Modelling and Prediction of Satellite Signal Path Loss using the ARIMA models at Ku-band in Lagos State, South Western Nigeria
No Thumbnail Available
Date
2025
Journal Title
Journal ISSN
Volume Title
Publisher
ICSSD 2024 (IOP Conf. Series: Earth and Environmental Science)
Abstract
Satellite signal path loss is the reduction of signal strength between the transmitting
satellite and the receiving antennas on the earth surface. It is cause by effects such as reflection, free
space loss, diffraction, refraction and absorption. It is also cause as a results of the environment,
terrain contours, medium of propagation, height of antennas location and transmitting distance
between the transmitter and receiver. Several methods such as the traditional methods and machine
learning techniques have been explored. However, statistical method of path loss prediction has been
completely utilized. In this work, prediction of path loss using the autoregressive integrated moving
average (ARIMA) technique was considered. Four different ARIMA models at Ku-band were explored
using eight years in Lagos State. The method consists of some steps in the run time analysis: the first stage
involved identifying the ARIMA model by creating an Autocorrelation Function (ACF) and Partial
Autocorrelation Function (PACF) plots and checking if the data are stationary. The second stage involved
estimating the ARIMA model, which consists in testing the parameters' significance after obtaining the model's
parameters. The ARIMA model is adopted if the parameters are significant. The third stage involved diagnostic
checking, in which a test was carried out to see if the adopted model was statistically significant. The fourth
stage involves predicting the path loss after the diagnostic process using the adopted or selected ARIMA model.
From the analysis, it can be concluded that the best ARIMA model for path loss prediction in Lagos at Ku-band
is ARIMA (1,0,0). The ARIMA (1,0,0) model is a better path loss prediction model than the other ARIMA
models, indicating that the model parameters are significant at 0.05 with the slightest MSE error, thereby
meeting the test of independence. Therefore, for sitting and planning of radio signal propagation technologies in the study location, policy-makers and engineers in telecommunication industry in Nigeria can take advantage
of the results within the area of study towards sustainable tele-communication and infrastructure (SDG goal 9).
Description
Keywords
ARIMA, Ku-band, Modelling, Nigeria, Path loss.