Yuhe Mao, Yunhua Wu, Chengfei Yue, Xibin Cao
With a large number of satellites and highly dynamic motion, route planning for mega-constellations faces significant challenges, such as frequent link switching and high complexity of routing calculations. To ensure efficient information transmission onboard, mega-constellations are grouped into multiple management domains with stable configurations, with the consideration of communication range and the stability of satellites' relative motion. A spatiotemporal grid (STG) is proposed to facilitate interdomain route planning, in which the Earth’s surface is divided into finite grids, and then the complex dynamic intersatellite routing problem is converted into static reference trajectory optimization in the spatial domain and grid-satellite matching in the temporal domain. Considering the network perception limitation of satellites, the dynamic programming (DP) algorithm and its optimization are proposed for the routing decision phase. In the DP algorithm, the current domain searches the next routing target according to the reference trajectory. To further reduce the routing hops, the lookahead dynamic programming (LDP) algorithm is proposed to optimize routing decisions through the matching of multistep grids. Simulation results demonstrate that the proposed strategies can effectively approach minimum-hop interdomain routing without requiring global constellation information or time-consuming orbital propagation, which is beneficial for the future on-orbit autonomous application of mega-constellations.