Page 329 - 《软件学报》2026年第1期

软件学报 ISSN 1000-9825, CODEN RUXUEW E-mail: jos@iscas.ac.cn
2026,37(1):326−343 [doi: 10.13328/j.cnki.jos.007410] [CSTR: 32375.14.jos.007410] http://www.jos.org.cn
©中国科学院软件研究所版权所有. Tel: +86-10-62562563

*
卫星边缘计算智能化技术研究进展

张其阳 1 , 邢若粼 1 , 李元哲 2 , 周 傲 1 , 徐梦炜 1 , 王尚广 1

1
(网络与交换全国重点实验室 (北京邮电大学), 北京 100876)
2
(清华大学 智能产业研究院, 北京 100084)
通信作者: 王尚广, E-mail: sgwang@bupt.edu.cn

摘 要: 近年来, 随着太空任务的日益复杂化, 太空数据呈现爆炸式增长. 然而, 受限于星地链路带宽限制和稀缺的
频谱资源, 传统弯管架构在星地数据传输中遭遇瓶颈. 此外, 星上数据必须等待卫星经过地面站上空才能下传, 而
大规模建设地面站不仅成本高昂, 还面临地缘政治风险和经济收益的不确定性. 卫星边缘计算作为一种有效的解
决方案, 通过在卫星边缘引入移动边缘计算技术, 能够显著提升用户体验, 同时有效减少网络冗余流量. 在轨处理
星上原始数据不仅缩短了数据获取时间, 还减少了对地面站的依赖. 此外, 卫星边缘计算结合人工智能技术, 为应
对当前挑战提供了高效且充满潜力的解决方案. 综述卫星边缘计算智能化技术的研究现状: 首先探讨其在多个典
型场景下的需求与应用; 随后分析该领域的关键挑战和研究进展; 最后归纳若干开放性研究课题, 并提出可借鉴的
新思路. 期望通过讨论, 为推动卫星边缘计算智能化技术创新与实际应用提供有价值的参考.
关键词: 卫星边缘计算; 智能化技术; 边缘智能; 星地协同
中图法分类号: TP393

中文引用格式: 张其阳, 邢若粼, 李元哲, 周傲, 徐梦炜, 王尚广. 卫星边缘计算智能化技术研究进展. 软件学报, 2026, 37(1): 326–343.
http://www.jos.org.cn/1000-9825/7410.htm
英文引用格式: Zhang QY, Xing RL, Li YZ, Zhou A, Xu MW, Wang SG. Research Progress on Intelligent Technologies for Satellite
Edge Computing. Ruan Jian Xue Bao/Journal of Software, 2026, 37(1): 326–343 (in Chinese). http://www.jos.org.cn/1000-9825/7410.
htm

Research Progress on Intelligent Technologies for Satellite Edge Computing
1
1
2
1
1
ZHANG Qi-Yang , XING Ruo-Lin , LI Yuan-Zhe , ZHOU Ao , XU Meng-Wei , WANG Shang-Guang 1
1
(State Key Laboratory of Networking and Switching Technology (Beijing University of Posts and Telecommunications), Beijing 100876,
China)
2
(Institute for AI Industry Research, Tsinghua University, Beijing 100084, China)
Abstract: In recent years, the increasing complexity of space missions has led to an exponential growth in space-generated data. However,
limited satellites-to-ground bandwidth and scarce frequency resources pose significant challenges to traditional bent-pipe architecture, which
faces severe transmission bottlenecks. In addition, onboard data must wait for satellites to pass over ground stations before transmission.
The large-scale construction of ground stations is not only cost-prohibitive but also carries geopolitical and economic risks. Satellite edge
computing has emerged as a promising solution to these bottlenecks by integrating mobile edge computing technology into satellite edges.
This approach significantly enhances user experience and reduces redundant network traffic. By enabling onboard data processing, satellite
edge computing shortens data acquisition times and reduces reliance on extensive ground station infrastructure. Furthermore, the integration
of artificial intelligence (AI) and edge computing technologies offers an efficient and forward-looking path to address existing challenges.
This study reviews the latest progress in intelligent satellite edge computing. First, the demands and applications of satellite edge
computing in various typical scenarios are discussed. Next, key challenges and recent research advancements in this field are analyzed.

* 基金项目: 国家自然科学基金 (62032003, 62372061, 62425203, U21B2016); 中央高校基本科研业务费专项资金 (2024ZCJH11)
收稿时间: 2024-07-28; 修改时间: 2024-12-05, 2024-12-31; 采用时间: 2025-02-11; jos 在线出版时间: 2025-06-04
CNKI 网络首发时间: 2025-06-05