Satellite routing method based on dynamic celestial sphere region division

Abstract

The invention relates to the technical field of satellite communication, in particular to a satellite routing method based on dynamic celestial sphere region division. The method comprises the following steps: dividing a plurality of nodes by using a clustering algorithm to obtain a plurality of load-balanced virtual areas; optimizing the plurality of virtual areas by using a local flooding optimization strategy, and generating a preset propagation range of the data packet; according to historical link delay data, utilizing a recurrent neural network model to predict delay and congestion conditions of links in a preset propagation range, and dynamically adjusting path selection and flow distribution strategies based on a prediction result; and calculating the additional delay of each path based on the delay and load conditions of different paths in the path selection and flow distribution strategy, and calculating the flow to be distributed of each path based on the additional delay. Therefore, the communication bottleneck problems of network delay, packet loss rate increase and the like caused by dynamic change of the network in the low-orbit satellite and the Internet of Things are solved, and efficient operation of the network in a dynamic environment is realized.

Description

Technical Field

[0001] The invention relates to the technical field of satellite communication, and in particular to a satellite routing method based on dynamic celestial sphere region division. Background Art

[0002] With the rapid development of satellite communication networks and Internet of Things technologies, global network connections have become increasingly complex. Especially in the application scenarios of Low Earth Orbit Satellite Networks (LEO) and Internet of Things (IoT), low-orbit satellites frequently connect and disconnect links when orbiting the earth; and a large number of mobile nodes such as sensors, devices, and vehicles in the Internet of Things also constantly change their positions and states at different times and spaces. These dynamic characteristics of the network directly affect the communication quality between nodes, such as delays and packet loss.

[0003] In current technology, commonly used routing algorithms include global flooding routing algorithm and static node partitioning strategy. The basic idea of ​​global flooding routing algorithm is to forward data packets to each node in the network through broadcasting and discard duplicate data packets when they are received. However, this method often fails to adjust the routing in time when the network load changes suddenly, resulting in increased network delay and increased packet loss rate. On the other hand, the static node partitioning strategy attempts to achieve reasonable resource utilization by pre-allocating network resources (such as frequency bands or time slots) to each node. However, this strategy lacks flexibility and cannot be adaptively adjusted according to the dynamic changes of the network, resulting in low resource utilization and decreased network efficiency.

[0004] Therefore, how to improve communication efficiency and effectively deal with problems such as delay and packet loss in a dynamically changing network environment has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The present invention provides a satellite routing method based on dynamic celestial sphere region division to solve communication bottleneck problems such as network delay and increased packet loss rate caused by dynamic changes in the network in low-orbit satellites and the Internet of Things, and realize efficient operation of the network in a dynamic environment.

[0006] To achieve the above object, a first aspect of the present invention provides a satellite routing method based on dynamic celestial sphere region division, comprising the following steps: Using a first preset clustering algorithm to initially divide the multiple nodes to obtain multiple load-balanced virtual areas; Utilizing a preset local flooding optimization strategy, the plurality of load-balanced virtual areas are optimized to generate a preset propagation range of a data packet; According to the historical link delay data, a preset recurrent neural network model is used to predict the delay and congestion of the links in the preset propagation range, and based on the prediction results, the path selection and traffic distribution strategies are dynamically adjusted; Based on the delay and load conditions of different paths in the path selection and traffic distribution strategy, the additional delay of each path is calculated, and the traffic to be distributed on each path is calculated based on the additional delay.

[0007] According to an embodiment of the present invention, the initial division of satellites or nodes using a preset clustering algorithm includes: Obtaining geographic coordinates and initial loads of the plurality of nodes, and selecting a preset number of initial center points as virtual node centers; Calculate the Euclidean distance from each node to each virtual node center respectively, and determine the virtual node center corresponding to each node based on the Euclidean distance; Calculate the average position coordinates and load of each virtual node center, and re-execute the steps of respectively calculating the Euclidean distance from each node to each virtual node center, and determining the virtual node center corresponding to each node based on the Euclidean distance, until the multiple nodes meet the preset division criteria.

[0008] According to an embodiment of the present invention, after the plurality of nodes meet the preset division criteria, the method further includes: Monitor the load distribution of each virtual node center in real time, and determine whether the load of each virtual node center is balanced based on the monitoring results; When the load is unbalanced, the node allocation in the corresponding virtual node center is dynamically adjusted using a second preset clustering algorithm to obtain the multiple load-balanced virtual areas.

[0009] According to an embodiment of the present invention, the method of optimizing the plurality of load-balanced virtual areas by using a preset local flooding optimization strategy to generate a preset propagation range of a data packet includes: Calculate the shortest hop count between any two nodes among all nodes; Based on the shortest hop count, determining a hop count limit of a corresponding node; The hop count of the current data packet is obtained, a node whose hop count limit is greater than or equal to the hop count of the current data packet is selected as a target node, and a preset propagation range of the data packet is generated based on the target node.

[0010] According to an embodiment of the present invention, dynamically adjusting the path selection and traffic distribution strategy based on the prediction result includes: Based on the prediction results, generating a delay matrix; According to the delay matrix, the flow distribution ratio of each path is calculated, and a flow distribution matrix is ​​generated; The path selection and traffic distribution strategy are dynamically adjusted according to the traffic distribution matrix.

[0011] According to a satellite routing method based on dynamic celestial sphere area division proposed in an embodiment of the present invention, multiple nodes are divided by using a clustering algorithm to obtain multiple load-balanced virtual areas; multiple virtual areas are optimized by using a local flooding optimization strategy to generate a preset propagation range of data packets; based on historical link delay data, a recurrent neural network model is used to predict the delay and congestion of links in the preset propagation range, and the path selection and traffic distribution strategy is dynamically adjusted based on the prediction results; the additional delay of each path is calculated based on the delay and load of different paths in the path selection and traffic distribution strategy, and the traffic to be distributed for each path is calculated based on the additional delay. In this way, the communication bottleneck problems such as network delay and increased packet loss rate caused by dynamic changes in the network in low-orbit satellites and the Internet of Things are solved, and the efficient operation of the network in a dynamic environment is achieved.

[0012] To achieve the above object, a second aspect of the present invention provides a satellite routing device based on dynamic celestial sphere region division, comprising: A partitioning module, used to perform initial partitioning on multiple nodes using a first preset clustering algorithm to obtain multiple load-balanced virtual areas; A generating module, configured to optimize the plurality of load-balanced virtual areas by using a preset local flooding optimization strategy, and generate a preset propagation range of a data packet; An adjustment module is used to predict the delay and congestion of the links in the preset propagation range according to the historical link delay data using a preset recurrent neural network model, and dynamically adjust the path selection and traffic distribution strategy based on the prediction results; A calculation module is used to calculate the additional delay of each path based on the delay and load conditions of different paths in the path selection and traffic distribution strategy, and calculate the to-be-allocated traffic of each path based on the additional delay.

[0013] According to an embodiment of the present invention, the division module is specifically used to: Obtaining geographic coordinates and initial loads of the plurality of nodes, and selecting a preset number of initial center points as virtual node centers; Calculate the Euclidean distance from each node to each virtual node center respectively, and determine the virtual node center corresponding to each node based on the Euclidean distance; Calculate the average position coordinates and load of each virtual node center, and re-execute the steps of respectively calculating the Euclidean distance from each node to each virtual node center, and determining the virtual node center corresponding to each node based on the Euclidean distance, until the multiple nodes meet the preset division criteria.

[0014] According to an embodiment of the present invention, after the plurality of nodes meet the preset division criteria, the division module is further configured to: Monitor the load distribution of each virtual node center in real time, and determine whether the load of each virtual node center is balanced based on the monitoring results; When the load is unbalanced, the node allocation in the corresponding virtual node center is dynamically adjusted using a second preset clustering algorithm to obtain the multiple load-balanced virtual areas.

[0015] According to one embodiment of the present invention, the generating module is specifically used for: Calculate the shortest hop count between any two nodes among all nodes; Based on the shortest hop count, determining a hop count limit of a corresponding node; The hop count of the current data packet is obtained, a node whose hop count limit is greater than or equal to the hop count of the current data packet is selected as a target node, and a preset propagation range of the data packet is generated based on the target node.

[0016] According to an embodiment of the present invention, the adjustment module is specifically used to: Based on the prediction results, generating a delay matrix; According to the delay matrix, the flow distribution ratio of each path is calculated, and a flow distribution matrix is ​​generated; The path selection and traffic distribution strategy are dynamically adjusted according to the traffic distribution matrix.

[0017] According to a satellite routing device based on dynamic celestial sphere area division proposed in an embodiment of the present invention, multiple nodes are divided by using a clustering algorithm to obtain multiple load-balanced virtual areas; multiple virtual areas are optimized by using a local flooding optimization strategy to generate a preset propagation range of data packets; based on historical link delay data, a recurrent neural network model is used to predict the delay and congestion of the link in the preset propagation range, and the path selection and traffic distribution strategy is dynamically adjusted based on the prediction results; the additional delay of each path is calculated based on the delay and load of different paths in the path selection and traffic distribution strategy, and the traffic to be allocated for each path is calculated based on the additional delay. In this way, the communication bottleneck problems such as network delay and increased packet loss rate caused by dynamic changes in the network in low-orbit satellites and the Internet of Things are solved, and the efficient operation of the network in a dynamic environment is achieved.

[0018] To achieve the above-mentioned purpose, the third aspect of the present invention proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a satellite routing method based on dynamic celestial region division as described in the above-mentioned embodiment.

[0019] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement a satellite routing method based on dynamic celestial region division as described in the above embodiments.

[0020] To achieve the above objectives, a fifth aspect of the present invention proposes a computer program product, which includes a computer program. When the computer program is executed by a processor, it is used to implement a satellite routing method based on dynamic celestial region division as in the above embodiment.

[0021] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which: Figure 1 A flowchart of a satellite routing method based on dynamic celestial sphere region division according to an embodiment of the present invention; Figure 2a A flowchart of another satellite routing method based on dynamic celestial sphere region division according to an embodiment of the present invention; Figure 2b A flow chart of a dynamic celestial sphere region division method, a local flood optimization method, and an intelligent path prediction method provided according to an embodiment of the present invention; Figure 2c A flowchart of a multipath load balancing method and an adaptive adjustment and fault tolerance mechanism method provided according to an embodiment of the present invention; Figure 3 A schematic block diagram of a satellite routing device based on dynamic celestial sphere region division according to an embodiment of the present invention; Figure 4 It is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0024] A satellite routing method based on dynamic celestial sphere region division according to an embodiment of the present invention is described below with reference to the accompanying drawings.

[0025] Figure 1 The present invention is a flowchart of a satellite routing method based on dynamic celestial region division according to an embodiment of the present invention.

[0026] Before introducing a satellite routing method based on dynamic celestial sphere region division proposed in an embodiment of the present invention, related technologies are first introduced.

[0027] Although the traditional global flooding routing algorithm can collect complete link information in the network, it has too high computing and communication overhead in complex dynamic networks, especially low-orbit satellite networks and large-scale Internet of Things, and it is difficult to meet the real-time requirements of the network. The global flooding routing algorithm needs to send routing information to each node in the network. As the scale of the network expands, data packets are flooded throughout the network, occupying a large amount of bandwidth and computing resources, causing the system to be overloaded. In addition, the traditional global flooding routing algorithm can only rely on the current link status and lacks the ability to predict future network changes. When the network load changes suddenly, it is often unable to adjust the path in time, resulting in network delays, increased packet loss rates and other communication bottlenecks.

[0028] At the same time, static node partitioning strategies also have significant limitations when dealing with dynamic network changes. The node load and topology in satellite networks and the Internet of Things change frequently, and statically divided node areas cannot be adaptively adjusted according to these changes, resulting in low resource utilization and reduced network efficiency. Fixedly divided routing areas are difficult to flexibly respond to load fluctuations on different paths, which can easily cause congestion on some paths, thereby affecting overall network performance. Especially under high load conditions, the overload problem of a single path or node will be exacerbated, and the traditional static partitioning strategy cannot perform effective load balancing adjustments, further increasing the risk of network communication bottlenecks.

[0029] It is based on the above problems that an embodiment of the present invention proposes a satellite routing method based on dynamic celestial region division. By combining the virtual node method based on dynamic celestial region division with the local flooding optimization strategy, it aims to solve the communication bottleneck problems such as network delay and increased packet loss rate caused by dynamic network changes in low-orbit satellites and the Internet of Things, improve the routing efficiency in complex dynamic networks such as low-orbit satellite networks and large-scale Internet of Things, reduce communication delays and enhance the network's adaptability and fault tolerance.

[0030] Next, a satellite routing method based on dynamic celestial sphere region division proposed in an embodiment of the present invention is described in detail.

[0031] For example, Figure 1 As shown, the satellite routing method based on dynamic celestial region division includes the following steps: In step S101, a first preset clustering algorithm is used to perform initial division on a plurality of nodes to obtain a plurality of load-balanced virtual areas.

[0032] Among them, the first preset clustering algorithm refers to the K-Means (K-Means Clustering Algorithm) clustering algorithm, which is an unsupervised learning algorithm that aims to divide samples in a data set into multiple clusters (or classes) according to a certain similarity measure, so that samples in the same cluster have high similarity to each other, while samples in different clusters have low similarity.

[0033] Specifically, the K-Means clustering algorithm is used to perform initial division of multiple satellites or nodes in the network, where K represents the number of virtual areas. Satellites or nodes are assigned to K clusters through an iterative optimization process, and each node can be assigned to a different virtual node area based on its geographical location and current load information. In this way, multiple nodes can be effectively divided into several clusters, each cluster represented by a virtual node, thereby achieving the initial division of nodes and obtaining multiple load-balanced virtual areas. This division helps to optimize the load balancing and resource allocation of the network and improve the overall performance and efficiency of the system.

[0034] The following describes in detail how to use the preset clustering algorithm to perform initial division of satellites or nodes.

[0035] As a possible implementation method, in some embodiments, a preset clustering algorithm is used to perform initial division of satellites or nodes, including: obtaining the geographic coordinates and initial loads of multiple nodes, and selecting a preset number of initial center points as virtual node centers; calculating the Euclidean distance from each node to each virtual node center, and determining the virtual node center corresponding to each node based on the Euclidean distance; calculating the average position coordinates and load of each virtual node center, and re-executing the steps of calculating the Euclidean distance from each node to each virtual node center, and determining the virtual node center corresponding to each node based on the Euclidean distance, until multiple nodes meet the preset division criteria.

[0036] Among them, Euclidean distance is a measurement method for the shortest distance between two points in geometry, and is often used to measure the straight-line distance between two points in multidimensional space. Initial load refers to the initial workload or task volume of a node, which is usually related to the performance and resource allocation of the node.

[0037] Specifically, we first input the geographic coordinates and initial load of each node in the network, and then select K initial cluster centers from these nodes (these points are preset to represent the center of each cluster) as virtual node centers to represent the center of each cluster. Next, we calculate the Euclidean distance from each node to these centers. , and assign each node to the nearest center point to form a preliminary cluster. Then, recalculate the new position of each cluster center point, that is, the average position of all node coordinates assigned to the center point, and update the load information of the cluster. Finally, using the updated virtual node center position and load, repeat the process of calculating the Euclidean distance of each node to these centers and reallocating nodes to the corresponding cluster centers according to this distance until the center point position and node allocation no longer change (i.e., meet the preset division criteria), indicating that the clustering process has been completed.

[0038] Furthermore, in some embodiments, after multiple nodes meet the preset division criteria, it also includes: real-time monitoring of the load distribution of each virtual node center, and judging whether the load of each virtual node center is balanced based on the monitoring results; when the load is unbalanced, dynamically adjusting the node allocation in the corresponding virtual node center using a second preset clustering algorithm to obtain multiple load-balanced virtual areas.

[0039] Among them, the second preset clustering algorithm refers to the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, which is an algorithm for data clustering. Its core idea is to group data points according to their density. In the DBSCAN algorithm, data points are divided into three categories: core points, boundary points, and noise points. A core point refers to a point that has enough other points in its neighborhood, a boundary point refers to a point that is in the neighborhood of a core point but is not a core point itself, and a noise point refers to a point that does not have enough points in the neighborhood of any core point. The advantage of the DBSCAN algorithm is that it does not require the number of clusters to be set in advance, can identify clusters of any shape, and can effectively handle noise points.

[0040] Specifically, the system monitors the load distribution of each virtual node center in real time, and then uses formula (1) combined with the monitored data to determine whether the workload of these virtual node centers has reached a balanced state.

[0041]

[0042] in, For nodes, is the load of the node.

[0043] When the load is unbalanced, the second preset clustering algorithm may be used to dynamically adjust the node allocation in the corresponding virtual node center to ensure that the workload of all virtual areas is balanced.

[0044] It should be noted that if the load of all nodes is close to its average value, it means that the load distribution is balanced; if the load of some nodes is much higher or lower than that of other nodes, it means that the load is unbalanced.

[0045] The specific calculation method to determine whether the load is balanced is as follows: (1) Calculate the average load of all nodes, as shown in formula (2):

[0046] in, For nodes, For load, is the number of nodes participating in the calculation in the network. Load It is defined as the total traffic (number of packets or bandwidth consumption) processed by the node per unit time.

[0047] (2) Calculate the standard deviation of node load, as shown in formula (3):

[0048] When the standard deviation Meet the preset threshold range When , the load distribution can be considered balanced. If the standard deviation Exceeds the preset threshold range , then the load distribution is considered to be unbalanced.

[0049] (3) Use the second preset clustering algorithm (DBSCAN algorithm) to dynamically adjust the node allocation in the corresponding virtual node center.

[0050] ① Define node positions: The position of each node can be defined by its Euclidean distance from the center of the virtual node, forming the spatial distribution of the data, that is, the distribution of nodes in space; ② Run DBSCAN clustering: Use the load and location of the nodes as input data and use DBSCAN for clustering. The main steps of the DBSCAN algorithm are as follows: a. Select appropriate neighborhood radius and minimum number of points as parameters of the DBSCAN algorithm.

[0051] b. Run the DBSCAN algorithm on all nodes and mark the core points, boundary points and noise points.

[0052] c. In the cluster, nodes with similar loads are divided into a virtual node cluster. By dynamically adjusting the division of virtual node centers, the load of each node is balanced.

[0053] ③ Adjust the virtual node center: According to the clustering results of the DBSCAN algorithm, adjust the position of the virtual node center to make it close to the area where the load is concentrated, ensuring that each virtual node can better handle the node load around it to improve the efficiency and response speed of the entire system.

[0054] ④ Cyclic monitoring and adjustment: Continuously monitor the node load distribution, and regularly recalculate the average load and load standard deviation of the node to evaluate the load balancing situation.

[0055] ⑤ Dynamically adjust node division: If the load distribution is uneven, you can rerun the DBSCAN algorithm to adjust the division of virtual nodes until a balanced state is reached, thereby obtaining multiple load-balanced virtual areas.

[0056] In step S102, a preset local flooding optimization strategy is used to optimize multiple load-balanced virtual areas to generate a preset propagation range of a data packet.

[0057] It is understandable that in order to prevent a large amount of aimless propagation when sending data packets in the network, that is, global flooding, which leads to a waste of network resources, a preset local flooding optimization strategy can be used to limit the area where data packets are propagated to improve network efficiency. By optimizing multiple load-balancing virtual areas, it is ensured that data packets are only propagated within the preset propagation range, rather than spreading indefinitely in the entire network, so as to reduce communication overhead.

[0058] Next, it is explained in detail how to use the preset local flooding optimization strategy to optimize multiple load-balanced virtual areas and generate the preset propagation range of the data packet.

[0059] As a possible implementation method, in some embodiments, a preset local flooding optimization strategy is used to optimize multiple load-balanced virtual areas to generate a preset propagation range of the data packet, including: calculating the shortest hop number between any two nodes among all nodes; based on the shortest hop number, determining the hop limit of the corresponding node; obtaining the hop number of the current data packet, selecting a node whose hop limit is greater than or equal to the hop number of the current data packet as the target node, and generating the preset propagation range of the data packet based on the target node.

[0060] Specifically, to avoid the high overhead of global flooding, embodiments of the present invention may use hop count limits to control the propagation range of data packets. First, a comprehensive calculation may be performed on all nodes in the network to determine all possible shortest paths between them. This step involves complex algorithms, such as the Dijkstra algorithm (also known as the Dijkstra algorithm or the Floyd-Warshall algorithm) to ensure that the shortest number of hops between any two nodes can be accurately found. After completing this step, a local hop count limit is set for each node. , used to control the propagation range of data packets in the network.

[0061] When a data packet propagates in the network, it can start from the source node and send it to its neighboring nodes by flooding. , which means that each node checks its hop counter after receiving a packet. If the hop counter value is less than or equal to , then the data packet will continue to be flooded to the adjacent nodes; on the contrary, if the value of the hop counter exceeds , then the data packet will no longer be transmitted, thus avoiding unnecessary communication overhead.

[0062] In this way, fast routing selection can be achieved in a local area. Since data packets are only transmitted within a limited number of hops, redundant data transmission in the network can be significantly reduced, thereby improving overall network efficiency.

[0063] in, Calculated by the shortest number of hops between nodes:

[0064] In step S103, based on the historical link delay data, a preset recurrent neural network model is used to predict the delay and congestion of the links in the preset propagation range, and based on the prediction results, the path selection and traffic distribution strategies are dynamically adjusted.

[0065] Among them, the preset recurrent neural network model can be LSTM (Long Short-Term Memory), which is a special recurrent neural network structure.

[0066] That is to say, by training and learning from historical link delay data, the recurrent neural network can capture the time series characteristics of link delay and predict future delay and congestion status based on these characteristics.

[0067] Predicted values ​​of link path delay and congestion status The calculation is as follows:

[0068] in, is the historical link delay data, is the link path.

[0069] According to the prediction results, the path selection follows the following formula:

[0070] in, is the actual delay of the current link, To adjust the weight of the impact of future delay prediction, is a small constant used to avoid the denominator being zero, is the expected distance from the source point O to the target point A, From the source point O to the node actual distance.

[0071] Based on these prediction results, corresponding measures can be taken to optimize network performance, that is, dynamically adjust the path selection strategy to ensure that data packets can be transmitted through low-latency and unobstructed paths, thereby improving the transmission efficiency of the overall network. At the same time, according to the predicted congestion situation, traffic can be reasonably allocated to avoid overload on certain links, further improving the stability and reliability of the network. Through this prediction-based dynamic adjustment mechanism, uncertainties and changes in the network can be effectively dealt with, ensuring the optimal configuration and utilization of network resources.

[0072] Among them, in the training process of the preset recurrent neural network model LSTM, in order to measure the prediction effect of the model, the mean square error can be used as the loss function to measure the difference between the predicted path delay and the actual delay. At the same time, the Adam (Adaptive Moment Estimation) algorithm has the advantage of adaptive learning rate, which can accelerate convergence, so the LSTM model can use the Adam optimizer for gradient descent and parameter update, and set the initial learning rate to 0.001. In addition, in the data processing stage, the collected historical data needs to be scientifically and reasonably divided into training sets, validation sets, and test sets. The specific ratio can be: 80% of the data is used for training, 10% of the data is used for validation, and the remaining 10% of the data is used for testing.

[0073] The specific model training steps are: (1) Initialize the LSTM model and its parameters.

[0074] (2) Input historical delay data into the LSTM model according to the time window for training.

[0075] (3) Calculate the loss (mean square error) and update the model weights through the Adam optimizer.

[0076] (4) After each round of training, use the validation set data to verify the model to avoid overfitting.

[0077] (5) Determine hyperparameters (number of hidden layer units, learning rate, etc.) through cross-validation.

[0078] (6) Continue training until the loss function converges (the error no longer decreases significantly).

[0079] After the LSTM model is trained, assuming that there is the latest historical delay data of a certain path, after inputting it into the LSTM model, the model will output the delay prediction value for a period of time in the future. Based on the prediction results of the LSTM model, the system can dynamically adjust the path selection and traffic distribution strategy.

[0080] The following describes in detail how to dynamically adjust path selection and traffic distribution strategies based on prediction results.

[0081] As a possible implementation method, in some embodiments, path selection and traffic distribution strategies are dynamically adjusted based on the prediction results, including: generating a delay matrix based on the prediction results; calculating the traffic distribution ratio of each path according to the delay matrix, and generating a traffic distribution matrix; and dynamically adjusting path selection and traffic distribution strategies according to the traffic distribution matrix.

[0082] Specifically, assuming that there are N paths in the network, the delay of each path can be predicted by the LSTM model and a delay matrix can be generated.

[0083] Delay Matrix is a matrix consisting of future path delays predicted by the LSTM model, with a dimension of 1×N. For the The delay prediction value of each path.

[0084] According to the delay matrix The delay prediction value of each path in , the traffic distribution ratio of the corresponding path can be calculated :

[0085] A traffic distribution matrix can be generated based on the traffic distribution ratio of each path.

[0086] Traffic distribution matrix The dimension is 1×N and satisfies the following constraints:

[0087] That is, the sum of the traffic distribution ratios of all paths is 1.

[0088] It can be seen from formula (7) that according to the delay value predicted by the LSTM model, the following principles are followed when allocating traffic: If the latency of a network path is high, the data traffic allocated to the path should be appropriately reduced to prevent network congestion; conversely, if the latency of a path is relatively low, its traffic allocation should be appropriately increased to improve the load balancing efficiency of the overall network.

[0089] It is understandable that in the actual deployment process, various dynamic changes are often experienced, such as the increase and decrease of equipment, link fluctuations, etc. In order to cope with these changes, the LSTM model can be continuously optimized through online learning capabilities. Specifically, whenever new path delay data is generated in the network, this data can be captured in real time and added to the model as new training samples. In this way, the LSTM model can update its internal parameters in real time to better adapt to changes in network status. This continuous online learning mechanism enables the LSTM model to maintain high prediction accuracy and robustness when facing complex network environments.

[0090] In step S104, based on the delay and load conditions of different paths in the path selection and traffic distribution strategy, the additional delay of each path is calculated, and the traffic to be distributed on each path is calculated based on the additional delay.

[0091] Specifically, the system can calculate the additional delay of each path based on the delay and load on different paths. :

[0092] in, is the packet length, is the bandwidth rate of the intersatellite link.

[0093] According to the additional delay of each path and the current load, the system can dynamically adjust and distribute network traffic in real time, thereby ensuring that the load of each path is evenly distributed, avoiding performance bottlenecks due to overload on some paths, and thus improving the stability and efficiency of the overall network.

[0094] The path load balancing formula is:

[0095] in, is the load of the path, is the adjustment factor.

[0096] The optimization formula for path selection is as follows:

[0097] The dynamic allocation method is: (1) Initial traffic allocation: The system first performs initial allocation based on the delay and load conditions of the path.

[0098] (2) Iterative adjustment: In each round of iteration, small adjustments are made to the traffic based on the actual delay changes and load distribution effects to avoid drastic fluctuations.

[0099] (3) Convergence condition: Set a convergence condition when the load standard deviation of all paths is less than a certain threshold. Stop flow adjustment.

[0100] Therefore, by selecting a path with lower latency, load balancing on multiple paths can be achieved, reducing network congestion.

[0101] In addition, by monitoring network status in real time and predicting potential failures, the algorithm's adaptive adjustment and fault-tolerant capabilities can automatically adjust routes when failures occur to ensure the continuity and reliability of data transmission.

[0102] Specifically, the system can use the LSTM model to monitor the status of nodes and links in real time and predict potential failures. The failure prediction results are as follows:

[0103] The system has the ability to automatically discover network topology. Through network scanning, routing protocol interaction, etc., it can build a topological structure view of the entire network. This topological view includes all nodes and links in the network and the connection relationships between them, providing a basic network map for path adjustment.

[0104] When the probability of system failure exceeds a preset threshold, the system can automatically adjust the path based on the network topology and use a path search algorithm to find an alternative path. This process ensures that the data flow can effectively avoid those nodes with failures, thereby ensuring the stability and reliability of the entire network. In this way, the system can respond quickly when a failure occurs and re-plan the data transmission path to avoid data being stranded or lost at the failed node, ensuring smooth and efficient data flow.

[0105] When the failed node or link is restored, the system will re-evaluate the network topology, identify the repaired nodes or links, and reintegrate them into the virtual node cluster. By updating the path selection, the system can redistribute data traffic to avoid passing through the still faulty parts, thereby ensuring the continuity and stability of data transmission.

[0106] To facilitate those skilled in the art to further understand a satellite routing method based on dynamic celestial sphere region division proposed in an embodiment of the present invention, the following is a schematic diagram of a satellite routing method based on dynamic celestial sphere region division. Figure 2a-2c For further explanation.

[0107] like Figure 2a As shown, another satellite routing method based on dynamic celestial sphere area division includes the following steps: Step S201, dynamic celestial sphere region division. That is, through dynamic celestial sphere region division, the network structure is adaptively adjusted according to the geographical location of the node and the real-time load to ensure load balancing.

[0108] Step S202: Local flooding optimization. That is, local flooding optimization technology is used to limit the propagation range of data packets, quickly collect link status, and reduce communication overhead.

[0109] Step S203: Intelligent path prediction: that is, using the LSTM model to predict link delays, select the optimal path in advance, and avoid future congestion risks.

[0110] Step S204: multi-path load balancing. That is, the system dynamically adjusts traffic distribution through multi-path load balancing to avoid overloading of a single path and improve transmission efficiency.

[0111] Step S205, adaptive adjustment and fault tolerance mechanism. That is, by real-time monitoring of network status and prediction of potential failures, based on the adaptive adjustment and fault tolerance capabilities of the algorithm, the route is automatically adjusted when a failure occurs to ensure the continuity and reliability of data transmission.

[0112] Furthermore, if Figure 2b As shown in (a) in the figure, the dynamic celestial sphere region division includes the following steps: Step S2011, input the geographic coordinates and initial load of the node.

[0113] Step S2012, select K initial centers as virtual node centers.

[0114] Step S2013, calculating the Euclidean distance from all nodes to the virtual node center.

[0115] Step S2014, assigning the nodes to the virtual node centers closest to them.

[0116] Step S2015, calculating the average position and load of each virtual node center.

[0117] Step S2016, recalculate the average position and load of each virtual node center.

[0118] Step S2017, determine whether the location of the virtual node center and the node allocation have changed. If so, execute step S2013, otherwise execute step S2018.

[0119] Step S2018, first define the node load, then calculate the average load of all nodes, and finally calculate the standard deviation of all nodes.

[0120] Step S2019, use the DBSCAN clustering algorithm to adjust the virtual node division.

[0121] Furthermore, if Figure 2b As shown in (b), the local Hongfan optimization includes the following steps: Step S2021, calculating the shortest hop count between all node pairs in the network.

[0122] Step S2022, determine whether the shortest hop count is less than or equal to the hop count limit. If so, execute step S2031, otherwise execute step S2013.

[0123] Furthermore, if Figure 2b As shown in (c), intelligent path prediction includes the following steps: Step S2031, training the LSTM model.

[0124] Step S2032: Use the LSTM model to predict the delay and congestion of the future path.

[0125] Step S2033, input the delay matrix predicted by the LSTM model.

[0126] Step S2034, calculating the flow distribution ratio of each path.

[0127] Step S2035, dynamically adjust the traffic distribution matrix.

[0128] Step S2036, dynamic flow adjustment.

[0129] Furthermore, if Figure 2c As shown in (d) in the figure, multipath load balancing includes the following steps: Step S2041, calculating the additional delay.

[0130] Step S2042, perform traffic scheduling.

[0131] Step S2043, determine whether the load standard deviation of all paths is less than a preset threshold. If yes, execute step S2044, otherwise execute step S2042.

[0132] Furthermore, if Figure 2c As shown in (e) in the figure, the adaptive adjustment and fault tolerance mechanism includes the following steps: Step S2051, perform fault prediction.

[0133] Step S2052, fault recovery.

[0134] In summary, the satellite routing method based on dynamic celestial sphere region division proposed in the embodiment of the present invention has the following advantages: (1) Efficient self-adaptation: By dynamically adjusting the division of celestial regions and updating the division of virtual nodes in real time, the system can adapt to changes in the network structure and load fluctuations in the dynamic network, thus ensuring efficient operation of the network.

[0135] (2) Reduce network overhead: The local flooding optimization mechanism reduces the excessive computing and communication overhead caused by traditional global flooding by limiting the number of hops of data packets, thereby improving the routing efficiency of the system.

[0136] (3) Foresight in path selection: Intelligent path prediction is not only based on the current link status, but can also predict future link delays through machine learning to avoid potential network congestion, thereby optimizing data transmission paths and improving network performance.

[0137] (4) Fault-tolerance mechanism: Through fault prediction and adaptive adjustment, the system can automatically switch paths when a fault occurs, ensuring the stability and fault tolerance of the network.

[0138] According to a satellite routing method based on dynamic celestial sphere area division proposed in an embodiment of the present invention, multiple nodes are divided by using a clustering algorithm to obtain multiple load-balanced virtual areas; multiple virtual areas are optimized by using a local flooding optimization strategy to generate a preset propagation range of data packets; based on historical link delay data, a recurrent neural network model is used to predict the delay and congestion of links in the preset propagation range, and the path selection and traffic distribution strategy is dynamically adjusted based on the prediction results; the additional delay of each path is calculated based on the delay and load of different paths in the path selection and traffic distribution strategy, and the traffic to be distributed for each path is calculated based on the additional delay. In this way, the communication bottleneck problems such as network delay and increased packet loss rate caused by dynamic changes in the network in low-orbit satellites and the Internet of Things are solved, and the efficient operation of the network in a dynamic environment is achieved.

[0139] Next, a satellite routing device based on dynamic celestial sphere region division according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0140] Figure 3 The present invention is a schematic block diagram of a satellite routing device based on dynamic celestial sphere region division according to an embodiment of the present invention.

[0141] like Figure 3 As shown, the satellite routing device 10 based on dynamic celestial region division includes: a division module 100 , a generation module 200 , an adjustment module 300 and a calculation module 400 .

[0142] The partitioning module 100 is used to perform initial partitioning on multiple nodes using a first preset clustering algorithm to obtain multiple load-balanced virtual areas; A generating module 200 is used to optimize multiple load-balanced virtual areas by using a preset local flooding optimization strategy to generate a preset propagation range of a data packet; The adjustment module 300 is used to predict the delay and congestion of the links in the preset propagation range according to the historical link delay data using the preset recurrent neural network model, and dynamically adjust the path selection and traffic distribution strategy based on the prediction results; The calculation module 400 is used to calculate the additional delay of each path based on the delay and load conditions of different paths in the path selection and traffic distribution strategy, and calculate the to-be-distributed traffic of each path based on the additional delay.

[0143] Furthermore, in some embodiments, the partitioning module 100 is specifically configured to: Obtaining geographic coordinates and initial loads of multiple nodes, and selecting a preset number of initial center points as virtual node centers; Calculate the Euclidean distance from each node to the center of each virtual node respectively, and determine the virtual node center corresponding to each node based on the Euclidean distance; Calculate the average position coordinates and load of each virtual node center, and re-calculate the Euclidean distance from each node to each virtual node center, and determine the virtual node center corresponding to each node based on the Euclidean distance, until multiple nodes meet the preset division criteria.

[0144] Further, in some embodiments, after a plurality of nodes meet a preset division criterion, the division module 100 is further configured to: Monitor the load distribution of each virtual node center in real time, and determine whether the load of each virtual node center is balanced based on the monitoring results; When the load is unbalanced, the node allocation in the corresponding virtual node center is dynamically adjusted using the second preset clustering algorithm to obtain multiple load-balanced virtual areas.

[0145] Furthermore, in some embodiments, the generating module 200 is specifically configured to: Calculate the shortest hop count between any two nodes among all nodes; Based on the shortest hop count, determine the hop count limit of the corresponding node; The hop count of the current data packet is obtained, a node whose hop count limit is greater than or equal to the hop count of the current data packet is selected as a target node, and a preset propagation range of the data packet is generated based on the target node.

[0146] Furthermore, in some embodiments, the adjustment module 300 is specifically configured to: Based on the prediction results, a delay matrix is ​​generated; According to the delay matrix, the traffic distribution ratio of each path is calculated and the traffic distribution matrix is ​​generated; Dynamically adjust path selection and traffic distribution strategies based on the traffic distribution matrix.

[0147] It should be noted that the above explanation of an embodiment of a satellite routing method based on dynamic celestial region division is also applicable to a satellite routing device based on dynamic celestial region division in this embodiment, which will not be repeated here.

[0148] According to a satellite routing device based on dynamic celestial sphere area division proposed in an embodiment of the present invention, multiple nodes are divided by using a clustering algorithm to obtain multiple load-balanced virtual areas; multiple virtual areas are optimized by using a local flooding optimization strategy to generate a preset propagation range of data packets; based on historical link delay data, a recurrent neural network model is used to predict the delay and congestion of the link in the preset propagation range, and the path selection and traffic distribution strategy is dynamically adjusted based on the prediction results; the additional delay of each path is calculated based on the delay and load of different paths in the path selection and traffic distribution strategy, and the traffic to be allocated for each path is calculated based on the additional delay. In this way, the communication bottleneck problems such as network delay and increased packet loss rate caused by dynamic changes in the network in low-orbit satellites and the Internet of Things are solved, and the efficient operation of the network in a dynamic environment is achieved.

[0149] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. The electronic device may include: Memory 401 , processor 402 , and a computer program stored in the memory 401 and executable on the processor 402 .

[0150] When the processor 402 executes the program, a satellite routing method based on dynamic celestial sphere area division provided in the above embodiment is implemented.

[0151] Furthermore, the electronic device further comprises: The communication interface 403 is used for communication between the memory 401 and the processor 402 .

[0152] The memory 401 is used to store computer programs that can be executed on the processor 402 .

[0153] The memory 401 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0154] If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 can be connected to each other through a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0155] Optionally, in a specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on a chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through an internal interface.

[0156] The processor 402 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0157] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the satellite routing method based on dynamic celestial region division as described above is implemented.

[0158] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the computer program implements the above satellite routing method based on dynamic celestial region division.

[0159] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0160] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms are not necessarily directed to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0161] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A satellite routing method based on dynamic celestial sphere region division, characterized in that: The following steps are involved: Using a first preset clustering algorithm to initially divide the multiple nodes to obtain multiple load-balanced virtual areas; Utilizing a preset local flooding optimization strategy, the plurality of load-balanced virtual areas are optimized to generate a preset propagation range of a data packet; According to the historical link delay data, a preset recurrent neural network model is used to predict the delay and congestion of the links in the preset propagation range, and based on the prediction results, the path selection and traffic distribution strategies are dynamically adjusted; Based on the delay and load conditions of different paths in the path selection and traffic distribution strategy, the additional delay of each path is calculated, and the traffic to be distributed on each path is calculated based on the additional delay.

2. The satellite routing method based on dynamic celestial sphere region division according to claim 1, characterized in that: The initial division of satellites or nodes by using a preset clustering algorithm includes: Obtaining geographic coordinates and initial loads of the plurality of nodes, and selecting a preset number of initial center points as virtual node centers; Calculate the Euclidean distance from each node to each virtual node center respectively, and determine the virtual node center corresponding to each node based on the Euclidean distance; Calculate the average position coordinates and load of each virtual node center, and re-execute the steps of respectively calculating the Euclidean distance from each node to each virtual node center, and determining the virtual node center corresponding to each node based on the Euclidean distance, until the multiple nodes meet the preset division criteria.

3. The satellite routing method based on dynamic celestial sphere region division according to claim 2, characterized in that: After the plurality of nodes meet the preset division criteria, the method further includes: Monitor the load distribution of each virtual node center in real time, and determine whether the load of each virtual node center is balanced based on the monitoring results; When the load is unbalanced, the node allocation in the corresponding virtual node center is dynamically adjusted using a second preset clustering algorithm to obtain the multiple load-balanced virtual areas.

4. The satellite routing method based on dynamic celestial sphere region division according to claim 1, characterized in that: The method of optimizing the plurality of load-balanced virtual areas by using a preset local flooding optimization strategy to generate a preset propagation range of a data packet includes: Calculate the shortest hop count between any two nodes among all nodes; Based on the shortest hop count, determining a hop count limit of a corresponding node; The hop count of the current data packet is obtained, a node whose hop count limit is greater than or equal to the hop count of the current data packet is selected as a target node, and a preset propagation range of the data packet is generated based on the target node.

5. The satellite routing method based on dynamic celestial sphere region division according to claim 1, characterized in that: The method of dynamically adjusting the path selection and traffic distribution strategy based on the prediction result includes: Based on the prediction results, generating a delay matrix; According to the delay matrix, the flow distribution ratio of each path is calculated, and a flow distribution matrix is ​​generated; The path selection and traffic distribution strategy are dynamically adjusted according to the traffic distribution matrix.

6. A satellite routing device based on dynamic celestial sphere region division, characterized in that: include: A partitioning module, used to perform initial partitioning on multiple nodes using a first preset clustering algorithm to obtain multiple load-balanced virtual areas; A generating module, configured to optimize the plurality of load-balanced virtual areas by using a preset local flooding optimization strategy, and generate a preset propagation range of a data packet; An adjustment module is used to predict the delay and congestion of the links in the preset propagation range according to the historical link delay data using a preset recurrent neural network model, and dynamically adjust the path selection and traffic distribution strategy based on the prediction results; A calculation module is used to calculate the additional delay of each path based on the delay and load conditions of different paths in the path selection and traffic distribution strategy, and calculate the to-be-allocated traffic of each path based on the additional delay.

7. The satellite routing device based on dynamic celestial sphere region division according to claim 6, characterized in that: The division module is specifically used for: Obtaining geographic coordinates and initial loads of the plurality of nodes, and selecting a preset number of initial center points as virtual node centers; Calculate the Euclidean distance from each node to each virtual node center respectively, and determine the virtual node center corresponding to each node based on the Euclidean distance; Calculate the average position coordinates and load of each virtual node center, and re-execute the steps of respectively calculating the Euclidean distance from each node to each virtual node center, and determining the virtual node center corresponding to each node based on the Euclidean distance, until the multiple nodes meet the preset division criteria.

8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a satellite routing method based on dynamic celestial region division as described in any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement a satellite routing method based on dynamic celestial region division as described in any one of claims 1 to 5.

10. A computer program product, characterized in that It comprises a computer program, which, when executed by a processor, is used to implement a satellite routing method based on dynamic celestial sphere area division as described in any one of claims 1 to 5.

Images