Research On Resource Scheduling Techniques Of Medium And Low Orbit Satellite Network System Based On Reinforcement Learning

With the rapid development of the new generation of medium-low-orbit satellite Internet constellation represented by SpaceX,Starlink and O3 b,etc.,it has significant advantages such as wide-area coverage,full-time interconnection and multi-star coordination compared with the traditional geostationary earth orbit communication satellite(GEO)and it has been widely paid attention in the world in the realization of rapid resource scheduling for mass communications services.However,unlike the traditional network environment of traditional single GEO satellite resource scheduling method,medium-low-orbit satellites generally work with a combination of multiple satellites as constellations,and there are multi-satellite collaborations,joint networking,and mass users in medium-low-orbit satellite network constellation.The existing traditional satellite resource scheduling methods are difficult to meet the resource scheduling needs of the satellite constellation communication services.This paper made a further study of the communication service resource scheduling technology in the medium-loworbit satellite network system.The specific research work is as follows:Firstly,the resource scheduling method of communication services in traditional satellite network systems is studied.There are many problems of the scheduling of mass communication service resources,such as the joint optimization of multiple satellite resources,the personalized demand of mass communication service user service customization and the quality of communication service in the environment of medium-low-orbit satellites.For the above issues,IRSUP designs an intelligent scheduling module and an intelligent user service preference optimization module.The intelligent scheduling module based on reinforcement learning uses reinforcement learning to autonomously learn and update knowledge from historical scheduling experience,so as to be fitted as the optimal decision strategy for the problems of medium-loworbit satellite network resource scheduling and give a reasonable global decision-making plan.At the same time,it greatly reduces the time required for the high complexity of massive business resource allocation calculations.The intelligent user service preference optimization module can effectively adopt customized services for different user needs,avoid waste of resources on the planet and business congestion,and greatly improve user business service quality and user satisfaction.The experimental results show that IRSUP can effectively improve the rationality of resource scheduling,link resource utilization and user satisfaction,among which the business capacity is increased by 30 to 60%,and user satisfaction is increased by more than twice.Secondly,for the problem that the resource scheduling mechanism in the medium-loworbit satellite network system can schedul the daily communication service,but it cannot timely and effectively guarantee the emergency communication service scheduling,a multi-star intelligent resource scheduling mechanism named IRSE on reinforcement learning for emergency business for the requirements of emergency services.Aiming at the requirements of emergency services,For the above issues,IRSE designs an emergency business service module and optimizes an emergency service module based on reinforcement learning intelligent scheduling.The emergency service module could provide emergency services for emergency services in order to achieve the effect of processing emergency services more faster and greatly improveing the emergency service rate.The emergency service module based on reinforcement learning intelligent scheduling could autonomously learn the past scheduling experience of daily and emergency services,intelligently select the corresponding scheduling strategy for different services,and monitor the scheduling condition in the medium-low-orbit satellite network system in real time in order to maximize the overall quality of service of communication services.The simulation results show that IRSE can effectively improve the emergency service rate,the service capacity and the rationality of resource scheduling,among which the emergency service rate is basically constant at 100%,the service capacity is increased by 58% to 84%,and the rationality of resource scheduling is increased by 21 % to 32%.