Artificial Intelligence Driven Direct-to-handset Satellite Communication
Direct-to-handset satellite communication can effectively fill the digital divide and becomes an important technology trend in 6G.However,it faces challenges in achieving uplink transmission with high data rate,highly reliable access of multi-users,real-time acquisition of channel state,high-performance digital beamforming under non-ideal radio hardware and efficient scheduling of multi-dimensional resource.To overcome the shortcomings of traditional methods,this paper proposes to leverage artificial intelligence technique and communication-domain knowledge to form an intelligent communication paradigm covering access,transmission and resource scheduling.Specifically,to enhance uplink transmission rate and the reliability of user access,this paper proposes the ideas of artificial intelligence based joint reception of multi-satellites,artificial intelligence based preamble detection,multi-user detection under non-orthogonal transmission of pilots and data,and sematic transmission.To overcome the negative impacts of delayed feedback of chan-nel state and signal impairment and solve beamforming optimization problem with high dimension,this paper proposes the ideas of light-weight prediction of downlink channel state and robust and intelligent on-board digital beamforming.Moreover,this paper also proposes the idea of intelligent scheduling of multi-dimensional resources,which can realize the good match between limited resource and differ-entiated service requirements under complex constraints.The proposals are expected to support the evolution of satellite internet towards intelligent paradigms.
国家科技期刊平台
NSTL
万方数据
