Research On Space Information Network Resource Scheduling And Management Technology

With the development of key technologies such as satellite-to-ground communications,satellite-based data processing,and heaven-to-earth integrated networking,space information networks has now become an important part of the next-generation communication system.Resource scheduling technology,as an important means to realize the optimal allocation of space information network resources,plays a vital role in many fields,such as ground observation,emergency communications,space measurement and control,and the expansion of national strategic interests.It has become one of the main research hotspots for domestic and foreign scholars in the space field in recent years.In order to meet the growing demand for space resources,it is necessary to design more reasonable resource scheduling strategies needs to be designed according to different resource constraints to ensure that the space information network can meet the needs of as many users as possible.Through in-depth understanding and analysis of the research status of space information network resource scheduling,it is found that there is no mature architecture of space information network and resource scheduling technology is not yet perfect.Scholars from various countries have proposed different resource scheduling algorithms for different research backgrounds,resource characteristics and task requirements.This thesis studies the resource scheduling of space information networks based on constraint satisfaction,which mainly includes the following four aspects:(1)Overview of resource scheduling technology.Through the in-depth understanding of the current research status in this field,the types and reliability of network resources are analyzed,and various resource scheduling algorithms proposed in recent years are classified and summarized according to different optimization objects,and the key issues that need to be addressed in this thesis are identified.(2)Modeling and simulation of resource scheduling problems.By analyzing the characteristics of resource scheduling problems,a space information network architecture based on Low Earth Orbit(LEO)satellites and Ground Receiving Stations(GRS)was established,and STK simulation tools were used to build network model,orbit data generation,visibility analysis,and the connection relationship data between LEO satellites and ground receiving stations during the network operation.Combined with the Constraints Satisfaction Problem(CSP) model,a mathematical model of resource scheduling that satisfies the problem description is established.(3)Optimization of resource scheduling based on constraint satisfaction.According to the needs of space information network users,this thesis analyzes the optimization goals and constraints of resource scheduling,and a resource scheduling model based on constraint satisfaction is established.For the previously established space information network architecture,a task un-weighted resource scheduling algorithm based on constraint satisfaction is proposed,which optimizes the task scheduling from the aspect of task satisfaction rate.Through constraint propagation and dynamic allocation strategies,it maximizes task completion under limited resource conditions to meet user service requirements.Simulation results show that the algorithm proposed in this thesis performs well in terms of task completion and satisfaction rate,and shows good adaptability when the number of tasks is large.(4)Resource scheduling algorithm based on resource reliability.By analyzing the factors affecting resource reliability,a task weighted resource scheduling algorithm based on resource reliability is proposed,and a conflict-elimination strategy for the scheduling algorithm in the case of resource conflict is designed.The task pruning principle,the time window pruning principle,and the uniqueness priority principle are three aspects to enhance the conflict resolution capability of the scheduling algorithm,and established a revenue weight model based on the task execution revenue and resource allocation revenue.Through simulation experiments and the contrast analysis,it is found that the algorithm has obvious performance in optimizing the comprehensive income goal of the task,and has a good conflict elimination effect,it also has a significant performance in improving the task completion number.