Energy-Efficiency Optimization in Relay Link Transmissions of Ground Relay Stations in Satellite Mobile Communication Networks

Equipped with solar energy and wind energy supply, satellite ground relay stations (SGRS) can be deployed in remote regions without the support of ground power network to provide communication access for ordinary mobile terminals in satellite mobile communications networks. Due to the instability and limitation of solar energy and wind energy supply, to reduce energy consumption and improve energy efficiency of SGRS is very important. Focusing on the unstable and limited energy problem on SGRS, this article proposes a soft actor-critic based joint resource allocation (SAC-JRA) algorithm to improve the transmission energy efficiency of SGRS. The proposed SAC-JRA algorithm includes a multi-dimensional state space, a hierarchical action space, and a dual-weight reward function to balance energy efficiency and user satisfaction. A lightweight dual-branch neural network is used to decouple power and bandwidth allocation, enhancing decision-making efficiency and adaptability. Simulation results demonstrate that the proposed SAC-JRA algorithm achieves significant improvements in transmission energy efficiency compared with existing methods, especially in low traffic load and mixed traffic load environments. The results also show the robustness of the proposed SAC-JRA algorithm in dynamic traffic loads, offering a sustainable solution for energy-constrained wireless communications.