AI Based Short-Term Prediction of Q-Band Rain Attenuation

In broadband satellite networks, adaptive coding and modulation (ACM) serves as a standard practice to mitigate the disruptive effects of rain fades, particularly at extremely high frequency (EHF) bands. ACM has emerged as an indispensable component of Fade Mitigation Techniques (FMTs) to bolster signal resilience. Nonetheless, a notable challenge lies in ACM's operation without real-time channel state information. To tackle this hurdle, this study leverages Deep Learning (DL) algorithms for rain attenuation prediction, facilitating efficient and optimal selection of modulation and coding (MODCODs) schemes within the ACM framework. The DL algorithm uses recent past rain attenuation data to forecast short-term trends. The investigation delves into the influence of recent past data on the accuracy of DL algorithms and evaluates their efficacy in enhancing predictions. The attained results furnish insights that underpin the development of a novel rain attenuation prediction framework. This framework aims to augment the overall MODCOD selection process of the Digital Video Broadcasting – Return Channel via Satellite – Second Generation (DVB-RCS2) Standards, bolstering reliability and resilience, particularly in adverse weather conditions.