Objectives  The low-orbit giant constellation is an important new infrastructure of the country, and it is also the frontier and hot spot of the current development of the aerospace field.

  Methods  In this paper, through the"cloud-edge-end"dynamic optimization collaboration of satellite clusters and ground resources, the key issues of cloud-edge-end collaborative intelligent service for low-orbit giant constellation are studied. Focusing on the goal of one satellite multi-purpose, star cluster networking, collaborative sensing, autonomous control, and intelligent service, we analyze the key technologies such as intelligent control and high-precision processing of ground center cloud, collaborative observation and fusion processing of space-based edge cloud, real-time processing and intelligent service of satellite and user.

  Results and Conclusions  Based on the characteristics of observation targets, we have formed a"cloud-edge-end"collaborative intelligent service model for real-time tasks, which can provide fast, accurate and flexible space-time continuous information service for typical real-time application scenarios, and realize the leapfrog development of intelligent earth observation.