PREDICTION OF TROPOSPHERIC SURFACE DUCTING IN ABUJA USING ARTIFICIAL NEURAL NETWORK (ANN)

Dr Ikharo Braimoh Abdullahi   (Published 2017)

Dr Braimoh Abdullahi
See Profile Page

Abstract

Meteorological forecasting is an ever-challenging area of investigation for scientists. Multipath fading is considered to arise from irregularities of the atmospheric refractive index, called radio ducts. By these irregularities, radio waves propagations over this medium are adversely impaired. In this paper, Artificial Neural Network is developed for the purposes of modelling meteorological conditions of Abuja – Federal Capital Territory. The work uses meteorological parameters like temperature, pressure and relative humidity for the purpose of predicting duct presence in Abuja over different periods of the year. The approach applied here uses feed forward artificial neural networks (ANNs) with back propagation for supervised learning using the data recorded at this locality. The trained ANN was used to predict the duct presence in the designated site. The results obtained are very encouraging and it is found that the ANN model can make predictions with high degree of accuracy (over 90%). Our results showed that duct occurrence is not prevalent but adequately visible in the Federal Capital Territory of Nigeria and duct occurrences have tendencies to be season dependent.


Item Type: Journal article(non-copyrighted)
Format: PDF document,   554.27 KB
Copyright: Creative Commons LicenseCreative Commons license
Keywords: Meteorological, Refractivity, Communication, Backpropagation, Ducting, Prediction
Department: Computer Engineering
Field of Study: Engineering- Electrical and Computer
Uploaded By: Agbodekhe Barnabas Philip
Date Added: 25 Feb 2019 8:37am
Last Modified: 25 Feb 2019
Journal URL: https://www.edouniversity.edu.ng/oer/journal/prediction_of_tropospheric_surface_ducting_in_abuja_using_artificial_neural_network_ann


Google Docs     Print     Download