A Heuristic Multi-Objective Routing Optimization Algorithm Based on Multidimensional Resource Information for Satellite Communication Networks

With the rapid advancement of global large-scale satellite networks, the construction of satellite constellations imposes stricter demands on inter-satellite routing algorithms. Traditional routing methods prioritize inter-satellite link states and topological structures, overlooking onboard computing, storage resources, and task load status, which becomes particularly problematic in scenarios involving computing tasks and inter-satellite data transmissions, where the efficient use of satellite resources is essential. To address these challenges, this paper constructs a joint delay and multidimensional resources (including computing, storage, and task load status) optimization model for the routing problem of satellite networks, and proposes an improved multi-objective genetic algorithm based on joint multidimensional resource information and shortest path(IMG-MRSP). The algorithm simultaneously optimizes the delay in path communication and the utilization of satellite resources, aiming to select paths through well-resourced satellites for Pareto-optimal routing decisions. First, a multidimensional resource cost model is developed to precisely capture the dynamic characteristics of resource utilization in satellite networks. The routing optimization problem is then formulated as a mixed integer non-linear programming problem. By injecting elite solutions, generated by the shortest path algorithm based on delay and resource costs, into the initial population of the genetic algorithm, the algorithm enhances population diversity and global search efficiency. The simulation results demonstrate that IMG-MRSP excels in convergence, diversity of solutions, and utilization of resources on board.The proposed IMG-MRSP algorithm can be applied in areas such as onboard computing in future satellite networks.