Research On Dynamic Scheduling Approach Of Relay Satellites

As the backbone of space-based network,the relay satellite system can provide tracking,telemetering and data relay services for many aircrafts in the space.The relay satellite scheduling is that the ground control center of the relay satellite,according to application demand,selects the suitable aerospace craft,the relay satellite and its payloads,assigns the relay satellite system resources scientifically to meet the demand of different users.On account of the contradiction between the growing task demand and the limited transmission resources in the relay satellite,better scheduling strategy has become the guarantee to maximize resource utilization in the relay satellite system.Up to now,most research work on dynamic task scheduling mainly focused on maximizing the system benefits with little consideration on the system resources utilization and the task delay,which leads to the inefficient utilization of system transmission resources.So some more efficient dynamic scheduling algorithm needs to be put forward with guaranteeing the maximum benefit.The contributions of this paper are listed as follows:Firstly,in this paper,according to the complex constraints such as different characteristics of tasks,time-varying visible time windows,different frequency bands of inter-satellite links,a mathematical model for tasks dynamic scheduling problem of the relay satellite system is presented in the paper.With this model,the available effective resources of the relay satellite system is analyzed.On the basis of theoretical analysis,a hybrid particle swarm algorithm based on rolling windows(HPSORS)is proposed in order to realize the pre-planning for the relay satellite system.The rolling window mechanism is used to adapt to the dynamic characteristic of tasks,and the hybrid particle swarm optimization algorithm is used for mission scheduling under the unit rolling window.Real satellite network scene is simulated by the STK(satellite tool kit)software,to verify the performance of the proposed algorithm.Compared with the existing scheduling algorithm,the proposed algorithm performs better in the aspects of the total profits from scheduling,unit profit from scheduling and the execution time.Secondly,the network graph model and dynamic scheduling algorithm are proposed based on on-demand scheduling.Facing with the inevitable trend that the relay satellite system is developing into satellite networking system,tasks will no longer rely only on a single relay satellite independently.In this paper the relationship between nodes in the network is represented by time-varying graph model based on the characteristics of intermittent link connectivity in multiple satellite relay transmission scenarios.After that,the storage enhanced time-varying graph(STVG)model is presented,considering the limited storage resources of each relay node.In order to realize the on-demand scheduling for tasks,a dynamic scheduling algorithm based on the shortest route of delay tolerant network(DTN)is put forward.In the proposed algorithm,each task independently selects the relay resource and finds the forwarding path based on the routing policy.The dynamic task scheduling of user aircrafts is realized,and the performance of the proposed algorithm is verified by numerical simulation.Finally,a simulation software platform of relay satellite scheduling system is constructed in this paper.The structure,the framework and function modules for the software platform are designed and implemented.The feasibility and effectiveness of the proposed model and algorithms are verified in multiple scenario demos under different running status of the relay satellite system.