Dynamic Inter-satellite Routing Strategy for Computation and Transmission
Low Earth Orbit(LEO)satellite networks have the characteristics of rapid topology changes,varying network nodes and fluctuations in node resource availability.An inter-satellite routing strategy for computation and transmission is proposed,which uses the Enhanced-Graph Convolutional Network(EGCN)to extract the spatiotemporal features from the satellite network and generates the hid-den states for each network node.As input to the Deep Reinforcement Learning(DRL)model,the DRL agent senses the key informa-tion of the next-hop node to make better decisions.Simulation results show that,compared to previous methods,the proposed method not only improves the overall throughput of the network,but also reduces the end-to-end transmission delay.
国家科技期刊平台
NSTL
万方数据
