Research On The Key Technologies In Mission-Driven Multi-Cluster Satellite Swarm Network

A satellite cluster(Satellite cluster)is a distributed spacecraft system composed of several satellites around a joint mission objective.Compared with constellation and formation,the distance between satellites inside the cluster is relatively close,and there are no strict requirements for satellite configuration.Since the functions of a traditional single large satellite are dispersed to several small satellites,satellite cluster has many outstanding advantages such as low construction cost and risk,reconfigurable function,and strong survivability,and has been widely concerned and studied in recent years.This paper carries out research work from three aspects: star cluster division,optimal cluster construction,and autonomous spatial networking.According to different mission requirements,a multi-cluster satellite swarm system is constructed through orbit clustering and satellite maneuvering,and a self-organizing time division multiple access networks are designed for the system.The main work and achievements of this paper are as follows:Firstly,a satellite clustering method based on orbit clustering is proposed to determine the satellite cluster.Thus,we establish the orbit distance model,obtain the orbit constraints of satellite clusters,and propose an orbit distance calculation method based on double-layer particle swarm optimization.According to the configuration of satellite clusters,we design two methods of orbit clustering and preliminarily construct the multicluster satellite swarm system using orbit clustering.The simulation results show that the two-layer particle swarm optimization algorithm has high accuracy in calculating the orbit distance and is suitable for the clustering index of orbit clustering.The two orbital clustering methods can reasonably cluster the selected orbital set according to the characteristics of orbital configuration.Secondly,we propose a construction strategy of a single satellite cluster based on a multi-satellite phase-modulation maneuver.We set up a Horman and Lambert phasemodulation mobility model on a single cluster of satellites in each orbit through multiple phase modulation motors constitutes a single satellite cluster,the phase modulation motor model to distinguish the star on the load,the motor fuel consumption,time,cost,fuel remaining three aspects put forward three kinds of mobile strategy,and use the Genetic Algorithm to solve.The simulation results show that the three strategies can construct a single star cluster according to the task requirements.Finally,a high real-time,self-organizing Time Division Multiple Access network is designed and implemented for the multi-cluster satellite swarm system — a kind of satellite cluster self-organizing network protocol is designed for network communication in a single cluster.We propose a time synchronization(MCTS)algorithm with motion compensation for the problem of time synchronization under node space motion.Then we use network simulation tool OMNe T++ to set up two simulation scenarios of ground and satellite clusters,respectively,and the self-organizing network protocol and MCTS algorithm are tested and analyzed.The simulation results show that both ground nodes and satellite clusters can achieve fast self-networking with high network reliability,and the MCTS algorithm can effectively improve the time synchronization accuracy in moving scenarios.In a word,this paper researches the classification of satellite clusters,the establishment of satellite clusters,and autonomous networking for satellite clusters’ characteristics and mission requirements.The relevant results can provide theoretical references for the construction and application of satellite clusters in the future.