End-to-end network slice distribution method and system and communication system
Through the self-learning neural network model, the problem of low resource utilization in the existing technology is solved, and more efficient resource allocation and network performance optimization is achieved.
Patent Information
- Application Number
- CN202411013263.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-26
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-26
AI Technical Summary
When facing diversified terminal devices and complex application scenarios, the existing satellite network slice allocation method fails to fully consider the matching degree between terminal devices and slice resources, resulting in low resource utilization and poor network performance and user experience.
Using a neural network model with self-learning ability, extract feature information of terminal devices and network slices, generate feature matrix for matching, realize intelligent resource allocation, cover the resource management of access network, transmission network and core network, and generate slice query tables To optimize resource allocation.
It improves the flexibility and efficiency of resource allocation, avoids resource waste and unbalanced allocation, adapts to different types of terminal equipment and complex scenarios, and optimizes the overall network performance.
Smart Images

Figure CN118945732A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of mobile communication technologies, and in particular to a satellite network-based end-to-end network slice allocation method, system and communication system. Background Art
[0002] With the continuous development of mobile communication technology, especially the rise of 5G and future 6G networks, traditional network architectures have been unable to meet diverse business needs and efficient resource allocation requirements. Network slicing technology, as a key technology that can effectively solve this problem, has received widespread attention. Network slicing technology virtualizes the physical network infrastructure into multiple logical networks, each of which can be independently configured and managed according to specific business needs, thereby achieving efficient resource utilization and ensuring service quality.
[0003] At present, network slicing is mainly used in terrestrial communication networks. However, with the continuous increase in global communication needs and the rapid development of satellite communication technology, network slicing allocation methods based on satellite networks have gradually become a research hotspot. Satellite networks have advantages such as wide coverage and low transmission delay, and are particularly suitable for remote areas and areas such as the sea where terrestrial networks are difficult to cover. However, the existing satellite network slicing allocation methods still have many problems when facing diverse terminal devices and complex application scenarios.
[0004] Specifically, first, when a terminal requests slice resources, network slices are usually allocated based on fixed parameters, without considering the matching degree between the terminal device and the allocated slice. Secondly, the existing slice allocation method mainly focuses on the configuration of core network slices, ignoring the management and allocation of access network slice resources and transmission network slice resources, resulting in low overall network resource utilization. The resource allocation of access network and transmission network has an important impact on network performance and user experience. Ignoring the allocation of these resources may lead to network congestion, increased latency and other problems. Summary of the invention
[0005] The purpose of the present invention is to provide an end-to-end network slice allocation method, system and communication system based on a satellite network, which can allocate network slices to terminal devices by comprehensively considering the matching degree between each sub-slice resource in the network slice and the terminal device.
[0006] In order to achieve the above object, the present invention provides an end-to-end network slice allocation method based on a satellite network, wherein the network slice includes an access network sub-slice, a transmission network sub-slice and a core network sub-slice, and is characterized in that it includes:
[0007] Extracting characteristic information of the terminal device when registering to join the network to obtain a first characteristic parameter, and processing the first characteristic parameter based on a first processing model to generate a first characteristic matrix related to the characteristics of the terminal device;
[0008] Extracting feature information of the access network sub-slice, the transmission network sub-slice, and the core network sub-slice in each network slice respectively to obtain a second feature parameter, and processing the second feature parameter based on a second processing model to generate a second feature matrix related to the feature of the network slice;
[0009] The first processing model and the second processing model are both neural network models with self-learning capability;
[0010] For any of the first feature matrices, searching for a second feature matrix having a higher matching degree with the first feature matrix from among a plurality of second feature matrices to obtain a target feature matrix;
[0011] Generate and store matching data pairs to obtain a slice query table including a plurality of matching data pairs, wherein the matching data pairs include first feature parameters associated with the first feature matrix and identity information of a network slice associated with the target feature matrix corresponding to the first feature matrix;
[0012] When the slice requirement information of the terminal device is received, the slice query table is queried based on the first characteristic parameter of the terminal device to obtain a network slice that matches the current first characteristic parameter, and the network slice is allocated to the terminal device.
[0013] Preferably, the method for obtaining the target feature matrix includes:
[0014] Performing a dot product sum operation on the first feature matrix and the second feature matrix, and generating a prediction score according to the operation result;
[0015] For the first feature matrix, the second feature matrix with a higher prediction score is selected as the target feature matrix.
[0016] Preferably, a matching threshold is set, all the prediction scores exceeding the matching threshold are put into a candidate set, and the feature matrix corresponding to one of the candidate sets is randomly selected as the target feature matrix.
[0017] Preferably, the first characteristic parameter includes any one or more of an identity identifier, a geographical location, and a service type of the terminal device.
[0018] Preferably, the first processing model includes a first embedding layer and two first fully connected layers, and the first feature parameters pass through the first embedding layer and the two first fully connected layers in sequence to generate the first feature matrix.
[0019] Preferably, the second characteristic parameter includes a first characteristic sub-parameter related to characteristic information of the access network sub-slice, a second characteristic sub-parameter related to the transport network sub-slice, and a third characteristic sub-parameter related to the core network sub-slice;
[0020] The first characteristic sub-parameters include transmission rate, delay, single satellite capacity of the communication satellite and constellation capacity;
[0021] The second characteristic sub-parameters include bandwidth, delay, jitter time and packet loss rate;
[0022] The third characteristic sub-parameters include the number of users, delay, transmission rate and service type.
[0023] Preferably, the second processing model includes three independent second embedding layers, three independent second fully connected layers and one third fully connected layer;
[0024] Each of the second embedding layers is connected to the third fully connected layer through a second fully connected layer to form three sub-processing modules;
[0025] The first characteristic sub-parameter, the second characteristic sub-parameter, and the third characteristic sub-parameter are processed respectively by the three sub-processing modules, and the processing results are transmitted to the third fully connected layer. The data transmitted by the three sub-processing modules are processed by the third fully connected layer to generate the second characteristic matrix.
[0026] The present invention also provides a satellite network-based communication system for allocating end-to-end network slices, wherein the network slices include access network sub-slices, transmission network sub-slices, and core network sub-slices, and the communication system allocates the network slices to terminal devices based on the end-to-end network slice allocation method as described above.
[0027] The present invention also provides an end-to-end network slice allocation system, which includes:
[0028] one or more processors;
[0029] Memory;
[0030] and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, the programs including instructions for executing the end-to-end network slice allocation method as described above.
[0031] The present invention also provides a computer-readable storage medium, which includes a computer program, which can be executed by a processor to complete the end-to-end network slice allocation method as described above.
[0032] Compared with the prior art, the end-to-end network slice allocation method provided by the above technical solution of the present invention, firstly, by using a neural network model with self-learning ability, the characteristic parameters of the terminal device and the characteristic parameters of the network slice are processed and matched, and network resources can be intelligently allocated according to the changes in the characteristics of the terminal device, thereby improving the flexibility and efficiency of resource allocation, avoiding resource waste and uneven allocation, and being able to better adapt to different types of terminal devices and complex application scenarios; secondly, since the second characteristic parameter of the network slice covers the access network sub-slice, the transmission network sub-slice and the core network sub-slice, it realizes the end-to-end all-round management of network slice resources, which can more comprehensively optimize the allocation of network resources and improve the overall network performance. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 This is a flow chart of the end-to-end network slice allocation method in an embodiment of the present invention.
[0034] Figure 2 Schematic diagram of the working principle of the processing model in the embodiment of the present invention. DETAILED DESCRIPTION
[0035] In order to explain the technical content, structural features, achieved objectives and effects of the present invention in detail, the following is a detailed description in conjunction with the implementation methods and the accompanying drawings.
[0036] This embodiment discloses an end-to-end network slice allocation method for slice resource allocation based on a satellite network. The network slice in this embodiment includes an access network sub-slice, a transmission network sub-slice, and a core network sub-slice. When allocating slices to a terminal device, the matching of each sub-slice with the terminal device is fully considered so that the network slice allocated to the terminal device has a better matching degree with the terminal device.
[0037] Specifically, Figure 1 , the allocation method comprises the following steps:
[0038] S10: extracting characteristic information of the terminal device when registering to join the network to obtain a first characteristic parameter, and processing the first characteristic parameter based on a first processing model M1 to generate a first characteristic matrix related to the characteristics of the terminal device;
[0039] S11: extracting feature information of the access network sub-slice, the transmission network sub-slice, and the core network sub-slice in each network slice respectively to obtain a second feature parameter, and processing the second feature parameter based on a second processing model M2 to generate a second feature matrix related to the features of the network slice;
[0040] The first processing model M1 and the second processing model M2 are both neural network models with self-learning capabilities;
[0041] S12: for any of the first feature matrices, searching for a second feature matrix having a higher matching degree with the first feature matrix from among a plurality of second feature matrices to obtain a target feature matrix;
[0042] S13: Generate and store matching data pairs to obtain a slice query table including a plurality of matching data pairs, wherein the matching data pairs include a first feature parameter associated with the first feature matrix and identity information of the network slice associated with the target feature matrix corresponding to the first feature matrix;
[0043] S14: When receiving the slice requirement information of the terminal device, query the slice query table based on the first characteristic parameter of the terminal device to obtain a network slice matching the current first characteristic parameter;
[0044] S15: Allocate the queried network slice to the terminal device.
[0045] For example, for a certain terminal device U, when it registers to join the network, the first characteristic parameter C10 of the terminal device U is extracted, and the first characteristic matrix M10 is generated based on the first processing model M1.
[0046] Assume that there are three network slices currently available, and the second characteristic matrices generated based on the second characteristic parameters C20, C21, and C22 of these three network slices Q1, Q2, and Q3 are M20, M21, and M22 respectively.
[0047] Then, M10 is matched with M20, M21, and M22 respectively. If the matching degree between M10 and M21 is the highest, M21 is used as the target feature matrix, and a data pair with the terminal device U is generated: [C10, Q2], and the data pair is stored in the slice query table.
[0048] When a request for slice resources sent by the terminal device U is received, the slice query table is queried based on the first characteristic parameter C10 of the terminal device U to obtain the network slice Q2, and the network slice Q2 is allocated to the terminal device U.
[0049] For the slice allocation method in this embodiment, first, by using a neural network model with self-learning ability, the characteristic parameters of the terminal device and the characteristic parameters of the network slice are processed and matched, and network resources can be intelligently allocated according to changes in the characteristics of the terminal device, thereby improving the flexibility and efficiency of resource allocation, avoiding resource waste and uneven allocation, and better adapting to different types of terminal devices and complex application scenarios, such as ultra-long-distance real-time services, intensive connection services, high-speed mobile broadband services, large-capacity transmission services, and strong security services.
[0050] Secondly, since the second characteristic parameter of network slicing covers access network sub-slicing, transmission network sub-slicing and core network sub-slicing, it realizes all-round management of end-to-end network slicing resources, which can more comprehensively optimize network resource allocation and improve overall network performance.
[0051] On the other hand, please refer to Figure 1 and Figure 2 , the methods for obtaining the target feature matrix include:
[0052] Performing a dot product sum operation on the first feature matrix and the second feature matrix, and generating a prediction score according to the operation result;
[0053] For the first feature matrix, the second feature matrix with a higher prediction score is selected as the target feature matrix.
[0054] In this embodiment, the interaction relationship between the terminal device and the network slice can be effectively captured through the point product sum operation of the first characteristic matrix and the second characteristic matrix. The calculation process is simple and efficient, and is suitable for processing large-scale data.
[0055] Furthermore, the working parameters of the first processing model M1 and the second processing model M2 can be optimized by obtaining the difference between the actual score and the predicted score corresponding to each network slice used by the terminal device from a third party, thereby improving the accuracy of the first feature matrix and the second feature matrix.
[0056] On the other hand, a matching threshold is set, all the prediction scores exceeding the matching threshold are put into a candidate set, and the feature matrix corresponding to one of the candidate sets is randomly selected as the target feature matrix. In this embodiment, the utilization efficiency of the network slice can be effectively improved by setting the candidate set, and.
[0057] On the other hand, the first characteristic parameter includes any one or more of the identity identification, geographic location and service type of the terminal device.
[0058] When the identity identifier, geographic location and service type are used as the first characteristic parameter at the same time, the first characteristic parameter in the matching data pair only needs to record the identity identifier. Therefore, when querying the slice query table, the corresponding network slice can be quickly found through the identity identifier of the current terminal device.
[0059] It should be noted that if the terminal device does not perform well when using the currently allocated network slice, it can repeatedly request to update the network slice. At this time, the server will regenerate the first feature matrix based on the three parameters of the current terminal device's identity, geographic location, and service type, and re-match the network slice based on the first feature matrix, and update the matching results to the data pair in the slice query table.
[0060] like Figure 2 The first processing model M1 includes a first embedding layer and two first fully connected layers, and the first feature parameters pass through the first embedding layer and the two first fully connected layers in sequence to generate the first feature matrix.
[0061] In this embodiment, the first embedding layer extracts features of the three parameters of identity, geographic location and business type in the first feature parameters to generate corresponding features, and then generates a first feature matrix through processing by two fully connected layers.
[0062] On the other hand, the second characteristic parameter includes a first characteristic sub-parameter related to characteristic information of the access network sub-slice, a second characteristic sub-parameter related to the transport network sub-slice, and a third characteristic sub-parameter related to the core network sub-slice;
[0063] The first characteristic sub-parameters include transmission rate, delay, single satellite capacity of the communication satellite and constellation capacity;
[0064] The second characteristic sub-parameters include bandwidth, delay, jitter time and packet loss rate;
[0065] The third characteristic sub-parameters include the number of users, delay, transmission rate and service type.
[0066] The above parameters are described in detail below.
[0067] Transmission rate refers to the data transmission rate between the terminal device and the access node.
[0068] Latency refers to the time it takes for data to be transmitted from a terminal device to an access node and back, usually measured in milliseconds (ms).
[0069] Single-satellite capacity refers to the maximum amount of data transmission that a single communications satellite can handle.
[0070] Constellation capacity refers to the total data transmission that the entire satellite constellation (a system consisting of multiple satellites) can provide.
[0071] Bandwidth refers to the maximum amount of data transmission that can be provided in a transmission network, usually measured in Mbps or Gbps.
[0072] Jitter time refers to the variation in the time interval between the arrival of data packets, usually measured in milliseconds.
[0073] Packet loss rate refers to the proportion of data packets lost during data transmission, usually expressed as a percentage.
[0074] Service type refers to the different service types supported by the core network, such as voice, video, data transmission, IoT, etc.
[0075] On the other hand, Figure 2 , the second processing model M2 includes three independent second embedding layers, three independent second fully connected layers and one third fully connected layer.
[0076] Each of the second embedding layers is connected to the third fully connected layer through a second fully connected layer to form three sub-processing modules.
[0077] The first characteristic sub-parameter, the second characteristic sub-parameter, and the third characteristic sub-parameter are processed respectively by the three sub-processing modules, and the processing results are transmitted to the third fully connected layer. The data transmitted by the three sub-processing modules are processed by the third fully connected layer to generate the second characteristic matrix.
[0078] In this embodiment, the characteristic information of the access network sub-slice, the transmission network sub-slice and the core network sub-slice are processed respectively through three sub-processing modules, and the respective processing results are uniformly transmitted to the third fully connected layer, and then the second characteristic matrix related to the access network sub-slice, the transmission network sub-slice and the core network sub-slice is output through the third fully connected layer.
[0079] In another preferred embodiment of the present invention, a satellite network-based communication system is also disclosed for allocating end-to-end network slices, wherein the network slices include access network sub-slices, transmission network sub-slices, and core network sub-slices, and is characterized in that the communication system allocates the network slices to the terminal device based on the end-to-end network slice allocation method in the above embodiment.
[0080] The present invention also discloses an end-to-end network slice allocation system, which includes one or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the program includes instructions for executing the end-to-end network slice allocation method as described above. The processor can adopt a general central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits to execute relevant programs to implement the functions required to be executed by the modules in the end-to-end network slice allocation system of the embodiment of the present application, or to execute the end-to-end network slice allocation method of the method embodiment of the present application.
[0081] The present invention also discloses a computer-readable storage medium, which includes a computer program, and the computer program can be executed by a processor to complete the end-to-end network slice allocation method as described above. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium can be a read-only memory (ROM), or a random access memory (RAM), or a magnetic medium, such as a floppy disk, a hard disk, a tape, a magnetic disk, or an optical medium, such as a digital versatile disc (DVD), or a semiconductor medium, such as a solid state disk (SSD), etc.
[0082] The embodiment of the present application also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of an electronic device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the electronic device performs the above-mentioned end-to-end network slice allocation method.
[0083] The above disclosure is only the preferred embodiment of the present invention, which certainly cannot be used to limit the scope of rights of the present invention. Therefore, equivalent changes made according to the scope of the patent application of the present invention are still within the scope covered by the present invention.
Claims
1. A method for allocating end-to-end network slices based on a satellite network, wherein the network slices include access network slices, transmission network slices, and core network slices, characterized in that: include: Extracting characteristic information of the terminal device when registering to join the network to obtain a first characteristic parameter, and processing the first characteristic parameter based on a first processing model to generate a first characteristic matrix related to the characteristics of the terminal device; Extracting feature information of the access network sub-slice, the transmission network sub-slice, and the core network sub-slice in each network slice respectively to obtain a second feature parameter, and processing the second feature parameter based on a second processing model to generate a second feature matrix related to the feature of the network slice; The first processing model and the second processing model are both neural network models with self-learning capability; For any of the first feature matrices, searching for a second feature matrix having a higher matching degree with the first feature matrix from among a plurality of second feature matrices to obtain a target feature matrix; Generate and store matching data pairs to obtain a slice query table including a plurality of matching data pairs, wherein the matching data pairs include a first feature parameter associated with the first feature matrix and identity information of the network slice associated with the target feature matrix corresponding to the first feature matrix; When the slice requirement information of the terminal device is received, the slice query table is queried based on the first characteristic parameter of the terminal device to obtain a network slice that matches the current first characteristic parameter, and the network slice is allocated to the terminal device.
2. The end-to-end network slice allocation method based on a satellite network according to claim 1, characterized in that: Methods for obtaining the target feature matrix include: Performing a dot product sum operation on the first feature matrix and the second feature matrix, and generating a prediction score according to the operation result; For the first feature matrix, the second feature matrix with a higher prediction score is selected as the target feature matrix.
3. The end-to-end network slice allocation method based on satellite network according to claim 2, characterized in that: A matching threshold is set, all the prediction scores exceeding the matching threshold are put into a candidate set, and the feature matrix corresponding to one of the candidate sets is randomly selected as the target feature matrix.
4. The end-to-end network slice allocation method based on a satellite network according to claim 1, characterized in that: The first characteristic parameter includes any one or more of the identity identification, geographic location and service type of the terminal device.
5. The end-to-end network slice allocation method based on satellite network according to claim 4, characterized in that: The first processing model includes a first embedding layer and two first fully connected layers, and the first feature parameters are sequentially passed through the first embedding layer and the two first fully connected layers to generate the first feature matrix.
6. The end-to-end network slice allocation method based on satellite network according to claim 1, characterized in that: The second characteristic parameter includes a first characteristic sub-parameter related to characteristic information of the access network sub-slice, a second characteristic sub-parameter related to the transport network sub-slice, and a third characteristic sub-parameter related to the core network sub-slice; The first characteristic sub-parameters include transmission rate, delay, single satellite capacity of the communication satellite and constellation capacity; The second characteristic sub-parameters include bandwidth, delay, jitter time and packet loss rate; The third characteristic sub-parameters include the number of users, delay, transmission rate and service type.
7. The end-to-end network slice allocation method based on satellite network according to claim 6, characterized in that: The second processing model includes three independent second embedding layers, three independent second fully connected layers, and one third fully connected layer; Each of the second embedding layers is connected to the third fully connected layer through a second fully connected layer to form three sub-processing modules; The first characteristic sub-parameter, the second characteristic sub-parameter, and the third characteristic sub-parameter are processed respectively by the three sub-processing modules, and the processing results are transmitted to the third fully connected layer. The data transmitted by the three sub-processing modules are processed by the third fully connected layer to generate the second characteristic matrix.
8. A satellite network-based communication system for allocating end-to-end network slices, wherein the network slices include access network sub-slices, transport network sub-slices, and core network sub-slices, characterized in that: The communication system allocates the network slice to the terminal device based on the end-to-end network slice allocation method described in any one of claims 1 to 7.
9. An end-to-end network slice allocation system, characterized in that: include: one or more processors; Memory; And one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the programs include instructions for executing the end-to-end network slice allocation method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: It includes a computer program that can be executed by a processor to complete the end-to-end network slice allocation method as described in any one of claims 1 to 7.
Citation Information
Patent Citations
Method and device for configuring network slices for terminal equipment
CN113133087A
Communication network slice distribution method, computer device and storage medium
CN114828261A
Network slice resource allocation method and device and storage medium
CN117014057A
Operation and maintenance method, apparatus and system for network slice, and storage medium
WO2022121735A1
