Utilization of Artificial Neural Networks for Estimation of Slant-Path Rain Attenuation

The ever-growing exploitation of various telecommunication services and radar applications resulted in saturation of low frequency bands supporting them and therefore being unable to handle new incoming traffic demands for services such as broad coverage network connection and supper high and extreme speed data transmission. Hence, a call for utilization of higher bands radio waves spectrums become imperative to address spectrum congestion and telecommunication service demands by the users. There are two platforms of communication at these upper frequencies, terrestrial and satellite. Although, rain attenuation affects the two platforms, the focus of this paper is on the satellite communication. Rain-induced attenuation is a main challenge of signal propagation in satellite communication networks operating in upper frequency bands and its impact becomes pronounced at frequency above 10 GHz. Therefore, accurate rain fading estimation for links path is of outmost importance in building reliable networks. This paper presents the adoption of Artificial Neural Networks (ANNs) as a technique in estimating rain-induced attenuation along the satellite links at the higher frequency in South Africa. The proposed ANNs model indicated improvement in estimation performance in terms of root mean square error and correlation coefficient.