Artificial intelligence for satellite communication: A review

Satellite communication offers the prospect of service continuity over uncovered and under-covered areas, ubiquity, scalability. However, several challenges must first be addressed to realize these benefits, as resource management, network control, security, spectrum energy usage satellite networks are more challenging than that terrestrial networks. Meanwhile, artificial intelligence (AI), including machine learning, deep reinforcement has been steadily growing a research field shown successful results in diverse applications, wireless communication. In particular, application AI wide variety aspects demonstrated excellent potential, beam-hopping, anti-jamming, traffic forecasting, channel modeling, telemetry mining, ionospheric scintillation detecting, interference managing, remote sensing, behavior space-air-ground integrating, managing. This work thus provides general overview AI, its sub-fields, state-of-the-art algorithms. Several facing systems then discussed, their proposed potential AI-based solutions presented. Finally, an outlook is drawn, future steps suggested.