The inter-layer connection planning strategy of the multi-layer satellite mega constellations is a key technology to guarantee reliable and rapid transmission for various traffic demands. However, satellites deployed at different orbital heights cause more complicated layer-relative motions and frequently intermittent satellite connections. How to plan inter-layer links (ILLs) to guarantee high-throughput communication and provide reliable satellite relays is extremely challenging. In this paper, we leverage the satellite regular trajectories of each layer to design a hierarchical planning architecture, which consists of two phases, critical satellite selection and dynamic ILL planning. In the first phase, critical satellites are selected by traffic load and delay requirements, and they can connect with other layers to search for efficient relays. Considering the high dynamic motion between layers, we further propose an ILL planning problem between critical satellites and relays to maximize throughput and obtain robust ILLs. To tackle the proposed problem in large spatio-temporal scales, we propose a multi-agent learning-based ILL planning strategy, which can adjust switching directions based on the orbit relative position from critical satellites to relay satellites of various layers and obtain efficient ILLs. Simulation results illustrate that the optimal ILL number is less than 1/3 of its layer scale, and the proposed strategy can enhance throughput by 14.6%, and reduce the switching rate by 60.9% compared to state-of-the-art baseline algorithms.

As the spaceborne resources are heterogeneous and the satellite network topology changes constantly, various time-varying derivative graphs are designed to represent the data acquisition and delivery process in satellite-terrestrial integrated networks (STNs). However, the time complexity of resource allocation approaches based on assorted time-varying graphs is remaining obstinately exponential. Derived from the satellite vertical coverage feature, we design a more general relationship to involve a group of nodes in a hyperedge rather than the bilateral relationship between two nodes. In this paper, we propose a time-expanded hypergraph (TEH) to contract the adjacency matrix of the network topology. Based on the proposed TEH, the problem is formulated to minimize the consumption of the communication resource in the resource-limited STN while completing the same number of tasks. Since the problem is intractable by exhaustive search, we further propose a hybrid hyperedge and Lagrangian relaxation algorithm to perform optimal resource allocation through an oriented search for the feasible hyperedges to reduce the scale of searching. The simulation results validate that the proposed algorithm can effectively complete the tasks with lower time complexity.

Satellite networks (SNs) are heterogeneous networks composed of typical functional domains, such as communication domains, observation domains, etc. Since the resources are independent among domains, and are highly dynamic with the movement of satellites, it is extremely difficult to capture the potential interaction relationship among various resources, even less to realize the coordinated scheduling of cross-domain resources in large-scale SNs. Motivated by these factors, we firstly propose a multi-aspect expanded hypergraph (MAEH) to accurately depict the Spatio-temporal features of resources in various domains as well as functional property, which is presented by different aspects. Particularly, the “aspect” can involve a group of resources with a similar feature in a hyperedge rather than a bilateral relationship between two individuals, thus the MAEH can dramatically reduce the redundant connections. By exploiting the MAEH, we model the multi-domain resource allocation problem in the form of mixed-integer linear programming to maximize the completed tasks. Through the topological nested characteristic of the MAEH, we propose a two-stage scheme to accomplish the resource allocation rapidly with lower computational complexity and to improve the resource utilization ratio efficiently. Simulation results validate that compared with the optimal performance, the executive time of the proposed scheme is average fivefold less under 4% extra communication resource cost. Besides, the resource utilization ratio improves over 15% by cross-domain collaboration.