Research On Task Cooperation Of Intelligent Satellite In Space Networked System

As an essential part of the network information system,the space networked system has unique location advantages and provides seamless coverage and continuous services for intelligent applications in various fields.As a service node of the future space networked system,the intelligent satellite will share computing,storage,communication,sensing and actuator resources through the space network,and cooperate with each other to provide flexible and accurate space intelligent information services,maximizing the utilization of resources.The space networked system will spawn a series of new paradigm of intelligent application,such as satellite Internet of Things,space cloud computing,and space-based early warning systems.How to efficiently coordinate services with intelligent satellites and how to effectively manage satellite resources has become important research of space networked system.The intelligent satellites are both resource providers and task executor.They have the characteristics of highly dynamics,limited resources,distribution,and selfishness.Therefore,there are the following urgent problems in task collaboration and resource management.Firstly,when and how users can offload tasks to which satellites to obtain the best quality of experience.Secondly,how to effectively balance the income and expenditure of intelligent satellite when sharing resources.Thirdly,what matching strategy is adopted between satellite resources and user tasks can maximize the quality of services.Given the above problems,this paper studies the problem of intelligent satellite task collaboration.An intelligent service architecture of space networked system is proposed,and integrated hardware architecture and modular layered software architecture of intelligent satellite are studied.On this basis,through the use of static game theory,differential game theory,matching game theory basis,in-depth research has been carried out in task migration,resource sharing and resource allocation.The main work and innovation are summarized as follow:(1)A task migration model for space networked system based on the static non-cooperative game is constructed,and an iterative algorithm for solving task offloading strategies under the Nash equilibrium is proposed.Compared with task offloading in other scenarios,the proposed model takes into account the characteristics of intermittent communication caused by the dynamics of satellites and introduces satellite visibility analysis and its impact on the task response delay.Simulation proves that the task migration strategy under the Nash equilibrium can optimize the processing delay of the task and the energy consumption of the satellite.(2)An intelligent satellite resource sharing model based on the differential game is constructed,and an optimal division strategy of intelligent satellite shared resource is proposed.Compared with the traditional static resource sharing model,the proposed model tasks into account the dynamics of the satellite orbit,the dynamics of requirements of the space resource pool,and the dynamics of interaction between different satellites.Differential game theory is used to solve the optimal shared resource division in real time.Simulation proves that the proposed division strategy of intelligent satellite resource sharing can effectively balance the income and expenditure of intelligent satellite in resource sharing,incentivize satellites to share resources,and improve the revenue of resource sharing.(3)A resource allocation model of space networked system based on the matching game is constructed,and an improved iterative algorithm based on “reject-receive” is proposed to solve the Pareto optimal resource allocation strategy.Since there are many types of satellite resource,the proposed model achieves joint matching of multiple resources.During the matching process,user satisfaction and intelligent satellite revenue are combined as user cost performance as an optimization goal,which effectively improves the overall quality of service of space networked system.Besides,for the “reject-receive” algorithm needs to preset the preference value between the user task and the space resource,which will result in a decrease in resource allocation performance.By introducing cooperation between users,an iterative algorithm for Pareto optimal resource allocation strategy is proposed.The simulation and ground experiments prove that the algorithm can effectively improve the cost performance of users and maximize the quality of service of the space networked system.