Dynamic satellite clustering method, device and equipment and storage medium

Abstract

The embodiment of the invention provides a dynamic satellite clustering method and device, equipment and a storage medium. The method comprises the following steps: acquiring first satellite information, inputting the first satellite information into a conversion model, extracting second satellite information, inputting the second satellite information into a cluster head determination model to predict the probability of each satellite serving as a cluster head, selecting the cluster head according to the probability of each satellite serving as the cluster head and a preset cluster head probability threshold, and then determining the cluster head according to the selected cluster head. And calculating the matching degree between the to-be-clustered satellites and the cluster heads, and clustering the to-be-clustered satellites. And at a preset clustering time interval, judging whether to re-cluster according to the total transmission data volume of each cluster head and the gateway station. In this way, the stability of satellite communication connection can be improved, and frequent oscillation of network topology is avoided.

Description

Technical Field

[0001] Embodiments of this application relate to the field of satellite communication technology, and in particular to a dynamic satellite clustering method, apparatus, device, and storage medium. Background Technology

[0002] In low-Earth orbit (LEO) satellite communication systems, the gateway station, acting as a ground access point, typically has multiple satellites within its coverage area that can establish connections with it. Satellites within the gateway station's coverage area are often managed using a clustering structure, with cluster heads elected to handle communication with the gateway station. However, the high-speed motion of satellites causes continuous dynamic changes within the gateway station's coverage area. Satellites constantly move out of the coverage area and lose connection, while new satellites enter the coverage area requiring connection. Current technologies, to cope with changes in the number of satellites within the coverage area, trigger a complete re-execution of the clustering process whenever an increase or decrease in the number of satellites is detected. This not only forces the selected cluster head satellite and the gateway station to continuously perform high-intensity computing tasks, significantly consuming the already limited onboard processing power and precious energy of the satellite platform, but also causes continuous and severe oscillations in the network topology. Frequent network topology reconstruction leads to repeated interruptions and reconstructions of satellite communication connections, resulting in significant communication latency, data loss, and severely damaging the stability and service continuity of the satellite network.

[0003] Therefore, there is an urgent need for an efficient dynamic satellite clustering method to solve the problems of excessive resource consumption and frequent network oscillations caused by frequent clustering and reconstruction. Summary of the Invention

[0004] According to an embodiment of this application, a dynamic satellite clustering scheme is provided, which can effectively reduce the satellite clustering frequency and significantly improve the stability of satellite communication connections.

[0005] In a first aspect of this application, a dynamic satellite clustering method is provided. The method includes: Acquire the first satellite information, input the first satellite information into the conversion model, and extract the second satellite information; The second satellite information is input into the cluster head determination model to predict the probability of each satellite as a cluster head, and the cluster head is selected based on the probability of each satellite as a cluster head and the preset cluster head probability threshold. Calculate the matching degree between the satellite to be clustered and the cluster head, and then cluster the satellite to be clustered. At a preset clustering time interval, it is determined whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station.

[0006] In one possible implementation, the first satellite information includes the satellite's horizontal coordinate, vertical coordinate, average signal-to-noise ratio of the input link, average signal-to-noise ratio of the output link, channel bandwidth, average connectivity, buffer utilization, and computing resource utilization.

[0007] In one possible implementation, the cluster head determination model consists of a deep learning module, a fully connected layer, and a binary classifier; The cluster head determination model was trained using the historical second satellite information dataset.

[0008] In one possible implementation, training the transformation model includes, Obtain the first satellite information of the first sample to be clustered, as well as the first satellite information of the first sample cluster head and the second sample cluster head; The transformation model outputs the second satellite information of the first sample to be clustered, the first sample cluster head, and the second sample cluster head, respectively. The second satellite information of the first sample to be clustered, the first sample cluster head, and the second sample cluster head is input into the fusion model to predict the probability that the first sample to be clustered will be added to the cluster where the first sample cluster head is located and the probability that the first sample to be clustered will be added to the cluster where the second sample cluster head is located. The probability of adding the first sample to be clustered into the cluster where the first sample cluster head is located is compared with the actual probability of the first sample to be clustered into the cluster where the first sample cluster head is located, and the first loss is obtained. The probability of adding the first sample to be clustered into the cluster where the second sample cluster head is located is compared with the actual probability of the first sample to be clustered into the cluster where the second sample cluster head is located, and the second loss is obtained. The conversion model is trained using the first loss and the second loss.

[0009] In one possible implementation, the matching degree between the satellite to be clustered and each cluster head is calculated, and the satellite to be clustered is then clustered, including... Obtain the channel bandwidth and signal-to-noise ratio of the satellites to be clustered and each cluster head, and calculate the channel capacity between the satellites to be clustered and each cluster head according to Shannon's formula; The matching degree between the satellite to be clustered and each cluster head is calculated based on the channel capacity between the satellite to be clustered and each cluster head, and the relative distance between the satellite to be clustered and each cluster head. The matching degree between the satellites to be clustered and each cluster head is sorted, and the satellites to be clustered are added to the cluster head with the highest matching degree.

[0010] In one possible implementation, at a preset clustering time interval, a decision is made on whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station, including: Calculate the mean and variance of the total data transmission volume between each cluster head and the gateway station in turn; If the variance of the total data transmitted between each cluster head and the gateway station is greater than or equal to a preset variance threshold, the satellites within the coverage area of ​​the gateway station will be re-clustered. If the variance of the total data transmitted between each cluster head and the gateway station is less than a preset variance threshold, then no re-clustering will be performed.

[0011] In one possible implementation, the method also includes, When a satellite flies out of the coverage area of ​​the gateway station, the satellite that flies out of the gateway station will be removed from the corresponding cluster; When a satellite enters the coverage area of ​​the gateway station, the matching degree between the satellite entering the gateway station and each cluster head is calculated, and the satellite entering the gateway station is added to the cluster of the cluster head with the highest matching degree.

[0012] In a second aspect of this application, a dynamic satellite clustering device is provided. The device includes: The acquisition module is used to acquire the first satellite information, input the first satellite information into the conversion model, and extract the second satellite information; The prediction module is used to input the second satellite information into the cluster head determination model to predict the probability of each satellite as a cluster head, and select the cluster head based on the probability of each satellite as a cluster head and the preset cluster head probability threshold. The clustering module is used to calculate the matching degree between the satellite to be clustered and each cluster head, and to cluster the satellite to be clustered. The judgment module is used to determine whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station at a preset clustering time interval.

[0013] In a third aspect of this application, an electronic device is provided. The electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the program to implement the method described above.

[0014] In a fourth aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the method according to the first aspect of this application.

[0015] The dynamic satellite clustering method provided in this application involves acquiring first satellite information, inputting it into a conversion model to extract second satellite information, and then inputting this second satellite information into a cluster head determination model to predict the probability of each satellite becoming a cluster head. Cluster heads are selected based on the probability of each satellite becoming a cluster head and a preset cluster head probability threshold. Then, the matching degree between the satellite to be clustered and the cluster head is calculated, and the satellites to be clustered are clustered. At a preset clustering time interval, it is determined whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station. This application optimizes the cluster head selection and clustering process by introducing an information conversion model and a probability prediction mechanism. Simultaneously, the dynamic re-clustering mechanism ensures the continuous performance of the network under high load or changing environments, reducing communication latency and energy consumption.

[0016] It should be understood that the description in the Summary Section is not intended to limit the key or essential features of the embodiments of this application, nor is it intended to restrict the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0017] The above and other features, advantages, and aspects of the embodiments of this application will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein: Figure 1 A flowchart of a dynamic satellite clustering method according to an embodiment of this application; Figure 2 This is a schematic diagram of a cluster head determination model according to an embodiment of this application; Figure 3 This is a schematic diagram illustrating the training of a conversion model according to an embodiment of this application; Figure 4 This is a schematic diagram illustrating the coverage area of ​​a gateway station according to an embodiment of this application; Figure 5 A block diagram of a dynamic satellite clustering device according to an embodiment of this application; Figure 6 This is a schematic diagram of the structure of a terminal device or server suitable for implementing the embodiments of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, 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 only 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 are within the scope of protection of this disclosure.

[0019] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0020] Figure 1 A flowchart of a dynamic satellite clustering method according to an embodiment of this disclosure is shown. See also Figure 1 The method includes: S101, Obtain the first satellite information, input the first satellite information into the conversion model, and extract the second satellite information.

[0021] In some embodiments, satellites within the coverage area of ​​the gateway station are represented. The gateway station uses ephemeris data to obtain satellite information. First satellite information , can be represented as:

[0022] Among them, satellite First satellite information This represents n satellite parameters in sequence: satellite x-coordinate, satellite y-coordinate, satellite vertices, average signal-to-noise ratio of the input link, average signal-to-noise ratio of the output link, channel bandwidth, average connectivity, buffer utilization, and computational resource utilization. Then, the gateway station inputs the first satellite information into the conversion model, captures the feature information within the first satellite information, and finally outputs the corresponding second satellite information. , can be represented as:

[0023] Among them, satellite Second satellite information This represents the precise satellite parameter information generated by the transformation model, which can accurately cluster the satellites.

[0024] In addition, the transformation model includes, but is not limited to, the BERT model (Bidirectional Encoder Representation from Transformers), which is used to extract feature information from the first satellite information and provide a reliable basis for subsequent satellite clustering.

[0025] In this embodiment, key information from the first satellite information is intelligently extracted through a transformation model to obtain the second satellite information, avoiding interference from redundant data and improving the efficiency of satellite information processing.

[0026] Optionally, the first satellite information includes satellite horizontal coordinate, satellite vertical coordinate, satellite vertical coordinate, average signal-to-noise ratio of input link, average signal-to-noise ratio of output link, channel bandwidth, average connectivity, buffer utilization rate, and computing resource utilization rate.

[0027] The average signal-to-noise ratio of the input link is the average signal-to-noise ratio of the data received by the satellite from other satellites, and the average signal-to-noise ratio of the output link is the average signal-to-noise ratio of the data sent by the satellite to other satellites.

[0028] In this embodiment, the first satellite information contains various satellite-related information data, which provides a foundation for subsequent extraction of satellite feature information.

[0029] S102, input the second satellite information into the cluster head determination model to predict the probability of each satellite as a cluster head, and select the cluster head based on the probability of each satellite as a cluster head and the preset cluster head probability threshold.

[0030] In some embodiments, the gateway station will transmit second satellite information. The matrix MY is input into the cluster head determination model, which predicts the probability Z of each satellite being a cluster head. This can be expressed as: , in, For satellite The probability of being the cluster head. For satellite The probability of not being a cluster head. For satellite The probability of being the cluster head. For satellite The probability of not being a cluster head, ... For satellite The probability of being the cluster head. For satellite The probability of not being a cluster head. And where MY = [Y1, Y2, ..., Y...]. L ].

[0031] In some embodiments, cluster heads are selected based on the probability of each satellite acting as a cluster head and a preset cluster head probability threshold. If the satellite... If the probability of a satellite being a cluster head is less than or equal to a preset cluster head probability threshold, it will not be considered a cluster head. If the probability of a satellite being selected as a cluster head is greater than a preset cluster head probability threshold, then it is selected as a cluster head. For example, if the preset cluster head probability threshold is 0.5, and the satellite... The probability of being a cluster head is 0.3, which is less than the preset cluster head probability threshold. Not as a cluster head, if satellite The probability of being a cluster head is 0.7, which is greater than the preset cluster head probability threshold. As a cluster head.

[0032] In this embodiment, the probability of a satellite becoming a cluster head is predicted using a cluster head determination model. A satellite can only become a cluster head if its probability of becoming a cluster head is higher than the cluster head probability threshold. This achieves efficient, accurate, and low-risk cluster head election, reduces manual configuration costs, and lays a reliable foundation for subsequent satellite clustering.

[0033] In one possible implementation, the cluster head determination model consists of a deep learning module, a fully connected layer, and a binary classifier; The cluster head determination model was trained using the historical second satellite information dataset.

[0034] Figure 2 This is a schematic diagram of a cluster head determination model according to an embodiment of this application, as shown below. Figure 2 As shown: The cluster head determination model consists of a deep learning module, a fully connected layer, and a binary classifier. First, a matrix MY composed of information from the second satellite is input into the cluster head determination model, where the deep learning module extracts features from the matrix MY. This deep learning module includes, but is not limited to, a multilayer perceptron (MLP). Then, the fully connected layer integrates and transforms the extracted features. Finally, binary classifiers 1 through L classify and predict the features of each satellite, generating the probability Z of each satellite being a cluster head. Among them, binary classifiers include, but are not limited to, softmax classifiers.

[0035] In some embodiments, the cluster head determination model is obtained by supervised training using a historical second satellite information dataset, which includes, but is not limited to, historical second satellite information and the actual probability of its corresponding satellites being cluster heads.

[0036] In this embodiment, the cluster head determination model utilizes a deep learning module to adaptively learn the dynamic characteristics of the second satellite information, ensuring the reliability of cluster head probability prediction.

[0037] Optionally, training the transformation model includes, Obtain the first satellite information of the first sample to be clustered, as well as the first satellite information of the first sample cluster head and the second sample cluster head; The transformation model outputs the second satellite information of the first sample to be clustered, the first sample cluster head, and the second sample cluster head, respectively. The second satellite information of the first sample to be clustered, the first sample cluster head, and the second sample cluster head is input into the fusion model to predict the probability that the first sample to be clustered will be added to the cluster where the first sample cluster head is located and the probability that the first sample to be clustered will be added to the cluster where the second sample cluster head is located. The probability of adding the first sample to be clustered into the cluster where the first sample cluster head is located is compared with the actual probability of the first sample to be clustered into the cluster where the first sample cluster head is located, and the first loss is obtained. The probability of adding the first sample to be clustered into the cluster where the second sample cluster head is located is compared with the actual probability of the first sample to be clustered into the cluster where the second sample cluster head is located, and the second loss is obtained. The conversion model is trained using the first loss and the second loss.

[0038] Figure 3 This is a schematic diagram illustrating the training of a conversion model according to an embodiment of this application, as shown below. Figure 3 As shown: First, the first satellite information aX of the first sample to be clustered, the first sample cluster head, and the first satellite information of the second sample cluster head are respectively... , Inputting the data into the transformation model, the second satellite information aY of the first sample to be clustered, the first sample cluster head, and the second sample cluster head is obtained. , .in, Figure 3 The image shows the first satellite information aX of the first sample to be clustered, the first sample cluster head, and the first satellite information of the second sample cluster head. , The transformation model used for processing is the same model.

[0039] Optionally, the first satellite information aX of the first sample to be clustered, the first satellite information of the first sample cluster head and the second sample cluster head. , The same transformation model can be input separately for each sample. The transformation model then outputs the second satellite information (aY) of the first sample to be clustered, the first sample cluster head, and the second sample cluster head, respectively. , .

[0040] Specifically, the first satellite information aX of the first sample to be clustered is input into the transformation model, and the transformation model processes aX to generate the second satellite information aY of the first sample to be clustered.

[0041] Then, the first satellite information of the first sample cluster head will be... Input transformation model, thereby transforming the model to Processing is performed to generate the second satellite information for the first sample cluster head. .

[0042] Then, the first satellite information of the second sample cluster head will be... Input transformation model, thereby transforming the model to Processing is performed to generate the second satellite information for the second sample cluster head. .

[0043] Then, the second satellite information aY of the first sample to be clustered, the first sample cluster head, and the second sample cluster head are... , The input is fed into the fusion model to predict the probability that the first sample to be clustered will be added to the cluster containing the first sample's cluster head. The probability of the first sample to be assigned to a cluster being added to the cluster containing the cluster head of the second sample. The fusion model consists of a multilayer perceptron and a binary classifier. The multilayer perceptron is responsible for generating fusion information between the second satellite information of the first sample to be clustered and the second satellite information of the first sample cluster head or the second sample cluster head. The binary classifier is responsible for generating the predicted probability based on the fusion information.

[0044] In the preset samples, the first sample to be clustered is a cluster member of the cluster where the first sample cluster head is located. Therefore, the probability of the actual first sample to be clustered being added to the cluster where the first sample cluster head is located is... ´ The probability of actually joining the cluster containing the second sample cluster head is 1. ´ It is 0.

[0045] Finally, the probability of adding the first sample to be clustered into the cluster containing the first sample's cluster head. The probability of the first sample to be clustered being added to the cluster containing the first sample's cluster head. ´ The input is compared with the first loss function to obtain the first loss. At the same time, the probability of adding the first sample to be clustered into the cluster where the second sample's cluster head is located is also calculated. The probability of the first sample to be clustered being added to the cluster containing the cluster head of the second sample. ´ The first and second losses are compared to obtain the second loss, and the conversion model is trained based on the first and second losses.

[0046] In some embodiments, the loss function of the transformation model can be expressed as: , in, To transform the overall loss of the model, For the first loss function, For the second loss function, and The preset loss weights are used. The first and second loss functions include, but are not limited to, mean squared error loss and cross-entropy loss.

[0047] In this embodiment, during the training of the conversion model, the probability of the sample to be clustered being added to the cluster where the first sample cluster head is located and the cluster where the second sample cluster head is located is calculated by the fusion model, which forces the conversion model to learn to distinguish the differential features of different cluster heads, effectively improving the reliability of feature extraction of the conversion model.

[0048] S103, calculate the matching degree between the satellite to be clustered and the cluster head, and cluster the satellite to be clustered.

[0049] In this embodiment, by calculating the matching degree between the satellite to be clustered and the cluster head, the association between the satellite and the cluster head is optimized, which can effectively improve the transmission efficiency of data within the cluster after clustering.

[0050] Optionally, the matching degree between the satellite to be clustered and each cluster head is calculated, and the satellite to be clustered is clustered, including, Obtain the channel bandwidth and signal-to-noise ratio of the satellites to be clustered and each cluster head, and calculate the channel capacity between the satellites to be clustered and each cluster head according to Shannon's formula; The matching degree between the satellite to be clustered and each cluster head is calculated based on the channel capacity between the satellite to be clustered and each cluster head, and the relative distance between the satellite to be clustered and each cluster head. The matching degree between the satellites to be clustered and each cluster head is sorted, and the satellites to be clustered are added to the cluster head with the highest matching degree.

[0051] In some embodiments, satellites are obtained through a cluster head determination model. The CCP Each cluster head is denoted as... In addition, satellites There is also One satellite to be clustered, denoted as... .in ,and, First, the gateway station acquires the satellites to be clustered. and cluster head Channel bandwidth between (i.e., the usable frequency range of the signal, in Hz) and signal-to-noise ratio (That is, the ratio of average signal power to average noise power), the channel capacity between the satellites to be clustered and each cluster head can be calculated using Shannon's formula and expressed as: , in, For satellites to be clustered and cluster head The channel capacity between them (i.e., the theoretically maximum data rate with no error transmission, measured in bits per second). Then, the gateway station acquires the satellites to be clustered. and cluster head relative distance between Calculate the satellites to be clustered With cluster head The formula for calculating the matching degree between them is: , in, For satellites to be clustered With cluster head The degree of matching between them and These are the preset matching degree calculation parameters. The larger the value, the more satellites it indicates that need to be clustered. With cluster head The better the match, The smaller the value, the more likely it is to be clustered into satellites. With cluster head The less compatible the satellite is, the better. Therefore, the gateway station can add the satellite to be clustered to the cluster head with the highest matching degree, making it a cluster member of that cluster, based on the matching degree between the satellite to be clustered and each cluster head.

[0052] In this embodiment, the traditional single-distance clustering mode is broken through. By integrating Shannon's formula to dynamically calculate the real-time communication potential and matching degree between the satellite to be clustered and the cluster head, the accurate resource matching between the satellite to be clustered and the cluster head is achieved, which significantly reduces the need for repeated clustering triggered by insufficient communication quality.

[0053] S104, within a preset clustering time interval, determine whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station.

[0054] In this embodiment, by setting a preset clustering time interval, the real-time computing pressure on the gateway station is reduced, and the topology maintenance overhead of satellite clustering is significantly reduced.

[0055] Optionally, at a preset clustering time interval, a decision is made on whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station, including: Calculate the mean and variance of the total data transmission volume between each cluster head and the gateway station in turn; If the variance of the total data transmitted between each cluster head and the gateway station is greater than or equal to a preset variance threshold, then the satellites within the coverage area of ​​the gateway station will be re-clustered. If the variance of the total data transmitted between each cluster head and the gateway station is less than a preset variance threshold, then no re-clustering will be performed.

[0056] Figure 4 This is a schematic diagram illustrating the coverage area of ​​a gateway station according to an embodiment of this application, as shown below. Figure 4 As shown: A gateway station's coverage area includes multiple satellites. Due to the high-speed movement of these satellites, the satellites within the gateway station's coverage area are constantly changing. When a satellite moves out of the gateway station's coverage area, the connection will be interrupted. At the same time, when a new satellite enters the gateway station's coverage area, a connection needs to be established, resulting in a continuous change in the number of satellites within the gateway station's coverage area.

[0057] In some embodiments, each cluster head Total data transmission volume with gateway station The formula for calculating the mean is: , in, For each cluster head Total data transmission volume with gateway station The average. The total amount of data transmitted between each cluster head and the gateway station. The formula for calculating the variance is: , in, Total data transmission volume between each cluster head and gateway station The variance.

[0058] In some embodiments, at a preset clustering time interval, the gateway station transmits each cluster head and the total amount of data transmitted by the gateway station. variance The data is compared with a preset variance threshold. If the variance is less than the preset variance threshold, the data is not re-clustered. For example, if the preset variance threshold is set to 20, and the variance of the total transmitted data between each cluster head and the gateway station is 15, which is less than the preset variance threshold, then the data is not re-clustered. If the variance of the total transmitted data between each cluster head and the gateway station is 25, which is greater than the preset variance threshold, then the data is re-clustered.

[0059] In this embodiment, the system determines whether to re-cluster by checking whether the variance of the total data transmitted between each cluster head and the gateway station is less than a preset variance threshold. This reduces the clustering frequency of the satellite, saves the computing resources of the gateway station, and significantly improves the working efficiency of the gateway station.

[0060] Optionally, the method also includes, When a satellite flies out of the coverage area of ​​the gateway station, the satellite that flies out of the gateway station will be removed from the corresponding cluster; When a satellite enters the coverage area of ​​the gateway station, the matching degree between the satellite entering the gateway station and each cluster head is calculated, and the satellite entering the gateway station is added to the cluster of the cluster head with the highest matching degree.

[0061] In some embodiments, gateway station A monitors the movement of satellites within its coverage area in real time. If gateway station A detects that satellite B1 has left its coverage area, satellite B1 is removed from the cluster associated with gateway station A. If gateway station A detects that satellite B2 has entered its coverage area, the matching degree between satellite B2 and each cluster head managed within the coverage area of ​​gateway station A is calculated, and satellite B2 is added to the cluster of the cluster head with the highest matching degree. For example, if the cluster heads managed within the coverage area of ​​gateway station A include C1, C2, C3, C4, and C5, and the matching degrees between satellite B2 and cluster heads C1, C2, C3, C4, and C5 are 0.1, 0.2, 0.1, 0.3, and 0.8, respectively, then satellite B2 is added to the cluster containing cluster head C5.

[0062] In this embodiment, local clustering adjustments are performed when satellites enter or leave the coverage area of ​​the gateway station. This avoids the large amount of computational operations caused by frequent clustering, reduces the workload and operating pressure of the gateway station, and ensures the relative stability of the network topology over a longer period of time, thereby ensuring reliable data transmission.

[0063] According to the embodiments of this disclosure, the following technical effects are achieved: 1) By using a cluster head determination model to predict the probability of each satellite becoming a cluster head, and combining this with a preset threshold to filter cluster heads, the accuracy and reliability of cluster head selection are improved, communication conflicts are reduced, and the throughput of the satellite network is increased.

[0064] 2) Clustering is performed by calculating the matching degree between the satellites to be clustered and the cluster head, which optimizes resource allocation, reduces data transmission path length and latency, and improves the overall network performance.

[0065] 3) Within a preset time interval, the system determines whether to re-cluster based on the total amount of data transmitted between each cluster head and gateway station. This avoids the resource consumption caused by frequent clustering calculations between cluster heads and gateway stations, enables real-time response to changes in network load, and enhances network reliability.

[0066] 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 this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0067] The above is an introduction to the method embodiments. The following describes the solution described in this application through device embodiments.

[0068] Figure 5 A block diagram of a dynamic satellite clustering apparatus according to an embodiment of this application is shown, such as... Figure 5 The following are included: The acquisition module 501 is used to acquire first satellite information, input the first satellite information into the conversion model, and extract second satellite information; The prediction module 502 is used to input the second satellite information into the cluster head determination model to predict the probability of each satellite as a cluster head, and select the cluster head based on the probability of each satellite as a cluster head and the preset cluster head probability threshold. Clustering module 503 is used to calculate the matching degree between the satellite to be clustered and each cluster head, and to cluster the satellite to be clustered. The judgment module 504 is used to determine whether to re-cluster based on the total amount of data transmitted between each cluster head and the gateway station at a preset clustering time interval.

[0069] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the described module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0070] Figure 6 A schematic diagram of a terminal device or server suitable for implementing embodiments of this application is shown.

[0071] like Figure 6 As shown, the terminal device or server includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from storage section 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the terminal device or server. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0072] The following components are connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 610 as needed so that computer programs read from it can be installed into storage section 608 as needed.

[0073] Specifically, according to embodiments of this application, the above method flow steps can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a machine-readable medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611. When the computer program is executed by central processing unit (CPU) 601, it performs the functions defined in the system of this application.

[0074] It should be noted that the computer-readable medium shown in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0075] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0076] The units or modules described in the embodiments of this application can be implemented in software or hardware. The described units or modules can also be located in a processor. The names of these units or modules do not, in certain circumstances, constitute a limitation on the unit or module itself.

[0077] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium stores one or more programs that, when used by one or more processors, execute the methods described in this application.

[0078] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing application concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions claimed in this application.

Claims

1. A dynamic satellite clustering method, characterized in that, The method comprises: obtaining first satellite information, inputting the first satellite information into a conversion model, and extracting second satellite information; inputting the second satellite information into a cluster head determination model to predict the probability of each satellite acting as a cluster head, and selecting a cluster head according to the probability of each satellite acting as a cluster head and a preset cluster head probability threshold; calculating the matching degree between the to-be-clustered satellite and each cluster head, and clustering the to-be-clustered satellite; judging whether to re-cluster according to the total transmission data amount of each cluster head and the gateway station in a preset clustering time interval.

2. The dynamic satellite clustering method of claim 1, wherein, The first satellite information comprises satellite horizontal coordinates, satellite vertical coordinates, input link average signal-to-noise ratio, output link average signal-to-noise ratio, channel bandwidth, average connectivity, buffer occupancy rate, and computing resource utilization rate.

3. The dynamic satellite clustering method of claim 1, wherein, The cluster head determination model is composed of a deep learning module, a full connection layer, and a binary classifier. The cluster head determination model is trained by historical second satellite information data sets.

4. The dynamic satellite clustering method of claim 1, wherein, The training of the conversion model comprises: obtaining the first satellite information of a first to-be-clustered sample and the first satellite information of a first sample cluster head and a second sample cluster head; outputting the second satellite information of the first to-be-clustered sample, the first sample cluster head, and the second sample cluster head through the conversion model respectively; inputting the second satellite information of the first to-be-clustered sample, the first sample cluster head, and the second sample cluster head into a fusion model to predict the probability of the first to-be-clustered sample joining the cluster in which the first sample cluster head is located and the probability of the first to-be-clustered sample joining the cluster in which the second sample cluster head is located; comparing the probability of the first to-be-clustered sample joining the cluster in which the first sample cluster head is located with the actual probability of the first to-be-clustered sample joining the cluster in which the first sample cluster head is located to obtain a first loss; comparing the probability of the first to-be-clustered sample joining the cluster in which the second sample cluster head is located with the actual probability of the first to-be-clustered sample joining the cluster in which the second sample cluster head is located to obtain a second loss; training the conversion model according to the first loss and the second loss.

5. The dynamic satellite clustering method of claim 1, wherein, The method comprises: calculating the matching degree between the to-be-clustered satellite and each cluster head, and clustering the to-be-clustered satellite, comprises: obtaining the channel bandwidth and signal-to-noise ratio of the to-be-clustered satellite and each cluster head, and calculating the channel capacity between the to-be-clustered satellite and each cluster head according to the Shannon formula respectively; calculating the matching degree between the to-be-clustered satellite and each cluster head according to the channel capacity between the to-be-clustered satellite and each cluster head and the relative distance between the to-be-clustered satellite and each cluster head respectively; 6. The dynamic satellite clustering method of claim 1, wherein, sorting the matching degree between the to-be-clustered satellite and each cluster head, and joining the to-be-clustered satellite into the cluster of the cluster head with the largest matching degree. The method comprises: calculating the mean and variance of the total transmission data amount of each cluster head and the gateway station in turn; if the variance of the total transmission data amount of each cluster head and the gateway station is greater than or equal to a preset variance threshold, re-clustering the satellites in the coverage range of the gateway station. If a variance of total transmission data amount of each cluster head and the gateway station is less than a preset variance threshold, no re-clustering is performed.

7. The dynamic satellite clustering method of claim 1, wherein, The method further comprises, When a satellite flies out of the coverage of the gateway station, the satellite flying out of the gateway station is removed from a corresponding cluster; When a satellite enters the coverage of the gateway station, a matching degree between the satellite entering the gateway station and each cluster head is calculated, and the satellite entering the gateway station is added to a cluster of the cluster head with which the matching degree is the largest.

8. A dynamic satellite clustering apparatus, characterized by, Comprise: An acquisition module, configured to acquire first satellite information, input the first satellite information into a conversion model, and extract second satellite information; A prediction module, configured to input the second satellite information into a cluster head determination model to predict a probability of each satellite as a cluster head, and select a cluster head according to the probability of each satellite as a cluster head and a preset cluster head probability threshold; A clustering module, configured to calculate a matching degree between a satellite to be clustered and each cluster head, and cluster the satellite to be clustered; A judgment module, configured to judge whether to re-cluster according to a total transmission data amount of each cluster head and a gateway station in a preset clustering time interval.

9. An electronic device comprising a memory and a processor, said memory having stored thereon a computer program, characterized in that, The processor implements the method of any one of claims 1-7 when executing the computer program.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the method of any one of claims 1-7 when executed by the processor.

Images