Pseudo-Random Asynchronous Multi-Satellite Cooperative Transmission Scheme for Cohesive Clustered Satellite Networks

Asynchronous interference mitigation has great potential for efficient multi-satellite cooperative transmission (MSCT) in cohesive clustered satellite networks. In this paper, we propose a pseudo-random asynchronous MSCT (PRA-MSCT) scheme to enhance the ergodic sum rate of asynchronous MSCT. We design a normalized waiting delay control mechanism to reshape the delay difference between satellites and user equipment (UEs), thereby maximizing the benefits of time asynchrony to mitigate inter- and intra-satellite asynchronous interference. We formulate an optimization problem for cooperative precoding and delay control to maximize the ergodic sum rate under power constraint of satellites. To address the decision coupling that complicates optimal problem-solving, we propose a graph neural network (GNN)-empowered PRA-MSCT approach, enabling satellites to make autonomous decisions based on their local observations. Simulation results demonstrate that the PRAMSCT scheme significantly improves the sum rate and computation efficiency compared to existing asynchronous MSCT schemes. Additionally, the GNN-empowered cooperative precoding approach can be applied effectively to both synchronous and asynchronous scenarios, with or without delay control.