Satellites will play a vital role in the future of the global Internet of Things (IoT); however, the resource shortage is the biggest limiting factor in the regional task of massiveequipment in the IoT for satellite service. Compared with the traditional isolated mode of satellite resources, the current research aims to realize resource sharing through satellite cooperation in satellite edge computing, to solve the problems of limited resources and low service quality of a single satellite. We propose a satellite resource pool architecture-oriented regional task in satellite edge computing. Different from fixed servers in ground systems, the satellite orbital motion brings challenges to the construction of the satellite resource pool. After the capacity planning of the satellite resource pool for regional tasks is given, an algorithm based on search matching is proposed to solve the dynamic satellite selection problem. A ground semi-physical simulation system is built to perform experiments and evaluate the performance of three modes of satellite resource sharing: isolated mode, cooperative mode, and pooled mode. The results show that the pooled mode, compared with the isolated mode, improves the task success rate by 19.52%, and at the same time increases network resources and energy consumption in the same scenario. Compared with the cooperation mode, the performance of task success rate and resource utilization rate is close to that of the pooled mode, but it has more advantages in response time and load balancing of satellite resources. This shows that in the IoT, the resource pool is of great benefit as it improves the task response time and improves the load balance of satellite resources without degrading the performance, which makes sense in task-demanding scenarios.

No abstract

Low earth orbit (LEO) mega-constellations have once again triggered a wave of space-based system construction. On the one hand, LEO communication, LEO navigation, LEO remote sensing constellations and so on are proposed. On the other hand, with the continuous development of software-defined satellite and intelligent satellite technology, space-based systems are developing in the direction of multi-function, integration and cross-domain integration. The whole space-based system is no longer the traditional working mode of a single functional constellation, but a genral cross-domain fusion constellation (CDFC) system for complex tasks. Like the terrestrial global Internet, the space-based system will serve as a global infrastructure for integrating communication, navigation and remote sensing, that is, the intelligent space-based system, to provide services for the global demand. The traditional method of designing constellation for a certain type of function is no longer applicable to this type of constellation design. To solve this problem, this paper proposes a design and optimization method of cross-domain fusion constellation of communication, navigation and remote sensing based on reverse design. The paper optimizes the CDFC through resource coverage. Through experiments, we prove that the number of satellites in the CDFC can be reduced by 30.60% compared with the independent and combined constellations in each domain, and the coverage and service performance of the constellation can be improved. The cost can be reduced by 18.31% compared with the combined constellation. When the same number of satellites is used, the resource coverage of the cross-domain fusion constellation is increased by at least eight times.

Low earth orbit (LEO) mega-constellation is in the stage of rapid construction, which is expected to serve more regions and apply to more scenarios, such as regional navigation enhancement, emergency communication, collaborative detection, etc. For such typical regional tasks, the construction of temporary LEO satellite clusters is an effective way to improve the quality of temporary services. However, LEO satellites move too fast, and relying on ground-based manual control can no longer meet the mission requirements. The cluster task coordination based on reinforcement learning can effectively improve the autonomous performance of the cluster, but it is difficult to apply to satellites, especially LEO satellites with extremely limited resources. Based on this, this paper proposes a lightweight collaboration method for LEO satellite constellation based on distributed reinforcement learning, which optimizes resources on the premise of ensuring the independent realization of task requirements. In this paper, the Q-learning reinforcement learning algorithm is used to compare the performance of the centralized Q-learning algorithm and distributed AC algorithm. The experiment of the star cluster cooperative detection task shows that the convergence time of the proposed algorithm is 81.5% less than that of the centralized algorithm, and it can run on embedded raspberry pi. It shows that the algorithm is expected to be applied to cluster task coordination of LEO satellites.