Network policy enhancement method, policy control function entity and communication system

Abstract

The present disclosure provides a network policy enhancement method, a policy control function entity and a communication system. The network policy enhancement method comprises: receiving terminal capability information reported by a user terminal through an AMF entity; receiving analysis information related to service execution efficiency of the user terminal sent by an NWDAF entity; generating a network policy related to AI and ML according to the terminal capability, the analysis information related to service execution efficiency and current network resources; and sending the network policy or a policy identifier corresponding to the network policy to the user terminal through the AMF entity, so that the user terminal performs access and mobility management according to the network policy or the network policy corresponding to the policy identifier. The present disclosure provides the user terminal with a network policy corresponding to the terminal capability, thereby enhancing the overall performance of the network.

Description

Technical Field

[0001] This disclosure relates to the field of communications, and in particular to a network policy enhancement method, a policy control functional entity, and a communication system. Background Technology

[0002] After achieving centralized network intelligence, 3GPP (3rd Generation Partnership Project) SA2 launched AI (Artificial Intelligence) and ML (Machine Learning) projects in R18, enabling 5GS to support distributed network intelligence and federated learning. Summary of the Invention

[0003] The inventors noted that existing network policy management primarily targets data transmission scenarios. The network issues policies such as AM (Access Management) and SM (Session Management) (including RFSP (RAT / Frequency Selection Priority), URSP (UE Route Selection Policy), and QoS (Quality of Service)) to base stations and terminals based on terminal subscriptions and session information. However, in AI / ML scenarios, training models need to be transferred between intelligent terminals and the network. Larger training models improve training accuracy but consume more communication resources. After reaching a certain level of accuracy, the accuracy gain from more complex training models gradually diminishes, thus failing to enhance overall network performance.

[0004] Accordingly, this disclosure provides a network policy enhancement scheme that can effectively enhance the overall network performance.

[0005] According to a first aspect of the present disclosure, a network policy enhancement method is provided, comprising: receiving terminal capability information reported by a user terminal through an Access and Mobility Management Function (AMF) entity; receiving service execution efficiency-related analysis information of the user terminal sent by a Network Data Analysis Function (NWDAF) entity; generating a network policy related to Artificial Intelligence (AI) and Machine Learning (ML) based on the terminal capability, the service execution efficiency-related analysis information, and current network resources; and sending the network policy or a policy identifier corresponding to the network policy to the user terminal through the AMF entity, so that the user terminal performs access and mobility management according to the network policy or the network policy corresponding to the policy identifier.

[0006] In some embodiments, the terminal capability information includes at least one of the following: the computing power of the user terminal, the memory size of the user terminal, the algorithms supported by the user terminal, and the intelligent services supported by the user terminal.

[0007] In some embodiments, the computing power includes at least one of the following: floating-point operations per second, the clock speed and number of cores of the CPU and GPU of the user terminal, and graph computing capability; the algorithm includes at least one of deep neural network algorithm and distributed learning algorithm; and the intelligent service includes at least one of image recognition, remote automated control, and autonomous driving.

[0008] In some embodiments, the network strategy includes at least one of the following: an operation partitioning scheme for AI and ML models, adaptive quality of service for transmitting AI and ML models and data, group quality of service, end-to-end quality of service, and a determined distributed learning scheme.

[0009] In some embodiments, the operation partitioning scheme includes a hierarchical scheme for AI and ML models; the adaptive quality of service is used to meet the needs of the distributed learning scheme; the group quality of service is used to limit the total resource consumption for AI and ML services; the end-to-end quality of service is used to ensure the end-to-end communication quality of the distributed learning scheme; the determined distributed learning scheme includes at least one of determining the terminals participating in the distributed learning and the learning task allocation strategy.

[0010] In some embodiments, the requirements of the distributed learning scheme include the timeliness requirements of the distributed learning scheme; the distributed learning scheme includes a federated learning scheme.

[0011] In some embodiments, the network policy or the policy identifier corresponding to the network policy is sent to the user plane function (UPF) entity through the session management function (SMF) entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

[0012] According to a second aspect of the present disclosure, a policy control function entity is provided, comprising: a first processing module configured to receive terminal capability information reported by a user terminal through an Access and Mobility Management Function (AMF) entity; a second processing module configured to receive service execution efficiency-related analysis information of the user terminal sent by a Network Data Analysis Function (NWDAF) entity; a third processing module configured to generate a network policy related to Artificial Intelligence (AI) and Machine Learning (ML) based on the terminal capabilities, the service execution efficiency-related analysis information, and current network resources; and a fourth processing module configured to send the network policy or a policy identifier corresponding to the network policy to the user terminal through the AMF entity, so that the user terminal performs access and mobility management according to the network policy or the network policy corresponding to the policy identifier.

[0013] In some embodiments, the terminal capability information includes at least one of the following: the computing power of the user terminal, the memory size of the user terminal, the algorithms supported by the user terminal, and the intelligent services supported by the user terminal.

[0014] In some embodiments, the computing power includes at least one of the following: floating-point operations per second, the clock speed and number of cores of the CPU and GPU of the user terminal, and graph computing capability; the algorithm includes at least one of deep neural network algorithm and distributed learning algorithm; and the intelligent service includes at least one of image recognition, remote automated control, and autonomous driving.

[0015] In some embodiments, the network strategy includes at least one of the following: an operation partitioning scheme for AI and ML models, adaptive quality of service for transmitting AI and ML models and data, group quality of service, end-to-end quality of service, and a determined distributed learning scheme.

[0016] In some embodiments, the operation partitioning scheme includes a hierarchical scheme for AI and ML models; the adaptive quality of service is used to meet the needs of the distributed learning scheme; the group quality of service is used to limit the total resource consumption for AI and ML services; the end-to-end quality of service is used to ensure the end-to-end communication quality of the distributed learning scheme; the determined distributed learning scheme includes at least one of determining the terminals participating in the distributed learning and the learning task allocation strategy.

[0017] In some embodiments, the requirements of the distributed learning scheme include the timeliness requirements of the distributed learning scheme; the distributed learning scheme includes a federated learning scheme.

[0018] In some embodiments, the fourth processing module is configured to send the network policy or the policy identifier corresponding to the network policy to the user plane function (UPF) entity through the session management function (SMF) entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

[0019] According to a third aspect of the present disclosure, a policy control function entity is provided, comprising: a memory configured to store instructions; and a processor coupled to the memory, the processor being configured to execute instructions stored in the memory to implement the method as described in any of the above embodiments.

[0020] According to a fourth aspect of the present disclosure, a communication system is provided, comprising: a policy control function (PCF) entity as described in any of the above embodiments; a user terminal configured to send terminal capability information to the PCF entity through an access and mobility management function (AMF) entity, and receive network policies or policy identifiers related to artificial intelligence (AI) and machine learning (ML) sent by the PCF entity through the AMF entity, so as to perform access and mobility management according to the network policies or network policies corresponding to the policy identifiers; and a network data analysis function (NWDAF) entity configured to collect statistical analysis information related to the service execution efficiency of the user terminal and send the statistical results to the PCF entity.

[0021] In some embodiments, the system further includes a User Plane Function (UPF) entity configured to receive the network policy or policy identifier through a Session Management Function (SMF) entity, so that the UPF entity performs routing management according to the network policy or the network policy corresponding to the policy identifier.

[0022] According to a fifth aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method as described in any of the above embodiments.

[0023] Other features and advantages of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1This is a flowchart illustrating a network policy enhancement method according to an embodiment of the present disclosure;

[0026] Figure 2 This is a schematic diagram of the structure of a PCF entity according to an embodiment of this disclosure;

[0027] Figure 3 This is a schematic diagram of the structure of a PCF entity according to another embodiment of this disclosure;

[0028] Figure 4 This is a schematic diagram of the structure of a communication system according to an embodiment of the present disclosure;

[0029] Figure 5 This is a schematic diagram of the structure of a communication system according to another embodiment of the present disclosure;

[0030] Figure 6 This is a flowchart illustrating a network policy enhancement method according to another embodiment of this disclosure. Detailed Implementation

[0031] 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. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this disclosure or its application or use. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0032] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of this disclosure.

[0033] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0034] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0035] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0036] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0037] The inventors noted the following shortcomings in existing network policy management:

[0038] Smart terminals, in their relentless pursuit of higher accuracy in intelligent analysis, consume a large amount of communication resources, while the resulting gains in accuracy are limited.

[0039] • When multiple smart terminals perform intelligent analysis / federated learning simultaneously, the resources occupied by each terminal are not excessive, but when all terminals are combined, they occupy too much communication resources, affecting other network services (such as voice, data, etc.).

[0040] • The network allocates communication resources evenly to terminals with different intelligence capabilities, resulting in low utilization of communication resources, which will reduce the overall performance of network communication resources.

[0041] Accordingly, this disclosure provides a network policy enhancement scheme that can effectively enhance the overall network performance.

[0042] Figure 1 This is a flowchart illustrating a network policy enhancement method according to an embodiment of the present disclosure. In some embodiments, the following network policy enhancement method is executed by a PCF (Policy Control Function) entity.

[0043] In step 101, the terminal capability information reported by the user terminal is received through the AMF (Access and Mobility Management Function) entity.

[0044] In some embodiments, the terminal capability information includes at least one of the following:

[0045] - Computing power of the user terminal: including at least one of the following: floating-point operations per second, clock speed and number of cores of the user terminal's CPU (Central Processing Unit) and GPU (Graphics Processing Unit), and graph processing capability.

[0046] -User terminal memory size

[0047] - Algorithms supported by the user terminal: for example, at least one of deep neural network algorithms and distributed learning algorithms.

[0048] - Intelligent services supported by the user terminal: such as at least one of image recognition, remote automation control, and autonomous driving.

[0049] In step 102, receive analysis information related to the service execution efficiency of user terminals sent by the NWDAF (Network Data Analytics Function) entity.

[0050] For example, the NWDAF entity obtains the historical data of the user terminal through OSS (Operation Support System) or BSS (Business Support System), and then determines the analytical information related to the business execution efficiency of the user terminal based on the historical data of the user terminal.

[0051] In step 103, based on the analysis information related to terminal capabilities, service execution efficiency, and current network resources, network policies related to AI and ML are generated.

[0052] In some embodiments, AI and ML-related network strategies include at least one of the following:

[0053] - Operational partitioning schemes for AI and ML models: including hierarchical schemes for AI and ML models.

[0054] - Adaptive Quality of Service (QoS) for transmitting AI and ML models and data: This is used to meet the needs of distributed learning solutions, such as their timeliness requirements.

[0055] - Group Service Quality: Used to limit the total resource consumption for AI and ML operations.

[0056] - Round Trip Service Quality: Used to ensure the quality of communication throughout the distributed learning scheme.

[0057] - The determined distributed learning scheme includes at least one of the following: determining the terminals participating in distributed learning and the learning task allocation strategy.

[0058] In some embodiments, distributed learning schemes include federated learning schemes.

[0059] In step 104, the network policy or the policy identifier corresponding to the network policy is sent to the user terminal through the AMF entity so that the user terminal can perform access and mobility management according to the network policy or the network policy corresponding to the policy identifier.

[0060] In the network policy enhancement method provided in the above embodiments of this disclosure, network policies related to AI and ML are generated based on the terminal capabilities of the user terminal, analysis information related to service execution efficiency, and current network resources, and the network policies are provided to the user terminal, thereby effectively enhancing the overall network performance.

[0061] In some embodiments, the network policy or the policy identifier corresponding to the network policy is sent to the UPF (User Plane Function) entity through the SMF (Session Management Function) entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

[0062] Figure 2 This is a schematic diagram of the structure of a PCF entity according to an embodiment of this disclosure. Figure 2 As shown, the PCF entity includes a first processing module 21, a second processing module 22, a third processing module 23, and a fourth processing module 24.

[0063] The first processing module 21 is configured to receive terminal capability information reported by user terminals through the AMF entity.

[0064] In some embodiments, the terminal capability information includes at least one of the following:

[0065] - Computing power of the user terminal: including at least one of the following: floating-point operations per second, CPU and GPU clock speeds and number of cores, and graph processing capabilities.

[0066] -User terminal memory size

[0067] - Algorithms supported by the user terminal: for example, at least one of deep neural network algorithms and distributed learning algorithms.

[0068] - Intelligent services supported by the user terminal: such as at least one of image recognition, remote automation control, and autonomous driving.

[0069] The second processing module 22 is configured to receive analysis information related to the service execution efficiency of user terminals sent by the NWDAF entity.

[0070] For example, the NWDAF entity obtains the historical data of the user terminal through OSS or BSS, and then determines the analytical information related to the service execution efficiency of the user terminal based on the historical data.

[0071] The third processing module 23 is configured to generate network policies related to AI and ML based on analysis information related to terminal capabilities, service execution efficiency, and current network resources.

[0072] In some embodiments, AI and ML-related network strategies include at least one of the following:

[0073] - Operational partitioning schemes for AI and ML models: including hierarchical schemes for AI and ML models.

[0074] - Adaptive Quality of Service (QoS) for transmitting AI and ML models and data: This is used to meet the needs of distributed learning solutions, such as their timeliness requirements.

[0075] - Group Service Quality: Used to limit the total resource consumption for AI and ML operations.

[0076] - Round Trip Service Quality: Used to ensure the quality of communication throughout the distributed learning scheme.

[0077] - The determined distributed learning scheme includes at least one of the following: determining the terminals participating in distributed learning and the learning task allocation strategy.

[0078] In some embodiments, distributed learning schemes include federated learning schemes.

[0079] The fourth processing module 24 is configured to send network policies or policy identifiers corresponding to network policies to user terminals through the AMF entity, so that user terminals can perform access and mobility management according to the network policies or network policies corresponding to policy identifiers.

[0080] In some embodiments, the fourth processing module 24 is configured to send a network policy or a policy identifier corresponding to the network policy to a UPF entity through an SMF entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

[0081] Figure 3 This is a schematic diagram of the structure of a PCF entity according to another embodiment of this disclosure. Figure 3 As shown, the PCF entity includes a memory 31 and a processor 32.

[0082] Memory 31 is used to store instructions, and processor 32 is coupled to memory 31. Processor 32 is configured to execute instructions based on memory storage, as shown in the example. Figure 1 The method involved in any of the embodiments.

[0083] like Figure 3 As shown, the PCF entity also includes a communication interface 33 for exchanging information with other devices. Additionally, the PCF entity includes a bus 34, through which the processor 32, communication interface 33, and memory 31 communicate with each other.

[0084] The memory 31 may include high-speed RAM, and may also include non-volatile memory, such as at least one disk storage device. The memory 31 may also be a memory array. The memory 31 may also be divided into blocks, and the blocks may be combined into virtual volumes according to certain rules.

[0085] Furthermore, processor 32 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present disclosure.

[0086] This disclosure also relates to a computer-readable storage medium storing computer instructions that, when executed by a processor, implement... Figure 1 The method involved in any of the embodiments.

[0087] Figure 4 This is a schematic diagram of the structure of a communication system according to an embodiment of this disclosure. Figure 4 As shown, the communication system includes user terminal 41, AMF entity 42, PCF entity 43, and NWDAF entity 44. PCF entity 43 is... Figure 2 or Figure 3 The PCF entity involved in any of the embodiments.

[0088] User terminal 41 sends terminal capability information to PCF entity 43 via AMF entity 42. NWDAF entity 44 collects analysis information related to the service execution efficiency of the user terminal and sends the statistical results to PCF entity 43. User terminal 41 receives network policies or policy identifiers related to AI and ML sent by PCF entity 43 via AMF entity 42, so as to perform access and mobility management according to the network policies or network policies corresponding to the policy identifiers.

[0089] Figure 5 This is a schematic diagram of the structure of a communication system according to another embodiment of the present disclosure. Figure 5 and Figure 4 The difference is that, in Figure 5 In the illustrated embodiment, the communication system further includes SMF entity 45 and UPF entity 46.

[0090] UPF entity 46 receives network policies or policy identifiers sent by PCF entity 43 through SMF entity 45, so that UPF entity 46 can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

[0091] Figure 6 This is a flowchart illustrating a network policy enhancement method according to another embodiment of this disclosure.

[0092] In step 601, the user terminal reports the terminal capability information to the AMF entity.

[0093] In some embodiments, the terminal capability information includes at least one of the following:

[0094] - Computing power of the user terminal: including at least one of the following: floating-point operations per second, CPU and GPU clock speeds and number of cores, and graph processing capabilities.

[0095] -User terminal memory size

[0096] - Algorithms supported by the user terminal: for example, at least one of deep neural network algorithms and distributed learning algorithms.

[0097] - Intelligent services supported by the user terminal: such as at least one of image recognition, remote automation control, and autonomous driving.

[0098] In step 602, the AMF entity reports the terminal capability information to the PCF entity.

[0099] In step 603, the NWDAF entity obtains the historical information of the user terminal through the OSS / BSS system.

[0100] In step 604, the NWDAF entity determines the analysis information related to the service execution efficiency of the user terminal based on the historical information of the user terminal, and sends the analysis information related to the service execution efficiency to the PCF entity.

[0101] In step 605, the PCF entity generates network policies related to AI and ML based on analysis information related to terminal capabilities, service execution efficiency, and current network resources.

[0102] In some embodiments, AI and ML-related network strategies include at least one of the following:

[0103] - Operational partitioning schemes for AI and ML models: including hierarchical schemes for AI and ML models.

[0104] - Adaptive Quality of Service (QoS) for transmitting AI and ML models and data: This is used to meet the needs of distributed learning solutions, such as their timeliness requirements.

[0105] - Group Service Quality: Used to limit the total resource consumption for AI and ML operations.

[0106] - Round Trip Service Quality: Used to ensure the quality of communication throughout the distributed learning scheme.

[0107] - The determined distributed learning scheme includes at least one of the following: determining the terminals participating in distributed learning and the learning task allocation strategy.

[0108] In some embodiments, distributed learning schemes include federated learning schemes.

[0109] In step 606, the PCF entity sends the network policy or the policy identifier corresponding to the network policy to the AMF entity.

[0110] In step 607, the AMF entity sends the network policy or the policy identifier corresponding to the network policy to the user terminal so that the user terminal can perform access and mobility management according to the network policy or the network policy corresponding to the policy identifier.

[0111] In step 608, the PCF entity sends the network policy or the policy identifier corresponding to the network policy to the SMF entity.

[0112] In step 609, the SMF entity sends the network policy or the policy identifier corresponding to the network policy to the UPF entity so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

[0113] The following beneficial effects can be obtained through the above embodiments of this disclosure:

[0114] Compared to existing AI / ML solutions that schedule all terminals, this disclosure configures different intelligent management strategies for terminals with different intelligent capabilities, thereby enhancing overall network performance and increasing overall network intelligence benefits.

[0115] In some embodiments, the functional units described above may be implemented as general-purpose processors, programmable logic controllers (PLCs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or any suitable combination thereof for performing the functions described herein.

[0116] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0117] The description in this disclosure is provided for illustrative and descriptive purposes only and is not intended to be exhaustive or to limit the disclosure to its forms. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of this disclosure and to enable those skilled in the art to understand this disclosure and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A network policy enhancement method, executed by a Policy Control Function (PCF) entity, comprising: The Access and Mobility Management Function (AMF) entity receives terminal capability information reported by user terminals. Receive analysis information related to the service execution efficiency of the user terminal sent by the Network Data Analysis Function (NWDAF) entity; Based on the terminal capabilities, the analysis information related to the service execution efficiency, and the current network resources, a network strategy related to artificial intelligence (AI) and machine learning (ML) is generated. The network strategy includes at least one of the following: an operation partitioning scheme for AI and ML models, adaptive quality of service for transmitting AI and ML models and data, group quality of service, end-to-end quality of service, and a determined distributed learning scheme. The network policy or the policy identifier corresponding to the network policy is sent to the user terminal through the AMF entity so that the user terminal can perform access and mobility management according to the network policy or the network policy corresponding to the policy identifier.

2. The method according to claim 1, wherein, The terminal capability information includes at least one of the following: the computing power of the user terminal, the memory size of the user terminal, the algorithms supported by the user terminal, and the intelligent services supported by the user terminal.

3. The method according to claim 2, wherein, The computing power includes at least one of the following: floating-point operations per second, the clock speed and number of cores of the CPU and GPU of the user terminal, and graph computing capability; The algorithm includes at least one of deep neural network algorithms and distributed learning algorithms; The intelligent services include at least one of image recognition, remote automated control, and autonomous driving.

4. The method according to claim 1, wherein, The operation partitioning scheme includes a hierarchical scheme for AI and ML models; The adaptive quality of service is used to meet the needs of distributed learning schemes; The group service quality is used to limit the total resource consumption for AI and ML operations; The overall quality of service is used to ensure the quality of communication throughout the distributed learning scheme. The determined distributed learning scheme includes at least one of the following: determining the terminals participating in distributed learning and the learning task allocation strategy.

5. The method according to claim 4, wherein, The requirements for the distributed learning solution include the timeliness requirement of the distributed learning solution; The distributed learning scheme includes the federated learning scheme.

6. The method according to any one of claims 1-5, further comprising: The network policy or the policy identifier corresponding to the network policy is sent to the user plane function (UPF) entity through the session management function (SMF) entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

7. A policy control function entity, comprising: The first processing module is configured to receive terminal capability information reported by user terminals through the Access and Mobility Management Function (AMF) entity. The second processing module is configured to receive analysis information related to the service execution efficiency of the user terminal sent by the Network Data Analysis Function (NWDAF) entity. The third processing module is configured to generate network strategies related to artificial intelligence (AI) and machine learning (ML) based on the terminal capabilities, analysis information related to the service execution efficiency, and current network resources. The network strategies include at least one of the following: an operation partitioning scheme for AI and ML models, adaptive quality of service for transmitting AI and ML models and data, group quality of service, end-to-end quality of service, and a determined distributed learning scheme. The fourth processing module is configured to send the network policy or the policy identifier corresponding to the network policy to the user terminal through the AMF entity, so that the user terminal can perform access and mobility management according to the network policy or the network policy corresponding to the policy identifier.

8. The strategy control function entity according to claim 7, wherein, The terminal capability information includes at least one of the following: the computing power of the user terminal, the memory size of the user terminal, the algorithms supported by the user terminal, and the intelligent services supported by the user terminal.

9. The strategy control function entity according to claim 8, wherein, The computing power includes at least one of the following: floating-point operations per second, the clock speed and number of cores of the CPU and GPU of the user terminal, and graph computing capability; The algorithm includes at least one of deep neural network algorithms and distributed learning algorithms; The intelligent services include at least one of image recognition, remote automated control, and autonomous driving.

10. The strategy control function entity according to claim 7, wherein, The operation partitioning scheme includes a hierarchical scheme for AI and ML models; The adaptive quality of service is used to meet the needs of distributed learning schemes; The group service quality is used to limit the total resource consumption for AI and ML operations; The overall quality of service is used to ensure the quality of communication throughout the distributed learning scheme. The determined distributed learning scheme includes at least one of the following: determining the terminals participating in distributed learning and the learning task allocation strategy.

11. The strategy control function entity according to claim 10, wherein, The requirements for the distributed learning solution include the timeliness requirement of the distributed learning solution; The distributed learning scheme includes the federated learning scheme.

12. The strategy control function entity according to any one of claims 7-11, wherein, The fourth processing module is configured to send the network policy or the policy identifier corresponding to the network policy to the user plane function (UPF) entity through the session management function (SMF) entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

13. A policy control function entity, comprising: The memory is configured to store instructions; A processor, coupled to memory, configured to implement the method as described in any one of claims 1-6 based on memory-stored instruction execution.

14. A communication system, comprising: The policy control function PCF entity as described in any one of claims 7-13; The user terminal is configured to send terminal capability information to the PCF entity through the Access and Mobility Management Function (AMF) entity, and receive network policies or policy identifiers related to Artificial Intelligence (AI) and Machine Learning (ML) sent by the PCF entity through the AMF entity, so as to perform access and mobility management according to the network policies or network policies corresponding to the policy identifiers. The Network Data Analysis Function (NWDAF) entity is configured to collect statistical analysis information related to the service execution efficiency of the user terminal and send the statistical results to the PCF entity.

15. The system according to claim 14, wherein, The system also includes: The User Plane Function (UPF) entity is configured to receive the network policy or policy identifier through the Session Management Function (SMF) entity, so that the UPF entity can perform routing management according to the network policy or the network policy corresponding to the policy identifier.

16. A non-transient computer-readable storage medium, wherein, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method as described in any one of claims 1-6.

Images