Study On The Efficient Management Method Of Giant Low-orbit Constellation Based On Clustering

The continuous development of satellite technology and the accumulated experience of rocket launch greatly reduce the satellite launch cost.Satellite networks evolve rapidly from low-density networks to high-density networks under the close attention of various countries around the world.Since satellite networks are not limited to geographical limitations and support long-distance communication,flexible networking,and other inherent features,globally efficient broadcast or multicast services can be provided anywhere.Today,with the growing network demand and the increasing ground network load,the satellite network can well supplement the ground stations,and cover the western region where the construction cost of ground base stations is too high to improve the network coverage in China.These features make the giant low-orbit constellation a research hotspot.In order to ensure the network expansion and maintainability of giant low-orbit constellations,this thesis introduces the hierarchical management of group and cluster in order to better manage the information of each network node.Meanwhile,a minute-level failure response system is designed to ensure the reliability of the satellite network.The main studies of this thesis are as follows:First,this graduation thesis proposes the operation and resource control method of giant low-orbit satellite network facing fast response to the integration of heaven and earth.The satellite constellation model used in this thesis is the low-orbit satellite constellation ——Starlink launched by Space X,supplemented by a double-layer hierarchical satellite constellation composed of several MEO satellites.Among them,MEO,as the management node,conducts hierarchical management of LEO satellites in the constellation Starlink based on their geographical location: group management and cluster management.LEO satellites are grouped according to the coverage of MEO,the corresponding MEO as the manager of the LEO group;due to the large number of LEO satellites in the same group,it adopts the geographical location-based method for cluster management:1.With the assistance of position service,LEO satellites in the packet are divided into N uniform initial clusters;2.Merge according to the corresponding rules.Stratified management can effectively avoid the problem of too much delay and too many interlayer links caused by too many satellite nodes.After the grouping is completed,the A~* algorithm is used to calculate the total number of hops from a single node to all other nodes in the cluster,and the node with the smallest total hops is selected as the cluster head.When the new time slot comes,the cluster head aggregate cluster information is handed over to the MEO,and the global topology information is formed by MEO arrangement.In the process of summary,there is a fault detection system and the corresponding fault response system to ensure the normal operation of constellation network.The fault response speed of this thesis can reach the minute level,which can effectively improve the robustness of constellation network.