Research On Key Technologies Of Wireless Resource Control And Simulation In Low-orbit Satellite Cooperative Networ

With the development of wireless communication technology,the popularization of intelligent terminals and the diversification of applications,the application scenarios of mobile communication networks are becoming more and more diverse.However,relying solely on the terrestrial mobile communication network cannot meet the beautiful vision of "global coverage and scene intelligence" in 6G.The satellite network has the outstanding characteristics of global wide area coverage,flexible networking and strong resistance to destruction,which has obvious advantages for providing communication services in the sea,remote areas that are difficult to be covered by ground network and disaster relief scenarios.However,compared with high-orbit satellite communication networks,the characteristics of low-orbit satellite communication networks with diverse types of massive services,limited radio resources,complex and changeable channel environment and high-speed dynamic change of topology are bound to impose a heavy burden on the dynamic management of radio resources.Aiming at the requirements of efficient and dynamic management of radio resources in low-orbit satellite networks,this thesis focuses on network traffic load balancing,radio resource allocation and network simulation mapping.It focuses on solving the problems of "low-latency and highreliability transmission of massive bursty non-uniform traffic","flexible and adaptive allocation of multi-dimensional resources under high dynamic topology" and "cross-platform joint simulation of large-scale complex networks".The main research contents and innovation points of this thesis include:(1)To address the network congestion problem caused by limited onboard processing capacity and instantaneous bursts of massive traffic,we extract traffic-related features using the temporal bursts of satellite network traffic,and propose a network traffic classification algorithm based on temporal discrete affiliation to effectively reduce the dependency among traffic features.Based on the traffic classification results to predict the resource consumption of processing each network traffic,a preemptive optimal scheduling algorithm based on queueing message prediction is proposed to reduce the resource contention between traffic flows.The results show that the proposed algorithm can improve the classification accuracy by 2.9%and reduce the packet loss rate by 1.2%compared with the time-continuous affiliation-based network traffic classification algorithm.(2)To address the problem of limited available resources for cooperative transmission due to large differences in beam channels in highly dynamic satellite network topologies,the influence factors of satellite channels are analyzed in depth,an attraction utility function based on channel state information is proposed,and an optimal radio resource evaluation model is established.The coupling between service demand and optimal resources is analyzed in depth,and the resource fragmentation optimization degree coefficient is introduced to propose an adaptive radio resource allocation algorithm based on the bidirectional demand of users and network to reduce the degree of resource fragmentation and to improve resource utilization.Simulation results show that the proposed algorithm can improve the system throughput by 8.7%and the user satisfaction index by 6.2%compared with the radio resource allocation algorithm that only considers the unidirectional demand.(3)To address the problem that the existing satellite communication system simulation lacks joint simulation of constellations and protocols for multiple constellations and multiple systems,we deeply analyze the independence within layers and correlation between layers in the integration scenario of communication,simulation and computation,establish a hypergraph hierarchical nested kriging model,and propose a local flexible network mapping algorithm based on basis functions to realize a unified characterization of multilayer networks.By setting a list of resource preferences for non-dominated layers and designing adaptation sharing coefficients,a multi-objective joint optimal resource allocation algorithm based on hierarchical preferences is proposed to avoid falling into a local optimum.By designing the interface between satellite constellation simulation and network protocol simulation,a constellation and protocol co-simulation platform is built based on parallel architecture,which effectively improves the simulation speed.The simulation results show that the proposed algorithm can improve the resource utilization by 3.7%and reduce the convergence time by 5.3%compared with other network mapping algorithms based on hierarchical relationships.