37.0 dB respectively). These results will provide useful information in mitigating signal outages due to rain for mobile communication systems.

Published
2021-09-13
Section
Articles

Journal Identifiers


eISSN: 2579-0617
print ISSN: 2579-0625