Research On Resource Scheduling Optimization Of Satellite Earth Fusion Network Based On Statistical Delay Guarantee

With the continuous development of satellite communication technology,the access volume and data transmission volume of satellite Internet of Things equipment continue to increase,which leads to the existence of obvious temporal and spatial imbalance,suddenness and randomness in the satellite Internet of Things business.Traditional satellite network The architecture is prone to problems such as high latency and high-energy writing.In order to meet these challenges,based on the satellite ground collaboration architecture integrating edge computing(MEC),this thesis studies the business modeling that can characterize the characteristics of satellite services and the delay model of multi satellite computing offload,the satellite ground collaboration resource allocation and computing offload algorithm based on statistical delay guarantee,and the satellite ground network simulation platform.It has important research significance for improving the efficiency and performance of Satellite Network and improving the reliability of satellite user quality of service(Qo S)guarantee.The main research work of this thesis is as follows:Firstly,this thesis proposes a satellite business modeling and analysis method based on stochastic network calculus to address the suddenness and randomness issues of satellite ground collaborative network business.The Poisson model and state Markov chain were used to characterize the arrival and channel service processes of satellite services,and the random arrival curve and random service curve of satellite services were derived;Based on the theory of stochastic network calculus and combined with the multi hop offloading workflow in satellite ground collaboration scenarios,a delay model for satellite tasks during multi hop offloading is derived;Next,simulation is conducted on the theoretical delay model to analyze the impact of different communication parameters on the multi hop offloading performance of satellites,and to verify the consistency between the theoretical performance boundary and actual simulation performance.Then,in response to the highly sensitive delay and resource management issues of satellite ground collaboration,this thesis proposes a satellite ground collaboration resource allocation and calculation offloading algorithm based on statistical delay guarantee.The statistical delay probability boundaries for local computing,access satellite computing,and multi satellite collaborative computing strategies were derived using random network calculus theory,and the delay models for the three strategies were obtained;Taking the system task completion rate as the optimization goal,the joint optimization model of resource allocation and task unloading is further constructed,and the joint problem is formulated as a mixed integer nonlinear programming problem;The resource allocation algorithm based on statistical delay guarantee is used to solve the fair resource allocation,and the deep Q-network(DQN)algorithm is used to learn the optimal unloading decision,so as to improve the reliability of task delay guarantee and the resource utilization of Satellite Network,and the simulation experiment and analysis and evaluation of unloading strategy are carried out.Finally,in response to the high cost testing problem of satellite dynamic resource allocation and offloading strategies,this thesis proposes a satellite ground network simulation platform based on STK.Build a satellite ground network simulation platform using STK and MATLAB software,and design the platform framework and functional modules based on functional requirements.The platform can simulate the process of interactive learning between satellite agents and the environment;Then,the construction and experiment of the satellite ground network simulation platform were carried out,and the initialization test of the simulation platform and the construction test of the satellite communication scene were carried out.The platform built can realize the network scene design and communication parameter adjustment of different network scale requirements,simulate multiple resource adjustment strategies,and provide multi-dimensional visual display of the Satellite Network scene.