Zero-Shot Learning for Beam Management and Resource Allocation in LEO Satellite Communications

Abstract

The integration of Non-Terrestrial Networks (NTN) with 5G in 3GPP (3rd Generation Partnership Project) standards represents a significant step toward creating a more connected world, where mobile users can access reliable and high-speed connectivity regardless of their location. LEO satellites can both extend coverage in rural and remote areas where terrestrial networks are sparse or non-existent and enhance capacity in densely populated urban areas where they may face congestion issues with lower latency and high data throughput compared with GEO satellites. However, despite their advantages, LEO satellites also face several technical challenges that need to be addressed for effective integration into global telecommunications networks. To ensure better performance of LEO satellite communications, in this thesis I tackle several research challenges in complex beam management and power allocation in LEO satellite communications.