Dynamic Multi-Routes Optimization Algorithm Based On Space Transformation

The load and communication link of the Low Earth Orbit (LEO) satellite network change in real time. How to search for a suitable route for each communication task in LEO satellite network with high dynamics and limited spectrum resources to minimize the total communication delay and balance the network load is the core of route optimization. Current route optimization only considers single-task scenarios and static problems, resulting in large route conflicts and serious waste of resources due to repeated calculations. In order to optimize multi-task routes efficiently, this paper proposes a dynamic multi-route optimization algorithm based on space transformation. The discrete route individuals are converted into latent space through the pre-trained model and then execute dynamic response strategy for each subpopulation in latent space, this allows the population to quickly track the optimal solution when the environment changes. Finally, the effectiveness of the method proposed in this paper is proved by experiments.