| 摘要: |
| 卫星网络作为国家新型基础设施建设的战略方向，面临高动态拓扑变化、星间链路时变不稳定、单星计算资源受限等关键技术挑战。针对传统星云架构中数据回传时延高、云端负载失衡等瓶颈问题，通过研究星地协同多轨道边缘计算调度架构，提出星地资源协同以及星间协同的分层式边缘计算架构。通过设计星载边缘自治以及自适应调度迁移框架，攻克星载异构资源动态管控、多模态业务自适应编排、星地计算效能协同优化等技术瓶颈，建立天地联合的边缘计算关键技术体系。研究成果为推进卫星网络云-边-端融合演进、增强空间信息基础设施实时服务能力提供了理论方法支撑与技术实践路径。 |
| 关键词: 星地协同 边缘计算 边缘自治 云原生化 微服务架构 |
| DOI：10.20079/j.issn.1001-893x.250416001 |
|
| 基金项目: |
|
| Onboard Multi-orbit Satellite-Ground Collaborative Scheduling Architecture and Key Technologies |
| LIU Yongfu,SHANG Kefeng,BIE Jia,LIU Yan,ZHU Kaige,WANG Yuwen,LI Kewei |
| (1.Southwest China Institute of Electronic Technology，Chengdu 610036，China;2.Shanghai Satellite Network Research Institute Co.，Ltd.，Shanghai 201210，China;3.CECloud Computing Technology Co.，Ltd.，Chengdu 610041，China) |
| Abstract: |
| As a strategic direction for national new infrastructure construction，satellite networks face key technical challenges such as highly dynamic topology changes，time-varying and unstable inter-satellite links，and limited computing resources of single satellite.For the bottlenecks in traditional nebula architectures，such as high data backhaul delay and unbalanced cloud load，by studying the satellite-ground collaborative multi-orbit edge computing scheduling architecture，the authors propose a hierarchical edge computing architecture featuring satellite-ground resource collaboration and on-board inter-satellite collaboration.By designing an on-board edge autonomy framework and an adaptive scheduling migration framework，technical bottlenecks are overcome including dynamic management and control of on-board heterogeneous resources，adaptive orchestration of multi-modal services，and collaborative optimization of satellite-ground computing efficiency，thus establishing a key technical system for satellite-ground integrated edge computing.The research results provide theoretical and methodological support as well as technical practice paths for promoting the cloud-edge-end integration evolution of satellite networks and enhancing the real-time service capability of spatial information infrastructure. |
| Key words: satellite-ground collaboration edge computing edge autonomy cloud-native microservices architecture |
《电讯技术》编辑部 