A routing decision method and device for a multi-orbit satellite and a storage medium

Abstract

The application discloses a routing decision method and device of a multi-orbit satellite and a storage medium, and comprises the following steps: constructing a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation respectively, the directed edges of the multi-layer graph structure are used for indicating the links between the satellites, and any layer graph structure of the multi-layer graph structure corresponds to one satellite orbit of the satellite constellation; acquiring satellite state information of the satellites and link state information of the corresponding links, and constructing graph data corresponding to the multi-layer graph structure; and determining an optimal node according to the graph data, so as to obtain an optimal routing path for data transmission.

Description

Technical Field

[0001] This application relates to the field of satellite communication technology, and in particular to a routing decision method, apparatus, and storage medium for multi-orbit satellites. Background Technology

[0002] Low-Earth orbit (LEO) satellite optical networks, with their advantages of wide-area coverage and low transmission latency, have become a core carrier for achieving seamless global communication. With the rapid development of satellite internet, multi-orbit satellite collaborative networking has become an important trend to meet the needs of wide-area communication, and its routing design needs to adapt to the large-scale, highly dynamic communication scenarios of multi-orbit collaborative coverage.

[0003] Furthermore, in multi-orbit coverage scenarios, due to the asynchronous motion characteristics of satellites in different orbits, the connectivity of communication links between satellites in different orbits will change dynamically, leading to an increase in the complexity of route planning.

[0004] However, existing technologies mostly focus on satellite routing planning within a single orbit, which can only achieve communication coverage in a local area and cannot meet the wide-area routing requirements for multi-orbit collaborative coverage.

[0005] There is currently no effective solution to the technical problem of high routing complexity caused by the asynchronous movement of multiple orbital satellites in the existing technologies mentioned above. Summary of the Invention

[0006] The embodiments of this disclosure provide a routing decision method, apparatus, and storage medium for multi-orbit satellites, to at least solve the technical problem of high routing complexity caused by the asynchronous movement of multi-orbit satellites in the prior art.

[0007] According to one aspect of the present disclosure, a routing decision method for multi-orbit satellites is provided, comprising: constructing a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate links between satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; acquiring satellite state information of the satellites and link state information of the corresponding links, constructing graph data corresponding to the multi-layer graph structure; and determining the optimal node based on the graph data to obtain the optimal routing path for data transmission.

[0008] According to another aspect of the present disclosure, a storage medium is also provided, the storage medium including a stored program, wherein, when the program is executed, a processor performs any of the methods described above.

[0009] According to another aspect of the present disclosure, a routing decision-making device for multi-orbit satellites is also provided, comprising: a multi-layer graph structure construction module, used to construct a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate links between satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; a graph data construction module, used to acquire satellite status information of the satellites and link status information of the corresponding links, and construct graph data corresponding to the multi-layer graph structure; and an optimal routing path determination module, used to determine the optimal node based on the graph data, and obtain the optimal routing path for data transmission.

[0010] According to another aspect of the present disclosure, a routing decision apparatus for multi-orbit satellites is also provided, comprising: a processor; and a memory connected to the processor, configured to provide the processor with instructions for processing the following steps: constructing a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate links between satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; acquiring satellite status information of the satellites and link status information of the corresponding links, constructing graph data corresponding to the multi-layer graph structure; and determining the optimal node based on the graph data to obtain the optimal routing path for data transmission.

[0011] The routing decision method for multi-orbit satellites provided in this application achieves coordinated routing across multiple satellite orbits by constructing a multi-layer graph structure corresponding to the multiple satellite orbits of a satellite constellation. This ensures the coverage of wide-area communication.

[0012] Furthermore, the constantly changing relative positions of satellites in various orbits lead to frequent disconnections and reconstructions of inter-orbit links, resulting in high complexity in routing decisions. The method provided in this application can avoid frequent fluctuations in the satellite constellation topology caused by the asynchronous motion of satellites in multiple orbits by dividing the topology into steady-state time intervals. This avoids frequent reconstruction of the multi-layered graph structure and reduces the complexity of routing decisions. Therefore, it solves the technical problem of high routing complexity caused by the asynchronous motion of multiple satellites in various orbits in existing technologies. Attached Figure Description

[0013] The accompanying drawings, which are included to provide a further understanding of this disclosure and form part of this application, illustrate exemplary embodiments of this disclosure and are used to explain this disclosure, but do not constitute an undue limitation of this disclosure. In the drawings: Figure 1A This is a schematic diagram of multiple satellite orbits according to Embodiment 1 of this disclosure; Figure 1BThis is a schematic diagram illustrating the data transmission relationship between the terminal device, satellite, and gateway station according to Embodiment 1 of this disclosure; Figure 2A This is a hardware structure block diagram of a satellite according to Embodiment 1 of this disclosure; Figure 2B This is a hardware structure block diagram of the gateway station according to Embodiment 1 of this disclosure; Figure 3 This is a flowchart illustrating the routing decision method for multi-orbit satellites according to Embodiment 1 of this disclosure; Figure 4 This is a schematic diagram of the multilayer diagram structure according to Embodiment 1 of this disclosure; Figure 5 This is a schematic diagram of the reachability graph structure according to Embodiment 1 of this disclosure; Figure 6 This is a schematic diagram of message passing in a graph neural network according to Embodiment 1 of this disclosure; Figure 7 This is a schematic diagram of a multi-orbit satellite routing decision device according to Embodiment 2 of this disclosure; and Figure 8 This is a schematic diagram of a multi-orbit satellite routing decision device according to Embodiment 3 of this disclosure. Detailed Implementation

[0014] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] Example 1

[0017] According to this embodiment, a method embodiment for routing decision-making of multi-orbit satellites is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0018] Figure 1A A schematic diagram of multiple satellite orbits according to Embodiment 1 of this disclosure is shown. Figure 1B A schematic diagram illustrating the data transmission relationship between the terminal device, satellite, and gateway station according to Embodiment 1 of this disclosure is shown.

[0019] refer to Figure 1A and Figure 1B As shown, multiple satellite orbits Orbiting the Earth, and in various satellite orbits Each of them has multiple satellites operating on it (10).

[0020] To achieve communication coverage for multiple terminal devices 20 on the ground, multiple satellite orbits are required. Multiple satellites 10 operating on the platform transmit data, thereby enabling communication services between different terminal devices 20. However, due to the multiple satellite orbits... The asynchronous motion of multiple satellites 10 in operation causes dynamic changes in the connectivity between satellites 10 in different orbits.

[0021] Therefore, it needs to be compatible with multiple satellite orbits. The routing decision method for multiple satellites 10 running on the ground is used, and the routing path for data transmission is managed in a unified manner by the gateway station 30, thereby achieving communication coverage for multiple terminal devices 20 on the ground.

[0022] Figure 2A Further shown Figure 1B A schematic diagram of the hardware architecture of Zhongwei Satellite 10. (Reference) Figure 2A As shown, satellite 10 includes an integrated electronic system, which comprises a processor, a memory, a bus management module, and a communication interface. The memory is connected to the processor, allowing the processor to access the memory, read program instructions stored in the memory, read data from the memory, or write data to the memory. The bus management module is connected to the processor and also to a bus such as a CAN bus. Thus, the processor can communicate with onboard peripherals connected to the bus through the bus managed by the bus management module. Furthermore, the processor also communicates with devices such as cameras, star sensors, telemetry and command transponders, and data transmission equipment via the communication interface. Those skilled in the art will understand that… Figure 2AThe structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, satellite 10 may also include components that are more... Figure 2A The more or fewer components shown, or having the same Figure 2A The different configurations shown.

[0023] It is worth noting that the aforementioned spaceborne peripherals connected to the CAN bus can be one or more. These spaceborne peripherals include, but are not limited to, GNSS modules, fiber optic gyroscopes, and high-torque flywheels. Further details will not be elaborated upon here.

[0024] Figure 2B Further shown Figure 1B A schematic diagram of the hardware architecture of CITIC Gateway 30. (Reference) Figure 2B As shown, the gateway station 30 may include one or more processors (processors may include, but are not limited to, microprocessors such as MCUs or programmable logic devices such as FPGAs), a memory for storing data, a transmission device for communication functions, and an input / output interface. The memory, transmission device, and input / output interface are connected to the processor via a bus. In addition, it may also include a display, keyboard, and cursor control device connected to the input / output interface. Those skilled in the art will understand that... Figure 2B The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, the gateway station 30 may also include... Figure 2B The more or fewer components shown, or having the same Figure 2B The different configurations shown.

[0025] It should be noted that, Figure 2A and Figure 2B One or more processors and / or other data processing circuits shown herein may generally be referred to as "data processing circuitry". This data processing circuitry may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuitry may be a single, independent processing module, or may be integrated, in whole or in part, into any other element in a computing device. As involved in embodiments of this disclosure, the data processing circuitry serves as processor control (e.g., selection of a variable resistor termination path connected to an interface).

[0026] Figure 2A and Figure 2BThe memory shown can be used to store software programs and modules for application software, such as the program instruction / data storage device corresponding to the multi-orbit satellite routing decision method in this embodiment of the present disclosure. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the multi-orbit satellite routing decision method of the aforementioned application. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory.

[0027] It should be noted here that, in some optional embodiments, the above... Figure 2A and Figure 2B The device shown may include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that... Figure 2A and Figure 2B This is only one instance of a specific particular instance, and is intended to illustrate the types of components that may exist in the aforementioned devices.

[0028] Under the aforementioned operating environment, according to the first aspect of this embodiment, a routing decision method for multi-orbit satellites is provided. This method comprises... Figure 2B The gateway station 30 shown is implemented. Figure 3 A flowchart illustrating the method is shown below. (Refer to...) Figure 3 As shown, the method includes: S302: Construct a multi-layer graph structure corresponding to the satellite constellation, where the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate the links between satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation. S304: Obtain satellite status information and corresponding link status information, and construct graph data corresponding to the multi-layer graph structure; and S306: Based on the graph data, determine the optimal node and obtain the optimal routing path for data transmission.

[0029] Specifically, Figure 4 A schematic diagram of a multilayer graph structure according to Embodiment 1 of this disclosure is shown. (Reference) Figure 1A and Figure 4 As shown, firstly, gateway station 30 establishes connections with multiple satellite orbits. Corresponding multi-layer graph structure Each layer of the graph structure All are based on this layer diagram structure Corresponding satellite orbit Multiple satellites 10 running on the network are nodes, and bidirectional directed edges are established between nodes corresponding to adjacent satellites 10 that can establish links. .in, This indicates the number of satellite orbits in the satellite constellation (corresponding to step S302).

[0030] For example, graph structure Corresponding to satellite orbit , graph structure nodes Corresponding to satellite orbit running Ten satellites are connected, and bidirectional directed edges are constructed between these ten satellites that can establish links. and .in, Indicates satellite orbit The number of satellites running on the platform. Indicates that by node To the node The directed edge, Indicates that by node To the node The directed edges. And, .

[0031] For example, graph structures Corresponding to satellite orbit , graph structure nodes Corresponding to satellite orbit running Ten satellites are connected, and bidirectional directed edges are constructed between these ten satellites that can establish links. and .in, Indicates satellite orbit The number of satellites running on the platform. Indicates that by node To the node The directed edge, Indicates that by node To the node The directed edges. And, .

[0032] And so on.

[0033] Graph structure Corresponding to satellite orbit , graph structure nodes Corresponding to satellite orbit running Ten satellites are connected, and bidirectional directed edges are constructed between these ten satellites that can establish links. and .in, Indicates satellite orbit The number of satellites running on the platform. Indicates that by node To the node The directed edge, Indicates that by node To the node The directed edges. And, , .

[0034] Furthermore, since the visible relationship and link conditions between satellites 10 in adjacent satellite orbits can be calculated according to the satellite ephemeris information, bidirectional directed edges can be dynamically constructed between nodes in adjacent graph structures.

[0035] For example, satellite orbit The running of the first 10 satellites and their orbits The running of the first Links can be established between 10 satellites, corresponding to a graph structure. nodes With graph structure nodes Capable of constructing bidirectional directed edges and .in, Indicates that by node To the node The directed edge, Indicates that by node To the node The directed edges. And, , Indicates satellite orbit The number of satellites running on the platform. Indicates satellite orbit The number of satellites running on the platform.

[0036] Furthermore, the satellite orbit is set. and satellite orbit For adjacent satellite orbits, graph structure Nodes and graph structure The nodes can be dynamically configured with bidirectional directed edges. This allows for the implementation of various satellite orbits. Complete connectivity between them enhances the flexibility and coverage integrity of routing paths.

[0037] Then, gateway station 30 acquires the satellite status information and corresponding link status information of satellite 10, and constructs a multi-layer graph structure. The corresponding graph data (corresponding to step S304). The specific operations for obtaining satellite status information and link status information, as well as constructing graph data, will be explained in detail later.

[0038] Finally, based on the graph data, gateway station 30 determines the optimal node and obtains the optimal routing path for data transmission (corresponding to step S306). The specific operations for determining the optimal node and the optimal routing path will be explained in detail later.

[0039] Therefore, the method provided in this application achieves collaborative routing across multiple satellite orbits by constructing a multi-layered graph structure corresponding to the multiple satellite orbits of a satellite constellation. This ensures wide-area communication coverage and adapts to dynamic and complex links between satellites operating asynchronously on multiple satellite orbits. Furthermore, it reduces the complexity of routing decisions.

[0040] Furthermore, this application also establishes a topologically stable time interval. Based on the satellite ephemeris, the motion trajectories, position coordinates, and relative motion relationships of satellites 10 operating in various satellite orbits are deduced, thus determining a time window. Within this time window, the relative position changes between satellites 10 operating in adjacent satellite orbits are relatively small, allowing for the establishment of stable link relationships, thereby yielding the topologically stable time interval.

[0041] Therefore, within a single topological steady-state time interval, the intra-layer and inter-layer graph structures of the multi-layer graph structure are fixed. Thus, the multi-layer graph structure can be reused for training and computation of graph neural networks. This avoids the problem of frequent fluctuations in satellite constellation topology caused by the asynchronous motion of satellites 10 operating in multiple orbits. Furthermore, it avoids the high complexity of routing decisions caused by frequent topology updates.

[0042] Furthermore, when the relative positions of satellites 10 operating in adjacent satellite orbits change significantly (exceeding a preset threshold), the pairing relationship is switched. Based on the satellite ephemeris, the link relationships that can be established between satellites 10 operating in adjacent satellite orbits are re-inferred, thereby forming a new topological steady-state time interval and a corresponding multi-layer graph structure.

[0043] This ensures that the topology of the satellite constellation always matches the actual motion state of the satellite 10, and reduces the computational complexity of the dynamic topology.

[0044] As described in the background section, low-Earth orbit (LEO) satellite optical networks have become a core carrier for achieving seamless global communication due to their advantages such as wide-area coverage and low transmission latency. With the rapid development of satellite internet, multi-orbit satellite collaborative networking has become an important trend to meet the needs of wide-area communication, and its routing design needs to be adapted to the large-scale, highly dynamic communication scenarios of multi-orbit collaborative coverage.

[0045] Furthermore, in multi-orbit coverage scenarios, due to the asynchronous motion characteristics of satellites in different orbits, the connectivity of communication links between satellites in different orbits will change dynamically, leading to an increase in the complexity of route planning.

[0046] However, existing technologies mostly focus on satellite routing planning within a single orbit, which can only achieve communication coverage in a local area and cannot meet the wide-area routing requirements for multi-orbit collaborative coverage.

[0047] In view of this, this application provides a routing decision method for multi-orbit satellites. This method achieves coordinated routing across multiple satellite orbits by constructing a multi-layered graph structure corresponding to the multiple satellite orbits of a satellite constellation. This ensures the coverage of wide-area communication.

[0048] Furthermore, the constantly changing relative positions of satellites in various orbits lead to frequent disconnections and reconstructions of inter-orbit links, resulting in high complexity in routing decisions. The method provided in this application can avoid frequent fluctuations in the satellite constellation topology caused by the asynchronous motion of satellites in multiple orbits by dividing the topology into steady-state time intervals. This avoids frequent reconstruction of the multi-layered graph structure and reduces the complexity of routing decisions. Therefore, it solves the technical problem of high routing complexity caused by the asynchronous motion of multiple satellites in various orbits in existing technologies.

[0049] Optionally, the operation of acquiring satellite status information and corresponding link status information, and constructing graph data corresponding to the multi-layer graph structure, includes: acquiring satellite transmit load rate, satellite receive load rate, satellite transmit remaining bandwidth, satellite receive remaining bandwidth, and satellite resource utilization rate as satellite status information to determine node attributes; acquiring one-way transmission delay, one-way maximum transmission bandwidth, one-way load rate, and remaining time coefficient of the link as link status information to determine directed edge attributes, wherein the remaining time coefficient is related to the one-way remaining connection time of the link; and constructing graph data corresponding to the multi-layer graph structure based on node attributes and directed edge attributes.

[0050] Specifically, the gateway station 30 acquires the satellite transmit load rate, satellite receive load rate, remaining satellite transmit bandwidth, remaining satellite receive bandwidth, and satellite resource utilization rate of each satellite 10 as satellite status information for each satellite 10. Among these, the satellite transmit load rate indicates the service busyness of the current transmit direction (i.e., satellite 10 as the transmitter); the satellite receive load rate indicates the service busyness of the current receive direction (i.e., satellite 10 as the receiver); the remaining satellite transmit bandwidth indicates the remaining bandwidth in the current transmit direction; the remaining satellite receive bandwidth indicates the remaining bandwidth in the current receive direction; and the satellite resource utilization rate indicates the overall resource utilization of the satellite 10, without distinguishing between the transmitter and receiver.

[0051] Then, the gateway station 30 determines the node attributes of the corresponding nodes based on the satellite status information of each satellite 10. For example, it obtains the satellite orbits. The satellite status information of satellite 10 running on the platform is used to determine the corresponding graph structure. nodes Node attributes .in, Indicates satellite orbit The number of satellites running on the platform.

[0052] Simultaneously, gateway station 30 acquires the link's one-way transmission delay, maximum one-way transmission bandwidth, one-way load rate, and remaining time coefficient as link status information. Among these, the one-way transmission delay represents the one-way signal transmission delay of the link; the maximum one-way transmission bandwidth represents the upper limit of the link's maximum one-way transmission bandwidth; and the one-way load rate represents the link's one-way traffic load rate, i.e., the current traffic occupancy level. Furthermore, the remaining time coefficient... The calculation formula is as follows: (1) in, Indicates the remaining time coefficient of the link; This represents the one-way remaining connectivity duration of the link, calculated from the satellite ephemeris. This indicates that the time-scale hyperparameters are obtained by fitting historical data of satellite orbits, and are not set based on human experience.

[0053] Furthermore, the remaining time coefficient The one-way residual connectivity time of links between 10 satellites operating in the same orbit. That is, the one-way remaining connectivity duration When it approaches infinity, the remaining time coefficient Approaching 1, that is When the link between satellites 10 operating in adjacent satellite orbits is about to break, i.e., the remaining one-way connectivity time... At that time, the remaining time coefficient Approaching 0 and always greater than 0, that is This avoids setting directed edges to null, preventing them from becoming invalid in routing decisions. Furthermore, it ensures the integrity of the graph structure, preventing any impact on the accuracy of routing decisions.

[0054] Then, gateway station 30 determines the directed edge attributes of the corresponding directed edges based on the link state information. For example, obtaining satellite orbits. The link status information between the 10 running satellites is used to determine the corresponding graph structure. nodes Directed edges between and Directed edge properties and .in, Indicates satellite orbit The number of satellites running on the platform (number of satellites 10). Indicates that by node To the node The directed edge, Indicates that by node To the node The directed edges. And, .

[0055] Similarly, obtaining satellite orbits Satellite 10 and its orbit The link status information between the 10 running satellites is used to determine the corresponding graph structure. nodes With graph structure nodes Directed edges between and Directed edge properties and .in, Indicates that by node To the node The directed edge, Indicates that by node To the node The directed edges. And, , Indicates satellite orbit The number of satellites running on the platform. Indicates satellite orbit The number of satellites running on the platform.

[0056] Finally, based on the aforementioned node attributes and directed edge attributes, gateway 30 constructs a multi-layer graph structure. The corresponding graph data.

[0057] Optionally, the operation of determining the optimal node based on graph data includes: determining reachable nodes and reachable directed edges based on the starting and destination nodes used for data transmission; obtaining reachable graph data corresponding to the reachable nodes and reachable directed edges based on the graph data; using a graph neural network to perform message passing on the reachable graph data to determine the graph features corresponding to the reachable graph data; determining the standardized score of the reachable directed edges; and determining the optimal node based on the standardized score using a greedy strategy.

[0058] Specifically, Figure 5 A schematic diagram of an reachability graph structure according to Embodiment 1 of this disclosure is shown. (See reference...) Figure 4 and Figure 5 As shown, firstly, for an end-to-end (i.e., from the sending terminal device 20 to the receiving terminal device 20) routing request, the gateway station 30 determines the corresponding originating node. and destination node .

[0059] In such Figure 4 In the current graph structure shown, gateway station 30 performs a directed reachability search to filter out nodes that can reach the starting node. Reach the destination node along the directed edge The nodes and directed edges are defined, and nodes connected to the starting node are removed. and destination node Unrelated nodes and directed edges. Thus, we obtain the connection to the starting node. and destination node The corresponding reachability graph structure. Furthermore, by reducing the number of nodes and directed edges that need to be involved in computation, the complexity of routing decisions can be reduced.

[0060] Then, based on the graph data, the gateway station 30 determines the reachable graph data corresponding to the reachable nodes and reachable directed edges.

[0061] Furthermore, the gateway station 30 utilizes a graph neural network to perform message passing on the reachable graph data and determine the graph features corresponding to the reachable graph data. The specific operations of message passing will be explained in detail later.

[0062] Finally, the standardized scores of reachable directed edges are determined. Then, based on these standardized scores, a greedy strategy is used to determine the optimal node. The specific operations for determining the standardized scores and the optimal node will be explained in detail later.

[0063] It should be noted that the above operations of message passing, standardization score determination, and optimal node determination are all performed within the reach graph structure.

[0064] Optionally, using a graph neural network, message passing is performed on the reachable graph data corresponding to reachable nodes and reachable directed edges to determine the graph features corresponding to the reachable graph data. This includes: obtaining the directed edge attributes of reachable directed edges and the corresponding source and target node attributes based on the reachable graph data; using the message passing layer of the graph neural network, determining the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target nodes based on the directed edge attributes of the reachable directed edges and the corresponding source and target node attributes; iteratively updating the directed edge features and node features using multiple message passing layers of the graph neural network; and determining the graph features corresponding to the reachable graph data based on the directed edge features and node features.

[0065] Specifically, firstly, the gateway station 30 obtains the directed edge attributes of the reachable directed edges and the corresponding source and target node attributes based on the reachable graph data.

[0066] For example, for the source node With the target node reachable directed edges between Get the source node Node attributes Target node Node attributes and reachable directed edges Directed edge properties .

[0067] In graph neural networks, information is passed through a single information passing layer to obtain directed edge features. and node features The specific operations for message passing using a single information passing layer will be explained in detail later.

[0068] Figure 6 A schematic diagram of message passing in a graph neural network according to Embodiment 1 of this disclosure is shown. (Reference) Figure 6 As shown, the information transmission process in a graph neural network, utilizing multiple information transmission layers, is as follows: First, the process of message passing using the first message passing layer is consistent with the process of message passing using a single message passing layer, resulting in reachable directed edges. First directed edge feature and target node First node features .

[0069] Second, the first directed edge features obtained through the first message passing layer. and the first node features The input is passed to the second message passing layer to obtain reachable directed edges. Second directed edge feature and target node Second node features .

[0070] Third, the second directed edge feature Second node features The message is then passed to the next message passing layer after the second message passing layer.

[0071] Similarly, through multiple message-passing layers of a graph neural network, reachable directed edges are... Directed edge characteristics and target node The node features are iteratively updated to output the final directed edge features. and node features Thus, the graph features corresponding to the reachable graph data are determined.

[0072] Optionally, using the message-passing layer of a graph neural network, the operation of determining the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target node based on the directed edge attributes of the reachable directed edges and the corresponding node attributes of the source and target nodes includes: determining a fusion vector based on the directed edge attributes of the reachable directed edges and the corresponding node attributes of the source and target nodes; using the message-passing layer and activation function, obtaining the directed edge features and the attention scores corresponding to the reachable directed edges based on the fusion vector; normalizing the attention scores of the incoming edges corresponding to the target node to obtain attention weights; obtaining aggregate information based on the attention weights and the node attributes of the source nodes corresponding to the incoming edges; and using the message-passing layer and activation function, obtaining the node features of the target node based on the aggregate information and the node attributes of the target node.

[0073] Specifically, in a graph neural network, the information transmission process using a single information transmission layer is as follows: First, merge source nodes Node attributes Target node Node attributes and reachable directed edges Directed edge properties This yields a multidimensional fusion vector. .

[0074] Second, the fusion vector The input is fed into a single message-passing layer, and an activation function is used to obtain reachable directed edges. First directed edge feature The activation function can be, for example, the ReLU function (Rectified Linear Unit). First directed edge feature. The specific calculation method is as follows: (2) in, Represents a reachable directed edge The first directed edge characteristic, This represents the activation function. This represents the fusion vector. Furthermore, This represents the weight matrix of a graph neural network. The bias vector and weight matrix of a graph neural network are represented by these values. and bias vector The learnable parameters corresponding to a graph neural network.

[0075] Third, the fusion vector Input is fed into a two-layer MLP (Multilayer Perceptron), and output can reach directed edges. Attention score The specific formula is as follows: (3) in, Represents a reachable directed edge Attention score, This represents the fusion vector. express Activation function (Leaky Rectified Linear Unit). This represents the second weight matrix of the graph neural network. This represents the second bias vector and the second weight matrix of the graph neural network. Second bias vector It corresponds to Learnable parameters of a graph neural network with activation functions. express Activation function (Rectified Linear Unit). This represents the first weight matrix of the graph neural network. This represents the first bias vector and the first weight matrix of the graph neural network. and the first bias vector It corresponds to Learnable parameters of a graph neural network with activation functions.

[0076] Fourth, based on reachable directed edges target node For all incoming edges, determine the attention score for each incoming edge. Then, the attention weights corresponding to each incoming edge are obtained by normalizing using the Softmax function. .

[0077] Among them, the target node All corresponding incoming edges represent the distance from adjacent reachable nodes to the target node. reachable directed edge Adjacent reachable nodes represent nodes that can reach the target node. Establish reachable nodes with bidirectional directed edges. Furthermore, the target node... Construct a set of adjacent reachable nodes for each of the corresponding adjacent reachable nodes. Set of adjacent reachable nodes Including source node .

[0078] For example, for reachable directed edges Its corresponding target node is The source node is At this time, the target node Corresponding reachable directed edges For the target node The incoming edges. Among them, the source node... For example, it could be an adjacent reachable node. , , wait.

[0079] Fifth, based on the attention weights corresponding to each incoming edge. and source node Node attributes Perform weighted aggregation to obtain the target node. Aggregated information The specific formula is as follows: (4) in, Represents the target node Aggregated information, Represents the target node The attention weights corresponding to each incoming edge, Indicates the source node Node attributes, Represents the target node The corresponding set of adjacent reachable nodes, Indicates the source node , Represents the target node .

[0080] Sixth, target node Node attributes and aggregated information After fusion, the data is input into a single information delivery layer, and an activation function is used to obtain the target node. First node features The specific formula is as follows: (5) in, Represents the target node The first node feature, This represents the activation function. Represents the target node Node attributes, Represents the target node Aggregated information, This represents the third weight matrix of the graph neural network. This represents the third bias vector of the graph neural network.

[0081] Thus, the first directed edge feature is obtained by message passing through a single information passing layer of the graph neural network. and the first node features .

[0082] It is important to note that when using a single message passing layer for message passing, the first directed edge feature obtained is... and the first node features This refers to the final directed edge characteristics. and node features .

[0083] When multiple message passing layers are used for information transmission, each message passing layer performs the above steps and reuses the node features and directed edge features of the previous message passing layer. For example, the second message passing layer utilizes the first directed edge feature of the first message passing layer. and the first node features Output the second directed edge feature Second node features and the obtained second directed edge features Second node features Input to the next message passing layer after the second message passing layer.

[0084] Optionally, the operation of determining the standardized score of reachable directed edges includes: taking the reachable node as the source node, determining the outgoing edge corresponding to the reachable node; and obtaining the standardized score corresponding to the outgoing edge based on the characteristics of the directed edge.

[0085] Specifically, the information gateway 30 is based on the characteristics of directed edges. Calculate the standardized scores of all outgoing edges corresponding to each reachable node.

[0086] First, determine the reachable directed edges with each reachable node as the source node, which will be the outgoing edges corresponding to each reachable node.

[0087] If a reachable node has only one outgoing edge, the normalized score of that outgoing edge is set to 1. That is, data transmission must occur from that outgoing edge.

[0088] When a reachable node has multiple outgoing edges, the directed edge features corresponding to each outgoing edge are input into the shared fully connected layer of the graph neural network, outputting the raw score for each outgoing edge. Then, the Softmax function is used to normalize the raw score for each outgoing edge, outputting the standardized score for each outgoing edge. The standardized score ranges from 0 to 1.

[0089] Furthermore, the higher the standardization score, the higher the priority of the outgoing edge corresponding to that standardization score, and the greater the likelihood that it will become the link for the next data transmission step.

[0090] In other words, at the starting node and destination node In the corresponding reachability graph structure, each directed edge can be assigned a unique standardized score. Furthermore, by sorting the standardized scores of all outgoing edges from the same reachable node, the next data transmission link is selected.

[0091] Optionally, the operation of determining the optimal node based on the standardized score using a greedy strategy includes: selecting the outgoing edge with the highest standardized score for data transmission based on the standardized score; taking the target node corresponding to the outgoing edge as the next node, determining the next outgoing edge corresponding to the next node, and determining the standardized score corresponding to the next outgoing edge, selecting the next outgoing edge with the highest standardized score for data transmission; and iterating through the reachable nodes to determine the optimal reachable node until the data is transmitted to the destination node, thereby obtaining the optimal routing path for data transmission.

[0092] Specifically, firstly, the gateway station 30 selects the outgoing edge with the highest standardized score for the next data transmission based on the standardized scores of the outgoing edges of each reachable node.

[0093] Then, taking the target node corresponding to the outgoing edge as the next node, determine the next outgoing edge corresponding to the next node. Furthermore, determine the standardized score corresponding to the next outgoing edge, and then select the next outgoing edge with the highest standardized score for data transmission.

[0094] For example, reachable nodes As the source node, with reachable nodes As the target node, the reachable node at this time The corresponding outgoing edges are directed edges. When there is a directed edge The standardized score at reachable nodes When the standardized score of all outgoing edges is the maximum, select the outgoing edge. This is the first step in data transmission.

[0095] Then, after completing the first step of data transmission, the target node of the first step... As the source node (i.e., the next node) for the second step of data transmission, and to determine reachable nodes. All outgoing edges (i.e., the next outgoing edge of the next node) and their corresponding standardized scores are considered. The reachable node with the highest standardized score is then selected. The second step of data transmission is performed on the outgoing edge.

[0096] Based on the above steps, iterate through the source node, its corresponding outgoing edges, and the corresponding target node to determine the optimal outgoing edge and corresponding optimal node for each data transmission step. Continue until the starting node is reached. Data transmission to the destination node This allows us to obtain the complete optimal routing path for data transmission.

[0097] Finally, the gateway station 30 sends the obtained optimal routing path to the terminal device 20 to complete the data transmission and ensure the quality of the satellite communication service.

[0098] Therefore, according to the first aspect of this embodiment, by constructing a multi-layered graph structure corresponding to multiple satellite orbits of a satellite constellation, collaborative routing of multiple satellite orbits is achieved. This ensures wide-area communication coverage and adapts to dynamic and complex links between satellites operating asynchronously on multiple satellite orbits. Furthermore, it reduces the complexity of routing decisions.

[0099] Furthermore, according to a second aspect of this embodiment, a storage medium is provided. The storage medium includes a stored program, wherein, when the program is executed, a processor performs the method described in any of the above embodiments.

[0100] Therefore, according to this embodiment, by constructing a multi-layered graph structure corresponding to multiple satellite orbits of a satellite constellation, collaborative routing of multiple satellite orbits can be achieved. This ensures the range of wide-area communication and adapts to the dynamic and complex links between satellites operating asynchronously on multiple satellite orbits. Furthermore, it reduces the complexity of routing decisions.

[0101] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0102] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0103] Example 2

[0104] Figure 7 A routing decision apparatus for multi-orbit satellites according to the first aspect of this embodiment is shown, which corresponds to the method described according to the first aspect of Embodiment 1. Reference Figure 7 As shown, the device includes: a multi-layer graph structure construction module 710, used to construct a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate the links between satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; a graph data construction module 720, used to acquire the satellite status information of the satellites and the link status information of the corresponding links, and construct graph data corresponding to the multi-layer graph structure; and an optimal route path determination module 730, used to determine the optimal node based on the graph data, and obtain the optimal route path for data transmission.

[0105] Optionally, the graph data construction module 720 includes: a node attribute determination submodule, used to obtain the satellite transmit load rate, satellite receive load rate, satellite transmit remaining bandwidth, satellite receive remaining bandwidth, and satellite resource utilization rate of the satellite as satellite status information to determine the node attributes; a directed edge attribute determination submodule, used to obtain the one-way transmission delay, one-way maximum transmission bandwidth, one-way load rate, and remaining time coefficient of the link as link status information to determine the directed edge attributes, wherein the remaining time coefficient is related to the one-way remaining connectivity duration of the link; and a graph data determination submodule, used to construct graph data corresponding to the multi-layer graph structure based on the node attributes and directed edge attributes.

[0106] Optionally, the optimal route path determination module 730 includes: a reachability graph structure determination submodule, used to determine reachable nodes and reachable directed edges based on the starting node and destination node used for data transmission; a reachability graph data acquisition submodule, used to acquire reachability graph data corresponding to reachable nodes and reachable directed edges based on graph data; a message passing submodule, used to perform message passing on the reachability graph data using a graph neural network to determine the graph features corresponding to the reachability graph data; a standardized score determination submodule, used to determine the standardized score of reachable directed edges; and an optimal node determination submodule, used to determine the optimal node based on the standardized score using a greedy strategy.

[0107] Optionally, the message passing submodule includes: an attribute acquisition submodule, used to acquire the directed edge attributes of reachable directed edges and the corresponding source and target node attributes based on the reachable graph data; a feature determination submodule, used to determine the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target nodes based on the directed edge attributes of the reachable directed edges and the corresponding source and target node attributes using the message passing layer of the graph neural network; an iterative update submodule, used to iteratively update the directed edge features and node features using multiple message passing layers of the graph neural network; and a graph feature determination submodule, used to determine the graph features corresponding to the reachable graph data based on the directed edge features and node features.

[0108] Optionally, the feature determination submodule includes: a fusion submodule, used to determine a fusion vector based on the directed edge attributes of reachable directed edges and the corresponding node attributes of source and target nodes; an attention score determination submodule, used to obtain directed edge features and attention scores corresponding to reachable directed edges based on the fusion vector using a message passing layer and activation function; a normalization submodule, used to normalize the attention scores of incoming edges corresponding to the target node to obtain attention weights; an aggregation submodule, used to obtain aggregation information based on the attention weights and the node attributes of the source nodes corresponding to the incoming edges; and a node feature determination submodule, used to obtain node features of the target node based on the aggregation information and the node attributes of the target node using a message passing layer and activation function.

[0109] Optionally, the standardized score determination submodule includes: an outgoing edge determination submodule, used to determine the outgoing edges corresponding to reachable nodes with reachable nodes as source nodes; and an outgoing edge standardized score determination submodule, used to obtain the standardized score corresponding to the outgoing edge based on the characteristics of the directed edge.

[0110] Optionally, the optimal node determination submodule includes: a first selection submodule, used to select the outgoing edge with the highest standardized score for data transmission based on the standardized score; a second selection submodule, used to take the target node corresponding to the outgoing edge as the next node, determine the next outgoing edge corresponding to the next node, determine the standardized score corresponding to the next outgoing edge, and select the next outgoing edge with the highest standardized score for data transmission; and an iteration submodule, used to iterate through reachable nodes to determine the optimal reachable node until the data is transmitted to the destination node, thereby obtaining the optimal routing path for data transmission.

[0111] Therefore, according to this embodiment, by constructing a multi-layered graph structure corresponding to multiple satellite orbits of a satellite constellation, collaborative routing of multiple satellite orbits can be achieved. This ensures the range of wide-area communication and adapts to the dynamic and complex links between satellites operating asynchronously on multiple satellite orbits. Furthermore, it reduces the complexity of routing decisions.

[0112] Example 3

[0113] Figure 8 A routing decision apparatus for multi-orbit satellites according to the first aspect of this embodiment is shown, which corresponds to the method described according to the first aspect of Embodiment 1. Reference Figure 8As shown, the device includes: a processor 810; and a memory 820 connected to the processor 810, for providing the processor 810 with instructions to perform the following processing steps: constructing a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate links between satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; acquiring satellite status information and corresponding link status information, constructing graph data corresponding to the multi-layer graph structure; and determining the optimal node based on the graph data to obtain the optimal routing path for data transmission.

[0114] Optionally, the operation of acquiring satellite status information and corresponding link status information, and constructing graph data corresponding to the multi-layer graph structure, includes: acquiring satellite transmit load rate, satellite receive load rate, satellite transmit remaining bandwidth, satellite receive remaining bandwidth, and satellite resource utilization rate as satellite status information to determine node attributes; acquiring one-way transmission delay, one-way maximum transmission bandwidth, one-way load rate, and remaining time coefficient of the link as link status information to determine directed edge attributes, wherein the remaining time coefficient is related to the one-way remaining connection time of the link; and constructing graph data corresponding to the multi-layer graph structure based on node attributes and directed edge attributes.

[0115] Optionally, the operation of determining the optimal node based on graph data includes: determining reachable nodes and reachable directed edges based on the starting and destination nodes used for data transmission; obtaining reachable graph data corresponding to the reachable nodes and reachable directed edges based on the graph data; using a graph neural network to perform message passing on the reachable graph data to determine the graph features corresponding to the reachable graph data; determining the standardized score of the reachable directed edges; and determining the optimal node based on the standardized score using a greedy strategy.

[0116] Optionally, using a graph neural network, message passing is performed on the reachable graph data corresponding to reachable nodes and reachable directed edges to determine the graph features corresponding to the reachable graph data. This includes: obtaining the directed edge attributes of reachable directed edges and the corresponding source and target node attributes based on the reachable graph data; using the message passing layer of the graph neural network, determining the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target nodes based on the directed edge attributes of the reachable directed edges and the corresponding source and target node attributes; iteratively updating the directed edge features and node features using multiple message passing layers of the graph neural network; and determining the graph features corresponding to the reachable graph data based on the directed edge features and node features.

[0117] Optionally, using the message-passing layer of a graph neural network, the operation of determining the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target node based on the directed edge attributes of the reachable directed edges and the corresponding node attributes of the source and target nodes includes: determining a fusion vector based on the directed edge attributes of the reachable directed edges and the corresponding node attributes of the source and target nodes; using the message-passing layer and activation function, obtaining the directed edge features and the attention scores corresponding to the reachable directed edges based on the fusion vector; normalizing the attention scores of the incoming edges corresponding to the target node to obtain attention weights; obtaining aggregate information based on the attention weights and the node attributes of the source nodes corresponding to the incoming edges; and using the message-passing layer and activation function, obtaining the node features of the target node based on the aggregate information and the node attributes of the target node.

[0118] Optionally, the operation of determining the standardized score of reachable directed edges includes: taking the reachable node as the source node, determining the outgoing edge corresponding to the reachable node; and obtaining the standardized score corresponding to the outgoing edge based on the characteristics of the directed edge.

[0119] Optionally, the operation of determining the optimal node based on the standardized score using a greedy strategy includes: selecting the outgoing edge with the highest standardized score for data transmission based on the standardized score; taking the target node corresponding to the outgoing edge as the next node, determining the next outgoing edge corresponding to the next node, and determining the standardized score corresponding to the next outgoing edge, selecting the next outgoing edge with the highest standardized score for data transmission; and iterating through the reachable nodes to determine the optimal reachable node until the data is transmitted to the destination node, thereby obtaining the optimal routing path for data transmission.

[0120] Therefore, according to this embodiment, by constructing a multi-layered graph structure corresponding to multiple satellite orbits of a satellite constellation, collaborative routing of multiple satellite orbits can be achieved. This ensures the range of wide-area communication and adapts to the dynamic and complex links between satellites operating asynchronously on multiple satellite orbits. Furthermore, it reduces the complexity of routing decisions.

[0121] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0122] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0123] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0124] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0125] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0126] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0127] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A routing decision method for multi-orbit satellites, applied to gateway stations, characterized in that, include: Construct a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate the links between the satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; Obtain the satellite status information of the satellite and the link status information of the corresponding link, and construct graph data corresponding to the multi-layer graph structure; as well as Based on the graph data, the optimal node is determined, and the optimal routing path for data transmission is obtained.

2. The method according to claim 1, characterized in that, The operation of acquiring the satellite status information of the satellite and the link status information of the corresponding link, and constructing graph data corresponding to the multi-layer graph structure, includes: The satellite transmit load rate, satellite receive load rate, satellite transmit remaining bandwidth, satellite receive remaining bandwidth, and satellite resource utilization rate of the satellite are obtained as satellite status information to determine the node attributes of the node. The one-way transmission delay, one-way maximum transmission bandwidth, one-way load rate, and remaining time coefficient of the link are obtained as the link state information to determine the directed edge attributes of the directed edge, wherein the remaining time coefficient is related to the one-way remaining connectivity duration of the link; and Based on the node attributes and the directed edge attributes, construct graph data corresponding to the multi-layer graph structure.

3. The method according to claim 1, characterized in that, The operation to determine the optimal node based on the graph data includes: Based on the originating and destination nodes used for data transmission, determine the reachable nodes and reachable directed edges; Based on the graph data, obtain the reachable graph data corresponding to the reachable nodes and the reachable directed edges; Using a graph neural network, message passing is performed on the reachable graph data to determine the graph features corresponding to the reachable graph data; Based on the graph features, determine the standardized score of the reachable directed edge; and Based on the standardized score, the optimal node is determined using a greedy strategy.

4. The method according to claim 3, characterized in that, The operation of using a graph neural network to perform message passing on the reachable graph data corresponding to the reachable nodes and the reachable directed edges, and to determine the graph features corresponding to the reachable graph data, includes: Based on the reachable graph data, obtain the directed edge attributes of the reachable directed edges and the corresponding source and target node attributes; Using the message passing layer of the graph neural network, based on the directed edge attributes of the reachable directed edges and the corresponding node attributes of the source and target nodes, the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target nodes are determined. The directed edge features and node features are iteratively updated using multiple message-passing layers of the graph neural network; and Based on the directed edge features and the node features, the graph features corresponding to the reachable graph data are determined.

5. The method according to claim 4, characterized in that, The operation of determining the directed edge features corresponding to the reachable directed edges and the node features corresponding to the target node, based on the directed edge attributes of the reachable directed edges and the corresponding source and target node attributes, using the message passing layer of the graph neural network, includes: The fusion vector is determined based on the directed edge attributes of the reachable directed edges and the corresponding node attributes of the source and target nodes. Using the message passing layer and activation function, and based on the fusion vector, the directed edge features and the attention scores corresponding to the reachable directed edges are obtained; The attention scores of the incoming edges corresponding to the target node are normalized to obtain the attention weights; Based on the attention weights and the node attributes of the source nodes corresponding to the incoming edges, aggregation information is obtained; and Using the message passing layer and the activation function, the node features of the target node are obtained based on the aggregated information and the node attributes of the target node.

6. The method according to claim 4, characterized in that, The operation of determining the standardized score of the reachable directed edge based on the graph features includes: Using the reachable node as the source node, determine the outgoing edges corresponding to the reachable node; and Based on the directed edge characteristics, the standardized score corresponding to the outgoing edge is obtained.

7. The method according to claim 3, characterized in that, Based on the standardized scores, the operations for determining the optimal node using a greedy strategy include: Based on the standardized score, the outgoing edge with the highest standardized score is selected for data transmission; Using the target node corresponding to the outgoing edge as the next node, determine the next outgoing edge corresponding to the next node, and determine the standardized score corresponding to the next outgoing edge. Select the next outgoing edge with the highest standardized score for data transmission; and Based on the reachable nodes, the optimal reachable node is determined iteratively until the data is transmitted to the destination node, thereby obtaining the optimal routing path for data transmission.

8. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, the method described in any one of claims 1 to 7 is performed by a processor.

9. A routing decision-making device for multi-orbit satellites, characterized in that, include: A multi-layer graph structure construction module is used to construct a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate the links between the satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation. The graph data construction module is used to obtain the satellite status information of the satellite and the link status information of the corresponding link, and construct graph data corresponding to the multi-layer graph structure. as well as The optimal route path determination module is used to determine the optimal node based on the graph data, and obtain the optimal route path for data transmission.

10. A routing decision-making device for multi-orbit satellites, characterized in that, include: processor; as well as A memory, connected to the processor, for providing the processor with instructions to perform the following processing steps: Construct a multi-layer graph structure corresponding to a satellite constellation, wherein the nodes of the multi-layer graph structure correspond to the satellites of the satellite constellation, the directed edges of the multi-layer graph structure are used to indicate the links between the satellites, and any layer of the multi-layer graph structure corresponds to a satellite orbit of the satellite constellation; Obtain the satellite status information of the satellite and the link status information of the corresponding link, and construct graph data corresponding to the multi-layer graph structure; as well as Based on the graph data, the optimal node is determined, and the optimal routing path for data transmission is obtained.

Images