On the use machine learning for flexible payload management in VHTS systems

Very High Throughput Satellites (VHTS) are next generation of satellite systems to meet the demands of increase on data traffic. The objective of VHTS systems is to achieve 1 Terabit/s by satellite communications in the near future. VHTS systems are based on multi-beam payloads with polarization and frequency reuse schemes, with VHTS using Q/V bands in the feeder link to increase available bandwidth. These systems provide a greater satellite capacity at a reduced cost per Gbps in orbit but further optimization is needed in order to use the full capacity of the satellite over the time. For instance, flexible payloads are required in VHTS to meet changing traffic demands. Whereby, this contribution presents a study of how and where Machine Learning algorithms can be used to manage a flexible payload architecture. The problem of resource allocation in a flexible payload architecture is analyzed to implement the application of ML as a solution for non-uniform traffic demand and its changes over the time in the service area