Policy negotiation method, policy generation method and device

Abstract

Disclosed are a policy negotiation method, policy generation method and device, belonging to the technical field of communications, the policy negotiation method of the embodiment of the present application comprising: a first device receiving a first request message sent by a second device, the first request message being used for requesting to formulate one or more AI service policies for ensuring artificial intelligence (AI) service, or is used for requesting to generate, update or delete the first target AI service strategy; and the first device sends a first response message to the second device, the first response message being used for providing one or more AI service policies for guaranteeing the AI service, or being used for indicating whether to successfully execute generation, update or deletion of a first target AI service policy.

Description

Technical Field

[0001] The present application belongs to the field of communication technology, and specifically relates to a policy negotiation method, a policy generation method and a device. Background Art

[0002] Artificial Intelligence (AI) is a technology that enables intelligent machines to think and judge similarly to humans through large amounts of data and model training. Its research includes robotics, speech recognition, image recognition, natural language, and other fields. With the continuous integration of computing and communications, AI, as one of the future mainstream technologies of computers, is increasingly favored by the communications field.

[0003] Current communication systems are usually designed with sessions as the unit. With the deepening of AI research, future communication systems are expected to introduce an architecture design based on AI services, that is, the communication system provides the required communication resources and computing resources for AI services. Therefore, how the communication network and external network elements negotiate policies for specific AI services and provide stable and reliable communication, computing and other aspects for AI services is a problem that technicians in this field need to solve. Summary of the invention

[0004] The embodiments of the present application provide a policy negotiation method, a policy generation method and a device, which can solve the problem of how a communication network and an external network element perform policy negotiation for a specific AI service.

[0005] In a first aspect, a policy negotiation method is provided, including:

[0006] The first device receives a first request message sent by the second device, where the first request message is used to request the formulation of one or more AI service policies to ensure artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0007] The first device sends a first response message to the second device, where the first response message is used to provide one or more AI service policies that guarantee the AI ​​service, or to indicate whether generation, update, or deletion of the first target AI service policy is successfully executed.

[0008] In a second aspect, a policy negotiation method is provided, including:

[0009] The second device sends a first request message to the first device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0010] The second device receives a first response message sent by the first device; the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0011] In a third aspect, a strategy generation method is provided, comprising:

[0012] The third device receives a second request message sent by the first device, where the second request message is used to request generation of one or more AI service policies to ensure artificial intelligence AI services;

[0013] The third device generates a candidate policy set corresponding to the AI ​​service based on the second request message.

[0014] In a fourth aspect, a policy negotiation method is provided, including:

[0015] The fourth device receives a third request message sent by the first device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy.

[0016] In a fifth aspect, a policy negotiation method is provided, including:

[0017] The fifth device sends a fourth request message to the third device, where the fourth request message is used to instruct to update or delete the second target AI service policy;

[0018] The fourth request message includes at least one of the following:

[0019] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0020] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0021] In a sixth aspect, a policy negotiation device is provided, including:

[0022] A receiving module, configured to receive a first request message sent by a second device, wherein the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0023] A sending module is used to send a first response message to the second device, where the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0024] In a seventh aspect, a policy negotiation device is provided, including:

[0025] A sending module, configured to send a first request message to a first device, wherein the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0026] A receiving module is used to receive a first response message sent by the first device; the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0027] In an eighth aspect, a strategy generation device is provided, including:

[0028] A receiving module, configured to receive a second request message sent by the first device, where the second request message is used to request generation of one or more AI service policies to ensure artificial intelligence AI services;

[0029] A processing module is used to generate a candidate policy set corresponding to the AI ​​service based on the second request message.

[0030] In a ninth aspect, a policy negotiation device is provided, including:

[0031] A receiving module is used to receive a third request message sent by the first device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy.

[0032] In a tenth aspect, a policy negotiation device is provided, including:

[0033] A sending module, configured to send a fourth request message to a third device, where the fourth request message is used to instruct to update or delete the second target AI service policy;

[0034] The fourth request message includes at least one of the following:

[0035] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0036] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0037] In the eleventh aspect, a first device is provided, which includes a processor and a memory, wherein the memory stores a program or instruction that can be executed on the processor, and when the program or instruction is executed by the processor, the steps of the method described in the first aspect are implemented.

[0038] In a twelfth aspect, a first device is provided, comprising a processor and a communication interface, wherein the communication interface is used to receive a first request message sent by a second device, the first request message being used to request the formulation of one or more AI service policies for ensuring artificial intelligence (AI) services, or being used to request the generation, update or deletion of a first target AI service policy; and send a first response message to the second device, the first response message being used to provide one or more AI service policies for ensuring AI services, or being used to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0039] In the thirteenth aspect, a second device is provided, which includes a processor and a memory, wherein the memory stores a program or instruction that can be executed on the processor, and when the program or instruction is executed by the processor, the steps of the method described in the second aspect are implemented.

[0040] In a fourteenth aspect, a second device is provided, comprising a processor and a communication interface, wherein the communication interface is used to send a first request message to a first device, the first request message being used to request the formulation of one or more AI service policies for ensuring artificial intelligence (AI) services, or to request the generation, update or deletion of a first target AI service policy; and to receive a first response message sent by the first device; the first response message being used to provide one or more AI service policies for ensuring AI services, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0041] In the fifteenth aspect, a third device is provided, which includes a processor and a memory, wherein the memory stores programs or instructions that can be executed on the processor, and when the program or instructions are executed by the processor, the steps of the method described in the third aspect are implemented.

[0042] In the sixteenth aspect, a third device is provided, including a processor and a communication interface, wherein the communication interface is used to receive a second request message sent by a first device, and the second request message is used to request the generation of one or more AI service policies to ensure artificial intelligence AI services; the processor is used to generate a candidate policy set corresponding to the AI ​​service based on the second request message.

[0043] In the seventeenth aspect, a fourth device is provided, which includes a processor and a memory, wherein the memory stores a program or instruction that can be executed on the processor, and when the program or instruction is executed by the processor, the steps of the method described in the fourth aspect are implemented.

[0044] In the eighteenth aspect, a fourth device is provided, comprising a processor and a communication interface, wherein the communication interface is used to receive a third request message sent by the first device, and the third request message is used to request the fourth device to generate, update or delete the first target AI service policy.

[0045] In the nineteenth aspect, a fifth device is provided, which includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the program or instructions are executed by the processor, the steps of the method described in the fifth aspect are implemented.

[0046] In the twentieth aspect, a fifth device is provided, comprising a processor and a communication interface, wherein the processor is used to send a fourth request message to the third device, and the fourth request message is used to indicate an update or deletion of a second target AI service policy.

[0047] In the twenty-first aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented, or the steps of the method described in the second aspect are implemented, or the steps of the method described in the third aspect are implemented, or the steps of the method described in the fourth aspect are implemented, or the steps of the method described in the fifth aspect are implemented.

[0048] In the twenty-second aspect, a wireless communication system is provided, including: a first device, a second device, a third device, a fourth device and a fifth device, wherein the first device can be used to execute the steps of the method described in the first aspect, the second device can be used to execute the steps of the method described in the second aspect, the third device can be used to execute the steps of the method described in the third aspect, the fourth device can be used to execute the steps of the method described in the fourth aspect, and the fifth device can be used to execute the steps of the method described in the fifth aspect.

[0049] In the twenty-third aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instructions to implement the method as described in the first aspect, or the method as described in the second aspect, or the method as described in the third aspect, or the method as described in the fourth aspect, or the method as described in the fifth aspect.

[0050] In the twenty-fourth aspect, a computer program / program product is provided, which is stored in a storage medium, and the program / program product is executed by at least one processor to implement the steps of the policy negotiation method as described in the first aspect or the second aspect, or the steps of the policy generation method as described in the third aspect, or the steps of the policy negotiation method as described in the fourth aspect or the fifth aspect.

[0051] In an embodiment of the present application, a first device receives a first request message sent by a second device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy; the first device sends a first response message to the second device, where the first response message is used to provide one or more AI service policies for guaranteeing AI services, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed. In the above scheme, AI service policies can be determined for different AI services, and the AI ​​service policy can provide communication guarantee and AI resource guarantee for an AI service, and can generate, update or delete the first target AI service policy, thereby realizing the negotiation of AI service policies between the first device and the second device. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a schematic diagram of the architecture of a wireless communication system provided in an embodiment of the present application;

[0053] Figure 2 This is one of the flow charts of the policy negotiation method provided in the embodiment of the present application;

[0054] Figure 3 This is one of the interactive schematic diagrams of the policy negotiation method provided in the embodiment of the present application;

[0055] Figure 4 This is the second interactive schematic diagram of the policy negotiation method provided in the embodiment of the present application;

[0056] Figure 5 This is the third interactive schematic diagram of the policy negotiation method provided in the embodiment of the present application;

[0057] Figure 6 This is the second flow chart of the policy negotiation method provided in the embodiment of the present application;

[0058] Figure 7 This is the second flow chart of the strategy generation method provided in the embodiment of the present application;

[0059] Figure 8 This is the second flow chart of the policy negotiation method provided in the embodiment of the present application;

[0060] Fig. 9 This is the second flow chart of the policy negotiation method provided in the embodiment of the present application;

[0061] Fig.10 This is one of the structural diagrams of the policy negotiation device provided in the embodiment of the present application;

[0062] Fig.11 This is the second structural diagram of the policy negotiation device provided in the embodiment of the present application;

[0063] Fig.12 This is the third structural diagram of the policy negotiation device provided in the embodiment of the present application;

[0064] Fig.13 This is the fourth structural diagram of the policy negotiation device provided in the embodiment of the present application;

[0065] Fig.14 This is the fifth structural diagram of the policy negotiation device provided in the embodiment of the present application;

[0066] Fig.15 is a schematic diagram of the structure of a communication device provided in an embodiment of the present application;

[0067] Fig.16 is a schematic diagram of the structure of a terminal in an embodiment of the present application;

[0068] Fig.17 It is a schematic diagram of the structure of the network side device of an embodiment of the present application. DETAILED DESCRIPTION

[0069] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of this application.

[0070] The terms "first", "second", etc. of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of one type, and the number of objects is not limited, for example, the first object can be one or more. In addition, "or" in the present application represents at least one of the connected objects. For example, "A or B" covers three schemes, namely, Scheme 1: including A but not including B; Scheme 2: including B but not including A; Scheme 3: including both A and B. The character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0071] The term "indication" in this application can be a direct indication (or explicit indication) or an indirect indication (or implicit indication). A direct indication can be understood as the sender explicitly informing the receiver of specific information, operations to be performed, or request results in the sent indication; an indirect indication can be understood as the receiver determining the corresponding information according to the indication sent by the sender, or making a judgment and determining the operation to be performed or the request result according to the judgment result.

[0072] It is worth noting that the technology described in the embodiments of the present application is not limited to the Long Term Evolution (LTE) / LTE-Advanced (LTE-A) system, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA) or other systems. The terms "system" and "network" in the embodiments of the present application are often used interchangeably, and the described technology can be used for the systems and radio technologies mentioned above as well as for other systems and radio technologies. The following description describes a New Radio (NR) system for illustrative purposes, and NR terminology is used in most of the following descriptions, but these technologies can also be applied to systems other than NR systems, such as the 6th generation (6 th Generation, 6G) communication system.

[0073] Figure 1A block diagram of a wireless communication system applicable to an embodiment of the present application is shown. The wireless communication system includes a terminal 11 and a network side device 12. Among them, the terminal 11 can be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer), a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (Ultra-mobile Personal Computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), an augmented reality (Augmented Reality, AR), a virtual reality (Virtual Reality, VR) device, a robot, a wearable device (Wearable Device), an aircraft (flight vehicle), a vehicle-mounted device (Vehicle User Equipment, VUE), a ship-mounted device, a pedestrian terminal (Pedestrian User Equipment, PUE), a smart home (home appliances with wireless communication functions, such as refrigerators, televisions, washing machines or furniture, etc.), a game console, a personal computer (Personal Computer, PC), a teller machine or a self-service machine and other terminal side devices. Wearable devices include: smart watches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. Among them, the vehicle-mounted device can also be called a vehicle-mounted terminal, a vehicle-mounted controller, a vehicle-mounted module, a vehicle-mounted component, a vehicle-mounted chip or a vehicle-mounted unit, etc. It should be noted that the specific type of the terminal 11 is not limited in the embodiment of the present application. The network side device 12 may include an access network device or a core network device, wherein the access network device may also be referred to as a radio access network (Radio Access Network, RAN) device, a radio access network function or a radio access network unit. The access network device may include a base station, a wireless local area network (Wireless Local Area Network, WLAN) access point (Access Point, AS) or a wireless fidelity (Wireless Fidelity, WiFi) node, etc.Among them, the base station may be referred to as a Node B (NB), an evolved Node B (eNB), a next generation Node B (gNB), a New Radio Node B (NR Node B), an access point, a Relay Base Station (RBS), a Serving Base Station (SBS), a Base Transceiver Station (BTS), a radio base station, a radio transceiver, a Basic Service Set (BSS), an Extended Service Set (ESS), a Home Node B (HNB), a Home Evolved Node B, a Transmission Reception Point (TRP) or other appropriate terms in the field. As long as the same technical effect is achieved, the base station is not limited to specific technical terms. It should be noted that in the embodiments of the present application, only the base station in the NR system is used as an example for introduction, and the specific type of the base station is not limited.

[0074] The core network equipment may include but is not limited to at least one of the following: core network node, core network function, mobility management entity (Mobility Management Entity, MME), access mobility management function (Access and Mobility Management Function, AMF), session management function (Session Management Function, SMF), user plane function (User Plane Function, UPF), policy control function (Policy Control Function, PCF), policy and charging rules function unit (Policy and Charging Rules Function, PCRF), edge application service discovery function (Edge Application Server Discovery Function, EASDF), unified data management (Unified Data Management, UDM), unified data storage (Unified Data Repository, UDR), home user server (Home Subscriber Server, HSS), centralized network configuration (CNC), network storage function (Network Repository Function, NRF), network exposure function (Network Exposure Function, NEF), local NEF (Local NEF, or L-NEF), binding support function (Binding Support Function, BSF), application function (Application Function, AF), etc. It should be noted that in the embodiments of the present application, only the core network device in the NR system is introduced as an example, and the specific type of the core network device is not limited.

[0075] First, the nouns and application scenarios involved in the embodiments of the present application are introduced:

[0076] NEF can provide an interactive interface for 3GPP internal network elements and external entities, and can provide external third parties with public services of capabilities and events, including capabilities, information and analysis; at the same time, it can also obtain data and information from external third parties; in this process, NEF can also provide parameter completion and information translation functions. As an important interface for internal and external interactions, it can also verify the identity of third parties to prevent information leakage and unsafe third parties from requesting data from the 3GPP network.

[0077] The existing 3GPP standard specifies the use of the Quality of Service (QoS) mechanism to ensure the quality of service of the terminal at the communication level. From the perspective of the transport layer or physical layer, no matter what data is transmitted, it will be converted into a corresponding bit stream. The bottom layer only needs to guarantee the corresponding communication level metrics such as bandwidth, bit error rate and latency. The 5G QoS feature should be understood as a guide to setting node-specific parameters for each QoS flow, such as for 3GPP wireless access link layer protocol configuration.

[0078] With the deepening of AI research, 6G or future communication systems are expected to introduce an architecture design based on AI services, which means that the 3GPP system can provide the required communication and computing resources for AI services.

[0079] If there is no negotiation of AI service policies, when the AF subsequently initiates AI services, the 3GPP network is unclear about how to provide guarantees for AI services. Furthermore, if there is no negotiation process, the AF directly requests AI services and requires guarantees, but the 3GPP network does not support it, which will cause 3GPP to reject AI services. In addition, different AI services require different guarantees. How can 3GPP support a large number of different AI services? Therefore, a policy negotiation solution is provided in the embodiments of this application.

[0080] The following describes in detail the policy negotiation method provided in the embodiment of the present application through some embodiments and their application scenarios in combination with the accompanying drawings.

[0081] Please refer to Figure 2 The embodiment of the present application provides a policy negotiation method. The execution subject of the embodiment is a first device. The policy negotiation method includes:

[0082] Step 101: The first device receives a first request message sent by the second device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0083] Specifically, the first device may be a NEF, and the second device may be an external network element, a non-3GPP network element, or a third-party entity (eg, an AF).

[0084] The AF sends a first request message to the NEF, such as an AI service policy negotiation request, for requesting the formulation of one or more AI service policies to guarantee the AI ​​service, that is, formulating an AI service policy that meets the business requirements for the AI ​​service. Optionally, formulating an AI service policy to guarantee the AI ​​service may be to create a new AI service policy for the AI ​​service, or to associate the AI ​​service with the AI ​​service policy.

[0085] Optionally, the first request message can be implemented by adding a new interaction message (for example, Nnef_AIServiceNegotiation service), indicating that AF requests 3GPP to negotiate and formulate an AI service policy.

[0086] Or, the first request message is used to request the generation, update or deletion of the first target AI service policy. For example, the AF needs to adjust the AI ​​service policy due to changes in the AI ​​service, and requests the 3GPP network to generate, update or modify the first target AI service policy for the target AI service.

[0087] Step 102: The first device sends a first response message to the second device, where the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update, or deletion of the first target AI service policy is successfully executed.

[0088] Specifically, in a case where the first request message is used to request formulation of one or more AI service policies for guaranteeing the AI ​​service, the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service;

[0089] In the case where the first request message is used to request generation, update or deletion of the first target AI service policy, the first response message is used to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0090] The method of this embodiment can determine AI service policies for different AI services. The AI ​​service policy can provide communication guarantee and AI resource guarantee for an AI service, and can generate, update or delete the first target AI service policy, thereby realizing the negotiation of AI service policies between the first device and the second device.

[0091] Optionally, when the first request message is used to request formulation of one or more AI service policies for guaranteeing AI services, the first request message includes at least one of the following:

[0092] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, target policy information, and first message type;

[0093] When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following:

[0094] ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information.

[0095] Optionally, the second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

[0096] Specifically, the first message type is used to represent that the second device (AF) requests to negotiate with 3GPP to determine an AI service policy for guaranteeing the AI ​​service.

[0097] The second message type is used to indicate that the second device (AF) requests the 3GPP network to generate, update or delete the first target AI service policy for the target AI service.

[0098] Application Service Provider (ASP) identification, used to identify the application server, including identification ID, IP address, etc.

[0099] The first request message may include one or more AI service IDs. For example, if multiple AI services are targeted, multiple AI service IDs may be carried. The AI ​​service ID is mainly used to indicate what kind of AI service this is, such as picture beautification, image recognition, speech recognition, etc.

[0100] The difference between AI service type and AI service identifier is that AI service type is mainly used to indicate whether it is a training service or an inference service. AI service type includes at least one of the following: inference service, training service, inference service and training service. Among them, the QoS requirement of AI service includes at least one of the following:

[0101] QoS Reference ID of the AI ​​service, such as the value of the standardized 5G QoS Identifier (5QI), the value of the AI ​​QoS index, etc.

[0102] Individual QoS parameters for AI services. The difference from the QoS association ID for AI services is that it carries more detailed parameter information, such as at least one of the following: priority, latency, packet loss rate, packet error rate, and peak rate.

[0103] The first request message may include one or more slice identifiers / data network names (Data Network Name, DNN). For example, if there are multiple AI services, multiple different slice identifiers or DNNs may be carried. The slice identifier and DNN are resource identifiers associated with the AI ​​service, that is, the second device can request to use the resource corresponding to the resource identifier to initiate the AI ​​service, and then the network side can formulate an AI service policy based on the resources used by the AI ​​service.

[0104] The expected policy application time refers to the application time of the AI ​​service policy expected by the second device. The application time can be a continuous time or a time window with multiple intervals.

[0105] The expected policy applicable area indicates the geographical area or network area to which the AI ​​service policy expected by the second device is applicable.

[0106] The first indication information is used to indicate whether the first device needs to notify the second device, such as AF, when the AI ​​service policy is upgraded or changed.

[0107] The parameter information defined by the second device, also called AF specific parameters, includes, for example, accuracy information defined by AF, QoS parameters defined by AF, etc. Here, the accuracy information is the accuracy understood by AF. After being transmitted to the first device (such as NEF), it needs to be translated on the first device to translate this information into the accuracy defined by 3GPP (for example, the training accuracy or inference accuracy of the model). The QoS parameters defined by AF include, for example, packet loss rate, packet error rate, maximum latency, etc. The QoS parameters defined by AF may be expressed differently from the QoS parameters understood by the first device and other network elements.

[0108] Optionally, the data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and total number of data sources; wherein the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: reasoning business, training business, and training and reasoning business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0109] The off-network data source refers to, for example, an AF data source or a World Wide Web data source, and the off-network data source identifier is, for example, an AF identifier or a network address.

[0110] The model training requirements include at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0111] Among them, the model training requirements are training requirements for the model associated with the AI ​​service, and the corresponding AI service strategy can be determined based on the model training requirements; for example, the second device can use the model associated with the AI ​​service to implement the AI ​​service.

[0112] For example, if the second device carries a model, it means that the second device wants to collect data for the model it carries to perform training and reasoning, so the receiver (such as the first device) needs to save the model to perform training and reasoning. If the second device does not carry a model, the receiver (such as the first device) needs to train a model according to the corresponding requirements.

[0113] Model size can be explicitly expressed, enumerated: 10M, 50M, 100M, etc., or implicitly expressed, such as small, medium, large, super large;

[0114] Model reasoning requirements include at least one of the following: reasoning latency, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method;

[0115] Among them, model reasoning requirements refer to the reasoning requirements of the model associated with the AI ​​service, based on which the corresponding AI service strategy can be determined.

[0116] The generalization range can be understood as the reasoning performance of the model facing unfamiliar data sets. This parameter includes the effective time (or applicability time) of the model.

[0117] The effective area is the applicable area of ​​the model, such as a network area or a geographical area;

[0118] The effective target refers to which terminals, network elements or AFs the model is applicable to;

[0119] Inference accuracy can be understood as the inference accuracy of the model;

[0120] The inference accuracy evaluation method includes a method for reflecting the quality of the inference result, a method for comparing accuracy, etc. The inference accuracy evaluation method can be represented by, for example, a matrix (inference matrix), and an element of the matrix represents an inference accuracy evaluation method.

[0121] The target policy information includes at least one of the following: a policy reference identifier and a QoS association ID of the AI ​​service.

[0122] When the first request message is used to request the generation, update or deletion of the first target AI service policy, the first request message may include target policy information, such as a policy reference identifier, that is, generating, updating or deleting the AI ​​service policy corresponding to the policy reference identifier, and may also include the QoS association ID of the AI ​​service, that is, the AI ​​service policy corresponding to the QoS association ID can be generated, updated or deleted.

[0123] Optionally, the method further comprises:

[0124] When it is determined that the second device meets the preset condition and the first request message is used to request the formulation of one or more AI service policies to ensure artificial intelligence AI services, the first device sends a second request message to the third device, where the second request message is used to request the third device to generate the AI ​​service policy to ensure the AI ​​service;

[0125] In a case where it is determined that the second device does not meet the preset condition, the first response message is used to indicate a rejection of the first request message of the second device, and the first response message includes a first rejection reason;

[0126] The first rejection reason includes at least one of the following:

[0127] The second device is not allowed to request an AI service policy, the second device is not allowed to initiate an AI service, the number of AI services associated with the first device reaches the upper limit, and the number of AI services associated with the second device reaches the upper limit;

[0128] The preset condition includes at least one of the following:

[0129] The second device is allowed to request an AI service policy, the second device is allowed to initiate an AI service, and the number of AI services associated with the second device does not exceed an upper limit.

[0130] Specifically, the first device checks the first request message of the second device to determine whether the second device is allowed to request an AI service policy, whether the second device is allowed to initiate an AI service, whether the number of AI services associated with the second device exceeds an upper limit, etc. If the check fails, the request of step 101 is directly rejected, and optionally, the first rejection reason may be carried.

[0131] Optionally, the third device may be a PCF.

[0132] If the first device sends a second request message to the third device through inspection, the second request message is used to request the third device to generate an AI service policy for guaranteeing the AI ​​service. Optionally, the second request message may include at least one parameter in the first request message.

[0133] In one implementation, the second request message can be implemented through a newly added interaction message Npcf_AIServicePolicyservice, indicating a request for a third device to formulate an AI service policy for the AI ​​service.

[0134] In the above implementation, the first device checks the first request message of the second device to determine whether the preset conditions are met, and then requests to generate an AI service policy, thereby ensuring the security and stability of the network.

[0135] Optionally, before step 102, the method further includes:

[0136] The first device receives a second response message sent by the third device, where the second response message is used to indicate acceptance or rejection of the second request message of the first device.

[0137] Optionally, after receiving the second request message, i.e., receiving the request to generate a policy, the third device sends a sixth request message, i.e., a data query message, to the fourth device (such as a UDR), requesting to obtain AI service policies for one or more AI services.

[0138] After the fourth device queries the database, if the requested AI service policy is saved, it will be sent to the third device in a reply message; if it is not saved, it will indicate in the reply message that the queried AI service policy information does not exist, and the third device will generate an AI service policy after receiving the reply message.

[0139] In one implementation method, the message between the third device and the fourth device can reuse the existing Nudr_DM_Queryservice message and enhance it on this basis, such as adding a new data type of AI service-related policy information to the existing UDR service, and the key is the user permanent identifier (Subscription Permanent Identifier, SUPI) or the AI ​​service identifier.

[0140] It should be noted that there is currently a mechanism for communication between PCF and UDR, but there is no relevant provision for requesting AI service policy information.

[0141] Optionally, the third device generates a candidate policy set corresponding to the AI ​​service based on the second request message and the local policy of the third device, wherein the candidate policy set corresponds to the AI ​​service identifier one by one;

[0142] Optionally, the third device may also generate a candidate policy set corresponding to the AI ​​service based on the second request message, the local policy of the third device, and third information, where the third information includes at least one of the following:

[0143] Whether the parameters of the first request message are satisfied;

[0144] An AI service policy associated with the AI ​​service stored in the fourth device.

[0145] Specifically, whether the parameters of the first request message are met, for example, whether at least one of the following AI services is supported; whether the required parameters are supported; whether the required parameters support, for example, the model accuracy in the model training requirements, the inference delay in the model inference requirements, the QoS parameters in the QoS requirements of the AI ​​service, etc.

[0146] In one implementation, the second response message may be implemented through a newly added interaction message Npcf_AIServicePolicyservice, indicating that the PCF replies with the AI ​​service policy formulated for the AI ​​service.

[0147] If the requirement of the parameter of the AI ​​service requested in step 101 cannot be met, the third device sends a second response message indicating rejection of the second request message.

[0148] Optionally, when the second response message is used to indicate rejection of the second request message of the first device, the second response message includes a second rejection reason; and the second rejection reason includes at least one of the following:

[0149] The model training requirements of the second device are not supported, the model inference requirements of the second device are not supported, the QoS requirements of the AI ​​service of the second device are not supported, and the AI ​​service of the second device is not supported.

[0150] Optionally, when the second response message indicates acceptance of the second request message of the first device, the second response message includes at least one of the following:

[0151] ASP logo;

[0152] Slice identification;

[0153] DNN;

[0154] One or more AI service identifiers;

[0155] A candidate strategy set corresponding to the AI ​​service identifier one by one;

[0156] The candidate policy set includes information of at least one candidate policy, and the information of the candidate policy includes at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of an AI service; a policy applicable time; and a policy applicable area.

[0157] The ASP identifier may be used to address the first device, and the ASP identifier may be included in a response message specifically used to reply to the policy negotiation request of the ASP.

[0158] The slice identifier and DNN are resource identifiers associated with the AI ​​service, and are used by the subsequent second device when initiating the AI ​​service.

[0159] Optionally, when the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, the first response message includes at least one of the following:

[0160] ASP logo;

[0161] Slice identification;

[0162] DNN;

[0163] One or more AI service identifiers;

[0164] A candidate strategy set corresponding to the AI ​​service identifier one by one;

[0165] The candidate policy set includes information of at least one candidate policy, and the information of the candidate policy includes at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of an AI service; a policy applicable time; and a policy applicable area.

[0166] The ASP identifier may be used to address the first device, and the ASP identifier may be included in a response message specifically used to reply to the policy negotiation request of the ASP.

[0167] The slice identifier and DNN are resource identifiers associated with the AI ​​service, and are used by the subsequent second device when initiating the AI ​​service.

[0168] It should be noted that the model training accuracy achieved is supplemented on the basis of the model accuracy requested by the first request message, and is an accuracy that can be identified and provided by the operator. One form of expression is to provide a set of specific accuracy information and corresponding indexes, from which the subsequent AF can select one or part of the indexes;

[0169] Optionally, if the first request message does not contain a QoS association ID and carries an individual QoS parameter, the QoS association ID of the AI ​​service is carried to the second device. The value of the QoS association ID of the AI ​​service can be a standardized AIQoS index value or 5QI, or it can be an operator-specific AIQoS index value or 5QI. The QoS association ID of the AI ​​service can be one or more AIQoS index values ​​or 5QIs.

[0170] It should be noted that the QoS association ID of the above-mentioned AI service may not be recognized by the second device. Therefore, the first device needs to support the conversion of the QoS association ID (such as AIQoS index) into parameters that the second device can understand, such as the following format: {AIQoS index; first parameter}. In this way, for the second device, when subsequently carrying the QoS parameters of the AI ​​service, it can directly indicate which AIQoS index is supported and which is not supported. From the AF's point of view, the corresponding parameters can be known from the AIQoS index.

[0171] Optionally, the QoS association ID of the AI ​​service in the first response message is represented by first information, where the first information includes:

[0172] QoS association ID and first parameter;

[0173] The first parameter includes at least one of the following: a parameter value characterizing the time dimension of the service experience; a parameter value characterizing the quality dimension of the service experience; a parameter value characterizing the data density dimension of the service experience; a parameter value characterizing the priority dimension of the service experience; a parameter value characterizing the resource assurance dimension of the service experience; a parameter value characterizing the stability dimension of the service experience; and a parameter value characterizing the resource overhead dimension of the service experience.

[0174] Optionally, the parameter value of the time dimension characterizing the service experience includes: service response delay, the service response delay is obtained based on at least one of the following information, and the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay;

[0175] The parameter value characterizing the quality dimension of the service experience includes: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value;

[0176] The parameter value of the data density dimension characterizing the service experience includes: the amount of data transmitted and processed, and the amount of data transmitted and processed includes at least one of the following: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time;

[0177] The parameter values ​​of the priority dimension characterizing the service experience include: priority level;

[0178] The parameter value of the resource assurance dimension characterizing the service experience includes: resource type, the resource type includes at least one of the following: a guaranteed resource type; a non-guaranteed resource type, a delay-sensitive guaranteed resource type;

[0179] The parameter values ​​representing the stability dimension of the service experience include: time window information;

[0180] The parameter value of the resource overhead characterizing the service experience includes at least one of the following:

[0181] The maximum data burst transfer amount, the maximum computing amount for each computing resource type, and the maximum storage amount.

[0182] Specifically, the parameter value of the time dimension characterizing the service experience includes the service response delay, wherein the service response delay is obtained based on at least one of the following information, wherein the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay;

[0183] One form of expression is: communication delay refers to the (communication) transmission delay of terminal data, and pure data collection and transmission delay refers to the collection and transmission delay of other data sources;

[0184] For example, service response delay = f(x1, x2, x3), where x1 is the terminal data communication delay, x2 is the AI ​​processing delay, and x3 is the pure data collection and transmission delay (x3 is, for example, the inference data collected in the network, the transmission delay from the data provider to the data consumer); f() is a mathematical function, and the specific function selection is determined by the operator's strategy and configuration. For example, F() = x1 + x2 + x3, F() = x1 + x2, etc.

[0185] The parameter values ​​representing the quality dimension of the service experience include: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value;

[0186] One way to express the communication reliability value is the bit error rate.

[0187] The data in the data integrity includes but is not limited to the data collected at the data surface.

[0188] Optionally, AI processing accuracy such as inference accuracy, or expected training accuracy;

[0189] For example, service accuracy = F(y1, y2, y3), where y1 is the communication reliability (such as bit error rate), y2 is the AI ​​processing accuracy (such as inference accuracy, or expected training accuracy), and y3 is the integrity / accuracy / timeliness of the data collected on the data plane; f() is a mathematical function, and the specific function selection is determined by the operator's strategy and configuration. For example, F() = y1*y2*y3, F() = y1*y2, etc.

[0190] Among them, the parameter value of the data density dimension that characterizes the service experience: the amount of data transmitted and processed, which may include one or more of the following parameters: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time;

[0191] Among them, the parameter value of the resource assurance dimension that characterizes the service experience: resource type, which can be one of the following: guaranteed Guarantee resource type; non-Guarantee resource type; delay-critical Guaranteed resource type;

[0192] Among them, the parameter values ​​that characterize the stability / sustainability dimension of the service experience are: time window averaging window (only for Guarantee type resources and Delay-critical Guaranteed resource types);

[0193] The time window refers to the time window in which the service experience is stable, and the service experience may not be guaranteed outside this time window.

[0194] The parameter value of the resource overhead that characterizes the service experience includes one or more of the following parameters: Maximum Data Burst Volume; Maximum computing volume per computing resource type; Maximum storage volume.

[0195] In the above implementation, the first device converts the QoS requirement, such as the QoS association ID, into information that can be recognized by the second device, and sends the information to the second device, thereby improving the processing efficiency of the second device.

[0196] Optionally, the method further comprises:

[0197] When it is determined that the second device meets the preset condition and the first request message is used to request generation, update or deletion of the first target AI service policy, the first device sends a third request message to the fourth device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy;

[0198] The first device receives a third response message sent by the fourth device;

[0199] The first device sends the third response message to the second device;

[0200] Optionally, the third request message includes at least one parameter in the first request message.

[0201] In the above implementation, specific AI service policies can also be generated, updated or deleted through the fourth device, which has greater flexibility.

[0202] For example, Figure 3 As shown, the method includes:

[0203] Step 1: AF sends an AI service policy generation, update, or deletion request message to NEF;

[0204] Specifically, the request message in step 1 includes, for example, an ASPID, an AI service identifier, and a selected policy reference identifier;

[0205] Step 2: NEF sends a policy creation, update, or deletion request message to UDR;

[0206] Specifically, the request message in step 2 may include at least one parameter of the request message in step 1;

[0207] Step 3: UDR sends a policy creation, update, or deletion reply message to NEF;

[0208] For example, the reply message in step 3 may include ASPID and ACK;

[0209] Step 4: NEF sends an AI service policy generation, update, or deletion reply message to AF;

[0210] For example, the reply message in step 4 may include ACK;

[0211] Step 5: UDR sends a policy generation, update, or deletion notification message to PCF;

[0212] Specifically, the notification message in step 5 includes, for example, AI service policy information associated with the AI ​​service.

[0213] In step 5, for example, the UDR sends a policy generation, update or deletion message to the contracted PCF, carrying policy information related to the AI ​​service. One implementation method is to add a new data type of policy information related to the AI ​​service to the service of the existing UDR, and the key is the user permanent identifier or the AI ​​service identifier.

[0214] For example, assuming that an update message is sent, in the previous negotiation process, 3GPP provides three different AI service policies with policy reference identifiers 1, 2, and 3, as well as policy reference identifiers and AI service QoS requirements corresponding to different AI service policies; AF finally selects the AI ​​service policy with policy reference identifiers 1 and 2. In this embodiment, AF expects to use AI service policies 2 and 3, so policy reference identifiers 2 and 3 are carried in the first request message, and the corresponding AI service QoS requirements also need to be replaced with the QoS requirements of policy reference identifiers 2 and 3.

[0215] Optionally, the method further comprises:

[0216] The first device receives a first notification message from the second device, where the first notification message is used to indicate information about the AI ​​service policy selected by the second device.

[0217] Optionally, the first notification message may also be called a service policy selection notification message.

[0218] Optionally, when the candidate policy set includes an AI service policy, the first notification message may be used to indicate whether the negotiation is successful or fails.

[0219] Specifically, when the second device receives one or more candidate policy sets corresponding to the AI ​​service, it evaluates the candidate policy set for each AI service according to the local policy and scoring mechanism.

[0220] Specifically, the first notification message, which may also be referred to as an AI service policy selection notification message, is used to notify the second device of information about the AI ​​service policy selected for the AI ​​service, which may be understood as the AI ​​service policy finally determined through negotiation.

[0221] In one implementation method, the first notification message can reuse the existing Nnef_ApplyPolicy service message, or use the newly added Nnef_ApplyAIServicePolicy service message to implement the interactive process of AI service policy negotiation.

[0222] If the second device receives only one policy in step 102, it replies with a confirmation message in the above step, including two situations: negotiation is successful or failed. Negotiation is successful, indicating that the only policy received in step 102 is used in the future, and negotiation failure indicates that the second device refuses to use the policy provided this time.

[0223] Optionally, the first notification message occurs after the first response message;

[0224] Optionally, the first notification message includes at least one of the following:

[0225] ASP logo;

[0226] Slice identification;

[0227] DNN;

[0228] AI service logo;

[0229] A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policy includes:

[0230] Policy reference identifier;

[0231] QoS association ID of the AI ​​service.

[0232] Optionally, the first device sends the first notification message to the third device, where the first notification message is used to notify the third device of information about the AI ​​service policy selected by the second device.

[0233] Specifically, the first device sends the first notification message to the third device to notify the third device of the AI ​​service strategy of the AI ​​service selected by the second device. The first notification message can be sent directly to the third device, or can be repackaged and sent. The embodiments of the present application are not limited to this.

[0234] Optionally, the third device sends a seventh request message to the fourth device, where the seventh request message is used to instruct the fourth device to store information about the AI ​​service policy selected by the second device;

[0235] Optionally, the seventh request message includes at least one parameter in the first notification message.

[0236] The third device stores the AI ​​service policy information finally determined for the AI ​​service in the UDR.

[0237] In one implementation method, the seventh request message may reuse the existing Nudr_DM_Create request message.

[0238] In the above implementation, the AI ​​service policy information determined for the AI ​​service is stored in the UDR, which facilitates subsequent use.

[0239] Optionally, the method further comprises:

[0240] The third device sends a fourth request message to the sixth device, where the fourth request message is used to request to obtain model-related information.

[0241] Specifically, the sixth device, such as the Network Data Analytics Function (NWDAF), can be used for AI model management; the fourth request message, which can be called a model request message, obtains information related to the model, including analysis results, training accuracy, reasoning accuracy, data set information, etc. This step may be performed multiple times, or it may take a long time to get the result. The information obtained can be used to provide a reference for PCF to formulate AI service strategies.

[0242] Optionally, implementation methods of the fourth request message include but are not limited to a model signing message and a model analysis request message.

[0243] If a model identifier (such as model ID, analytics ID) is provided in step 101, the fourth request message is used for the third device to sign a contract for the model with the sixth device, and obtain the subscription association ID, model analysis accuracy, model training information, and model reasoning information from the sixth device.

[0244] If the model identification information is not provided in step 101, the third device, after determining the strategy and the model, requests the sixth device to allocate a model identification, and sends the model training requirements and the model inference requirements to the sixth device.

[0245] In one implementation, the fourth request message may reuse the existing message Nnwdaf_AnalyticsSubscription_Subscribe message.

[0246] For example, Figure 4 As shown, the method includes:

[0247] Step 1: AF sends an AI service negotiation request to NEF to request the formulation of a corresponding AI service policy for AI service negotiation;

[0248] Specifically, the AI ​​service negotiation request may include AI service-related information, such as at least one of the following: ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, AI service quality QoS requirements, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, and target policy information.

[0249] Step 2: NEF sends a request to generate a policy to PCF;

[0250] The generating policy request may include at least one parameter of the AI ​​service negotiation request.

[0251] Step 3: The PCF sends a query request to the UDR. The query request is implemented, for example, through a Nudr_DM_Query request message.

[0252] Optionally, the UDR sends a response message corresponding to the query request to the PCF.

[0253] Step 4: PCF sends a subscription request to NWDAF to obtain model-related information, for example, through the Nnwdaf_AnalyticsSubscription_Subscribe message.

[0254] Optionally, the NWDAF sends a response message corresponding to the subscription request to the PCF.

[0255] Step 5: PCF generates AI service strategy;

[0256] Step 6: PCF sends a generated policy reply to NEF;

[0257] The generated policy response includes: ASPID and candidate policy set.

[0258] Step 7: NEF sends an AI service negotiation reply to AF;

[0259] Among them, the AI ​​service negotiation reply may include a candidate strategy set.

[0260] Step 8: AF sends AI service policy selection notification to NEF;

[0261] The AI ​​service policy selection notification may include information of the selected AI service policy, such as a policy reference identifier and a QoS association ID of the AI ​​service.

[0262] Step 9: NEF sends a policy selection notification to PCF;

[0263] The policy selection notification includes, for example, an ASPID, an AI service identifier, and information about the selected AI service policy.

[0264] Step 10: The PCF sends a storage request message to the UDR to request that the AI ​​service policy information finally determined for the AI ​​service be stored in the UDR. The request message may include the content in step 8.

[0265] In one implementation method, the storage request message may reuse an existing Nudr_DM_Create request message.

[0266] Optionally, the UDR sends a response message corresponding to the request message to the PCF.

[0267] Optionally, when the AI ​​service that can be guaranteed and provided by the terminal or network function (NF) side (for example, NWDAF, AMF) changes, the AI ​​service policy needs to be adjusted; if the adjusted AI service policy affects the AI ​​service policy previously negotiated and determined by the AF or related network elements, the process is triggered to notify the AF, or to notify the network elements that have signed or are using the policy.

[0268] Optionally, the method further comprises:

[0269] The first device receives a second notification message from the third device, where the second notification message is used to notify that the second target AI service policy has been updated or deleted;

[0270] The first device sends a first notification response message to the third device.

[0271] Optionally, the second notification message may be sent after the third device receives the following message, that is, a fourth request message sent by the fifth device, where the fourth request message is used to instruct to update or delete the second target AI service policy;

[0272] The fourth request message includes at least one of the following:

[0273] Message type, slice identifier, DNN, one or more AI service identifiers, updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner;

[0274] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0275] Specifically, the fifth device may be a terminal or an NF. The terminal or NF may obtain policy information from the PCF when requesting an AI service policy, such as a policy reference identifier and QoS information of the AI ​​service, or the network may inform the terminal or NF when establishing the AI ​​service, so that the above parameters may be carried when modifying or deleting.

[0276] In the above implementation, the update or deletion of a specific AI service policy can also be initiated through a terminal or a network-side device.

[0277] Optionally, after the first device receives the second notification message from the third device, the method further includes:

[0278] The first device sends the second notification message to the second device;

[0279] The first device receives a second notification response message sent by the second device.

[0280] Optionally, the second notification message includes at least one of the following:

[0281] A third message type, wherein the third message type includes one of the following: updating the second target AI service policy, deleting the second target service policy;

[0282] Slice identification;

[0283] DNN;

[0284] One or more AI service identifiers;

[0285] A policy set, wherein the policy set corresponds to the AI ​​service identifier in a one-to-one manner, each of the policy sets includes at least one AI service policy, and the AI ​​service policy includes: a policy reference identifier; and a QoS association ID of the AI ​​service.

[0286] For example, Figure 5 As shown, the method includes:

[0287] Step 1: The terminal or network function NF sends an AI service policy update or deletion request to the PCF;

[0288] The AI ​​service policy update or deletion request may include the AI ​​service identifier and relevant information of the AI ​​service policy;

[0289] Step 2: PCF and UDR interact to update or delete policies;

[0290] Step 3: PCF sends a policy update or deletion notification to NEF;

[0291] For example, the update or deletion notification includes: ASPID, AI service identifier, and related information of AI service policy;

[0292] Step 4: NEF sends a policy notification to AF.

[0293] For example, the policy notification includes: ASPID, AI service identifier, and relevant information of the AI ​​service policy;

[0294] Step 5: AF sends a policy notification reply to NEF;

[0295] For example, the notification reply includes: ASPID and ACK;

[0296] Step 6: NEF sends a policy update or deletion reply to PCF;

[0297] For example, the reply message includes: ASPID and ACK;

[0298] Step 7: PCF sends an AI service policy update or deletion reply to the terminal or network function NF;

[0299] For example, the reply message is ACK.

[0300] Please refer to Figure 6 The embodiment of the present application provides a policy negotiation method. The execution subject of the embodiment is a second device. The policy negotiation method includes:

[0301] Step 201: The second device sends a first request message to the first device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0302] Step 202: The second device receives a first response message sent by the first device; the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0303] Optionally, the method further comprises:

[0304] In the case where the first request message is used to request the formulation of one or more AI service policies to guarantee the AI ​​service, the first request message includes at least one of the following:

[0305] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type;

[0306] When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following:

[0307] ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information;

[0308] The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service;

[0309] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0310] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0311] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0312] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method;

[0313] The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service;

[0314] The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

[0315] Optionally, when the first device determines that the second device does not meet a preset condition, the first response message is used to indicate a rejection of the first request message of the second device, and the first response message includes a first rejection reason;

[0316] The first rejection reason includes at least one of the following:

[0317] The second device is not allowed to request an AI service policy, the second device is not allowed to initiate an AI service, the number of AI services associated with the first device reaches the upper limit, and the number of AI services associated with the second device reaches the upper limit;

[0318] The preset condition includes at least one of the following:

[0319] The second device is allowed to request an AI service policy, the second device is allowed to initiate an AI service, and the number of AI services associated with the second device does not exceed an upper limit.

[0320] Optionally, when the first response message is used to provide one or more AI service policies for guaranteeing AI services, the first response message includes at least one of the following:

[0321] ASP logo;

[0322] Slice identification;

[0323] DNN;

[0324] One or more AI service identifiers;

[0325] A candidate strategy set corresponding to the AI ​​service identifier one by one;

[0326] The candidate policy set includes information of at least one candidate policy, and the information of the candidate policy includes at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of an AI service; a policy applicable time; and a policy applicable area.

[0327] Optionally, the QoS association ID of the AI ​​service in the first response message is represented by first information, and the first information includes:

[0328] QoS association ID and first parameter;

[0329] The first parameter includes at least one of the following: a parameter value characterizing the time dimension of the service experience; a parameter value characterizing the quality dimension of the service experience; a parameter value characterizing the data density dimension of the service experience; a parameter value characterizing the priority dimension of the service experience; a parameter value characterizing the resource assurance dimension of the service experience; a parameter value characterizing the stability dimension of the service experience; and a parameter value characterizing the resource overhead dimension of the service experience.

[0330] Optionally, the parameter value of the time dimension characterizing the service experience includes: service response delay, the service response delay is obtained based on at least one of the following information, and the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay;

[0331] The parameter value characterizing the quality dimension of the service experience includes: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value;

[0332] The parameter value of the data density dimension characterizing the service experience includes: the amount of data transmitted and processed, and the amount of data transmitted and processed includes at least one of the following: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time;

[0333] The parameter values ​​of the priority dimension characterizing the service experience include: priority level;

[0334] The parameter value of the resource assurance dimension characterizing the service experience includes: resource type, the resource type includes at least one of the following: a guaranteed resource type; a non-guaranteed resource type, a delay-sensitive guaranteed resource type;

[0335] The parameter values ​​representing the stability dimension of the service experience include: time window information;

[0336] The parameter value of the resource overhead characterizing the service experience includes at least one of the following:

[0337] The maximum data burst transfer amount, the maximum computing amount for each computing resource type, and the maximum storage amount.

[0338] Optionally, the method further comprises:

[0339] The second device evaluates the candidate policy set corresponding to the AI ​​service based on the second information;

[0340] The second device sends a first notification message to the first device based on the evaluation result, where the first notification message is used to indicate information about the AI ​​service policy selected by the second device.

[0341] Optionally, the first notification message occurs after the first response message;

[0342] The first notification message includes at least one of the following:

[0343] ASP logo;

[0344] Slice identification;

[0345] DNN;

[0346] AI service logo;

[0347] A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policies include:

[0348] Policy reference identifier;

[0349] QoS association ID of the AI ​​service.

[0350] Optionally, the method further comprises:

[0351] The second device receives a second notification message sent by the first device; the second notification message is used to notify the second target AI service policy that the update or deletion has occurred; the second notification message is sent by the third device to the first device;

[0352] The second device sends a second notification response message to the first device.

[0353] Optionally, the second notification message includes at least one of the following:

[0354] A third message type, wherein the third message type includes one of the following: updating the second target AI service policy, deleting the second target service policy;

[0355] Slice identification;

[0356] DNN;

[0357] One or more AI service identifiers;

[0358] A policy set, wherein the policy set corresponds to the AI ​​service identifier in a one-to-one manner, each of the policy sets includes at least one AI service policy, and the AI ​​service policy includes: a policy reference identifier; and a QoS association ID of the AI ​​service.

[0359] The specific implementation process and technical effects of the method of this embodiment are the same as those in the first device side method embodiment. For details, please refer to the detailed introduction in the first device side embodiment, and will not be repeated here.

[0360] Please refer to Figure 7 The embodiment of the present application provides a policy generation method. The execution subject of the embodiment is a third device. The policy generation method includes:

[0361] Step 301: The third device receives a second request message sent by the first device, where the second request message is used to request generation of one or more AI service policies for guaranteeing artificial intelligence AI services;

[0362] Step 302: The third device generates a candidate policy set corresponding to the AI ​​service based on the second request message.

[0363] Optionally, the third device generates a candidate policy set corresponding to the AI ​​service based on the second request message, including:

[0364] The third device generates a candidate policy set corresponding to the AI ​​service based on the second request message and the local policy of the third device. Optionally, the second request message includes at least one of the following:

[0365] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device;

[0366] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0367] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0368] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0369] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method.

[0370] Optionally, the method further comprises:

[0371] The third device sends a second response message to the first device, where the second response message is used to indicate acceptance or rejection of a second request message of the first device.

[0372] Optionally, when the second response message indicates acceptance of the second request message of the first device, the second response message includes at least one of the following:

[0373] ASP logo;

[0374] Slice identification;

[0375] DNN;

[0376] AI service logo;

[0377] A candidate policy set corresponding to an AI service identifier one by one, the candidate policy set including information of at least one candidate policy, the information of the candidate policy including at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of the AI ​​service; a policy applicable time; and a policy applicable area.

[0378] Optionally, in a case where the second response message is used to indicate rejection of a second request message of the first device, the second response message includes a second rejection reason;

[0379] The second rejection reason includes at least one of the following:

[0380] The model training requirements of the second device are not supported, the model inference requirements of the second device are not supported, the QoS requirements of the AI ​​service of the second device are not supported, and the AI ​​service of the second device is not supported.

[0381] Optionally, the method further comprises:

[0382] The third device receives a first notification message sent by the first device, where the first notification message is used to notify the third device of information about the AI ​​service policy selected by the second device;

[0383] The first notification message includes at least one of the following:

[0384] ASP logo;

[0385] Slice identification;

[0386] DNN;

[0387] AI service logo;

[0388] A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policy includes:

[0389] Policy reference identifier;

[0390] QoS association ID of the AI ​​service.

[0391] Optionally, the method further comprises:

[0392] The third device receives a fourth request message from the fifth device, where the fourth request message is used to request to update or delete the second target AI service policy;

[0393] The fourth request message includes at least one of the following:

[0394] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0395] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0396] Optionally, after the third device updates or deletes the second target AI service policy, the method further includes:

[0397] The third device sends a second notification message to the first device, where the second notification message is used to notify the first device that the second target AI service policy has been updated or deleted;

[0398] The third device receives the first notification response message sent by the first device.

[0399] Optionally, after the third device receives the first notification response message sent by the first device, the method further includes:

[0400] The third device sends a fourth response message to the fifth device, indicating whether the update or deletion of the second target AI service policy is successfully executed.

[0401] Optionally, the method further comprises:

[0402] When the third device receives the fourth request message of the fifth device, the third device sends a fifth request message to the fourth device, where the fifth request message is used to request to update or delete the second target AI service policy;

[0403] The fifth request message includes at least one of the following:

[0404] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0405] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0406] Optionally, the method further comprises:

[0407] When the third device receives the second request message sent by the first device, the third device sends a sixth request message to the fourth device, where the sixth request message is used to request to obtain an AI service policy for the AI ​​service;

[0408] The sixth request message includes at least one of the following:

[0409] The type of data obtained is AI service strategy;

[0410] ASP logo;

[0411] AI service logo;

[0412] Slice identification;

[0413] DNN.

[0414] Optionally, the method further comprises:

[0415] The third device sends a seventh request message to the fourth device, where the seventh request message is used to instruct the fourth device to store information about the AI ​​service policy selected by the second device;

[0416] The seventh request message includes at least one of the following:

[0417] ASP logo;

[0418] Slice identification;

[0419] DNN;

[0420] One or more AI service identifiers;

[0421] A selected policy set, wherein the selected policy set corresponds to the AI ​​service identifier in a one-to-one manner, each of the selected policy sets includes at least one selected AI service policy, and the selected AI service policy includes:

[0422] Policy reference identifier;

[0423] QoS association ID of the AI ​​service.

[0424] Optionally,

[0425] The specific implementation process and technical effects of the method of this embodiment are the same as those in the first device side method embodiment. For details, please refer to the detailed introduction in the first device side embodiment, and will not be repeated here.

[0426] Please refer to Figure 8 The embodiment of the present application provides a policy negotiation method. The execution subject of the embodiment is a fourth device. The policy negotiation method includes:

[0427] Step 401: The fourth device receives a third request message sent by the first device, where the third request message is used to request the fourth device to generate, update or delete a first target AI service policy.

[0428] Optionally, the third request message includes at least one of the following:

[0429] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type;

[0430] When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following:

[0431] ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information;

[0432] The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service;

[0433] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0434] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0435] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0436] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method;

[0437] The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service;

[0438] The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

[0439] Optionally, the fourth device receives a sixth request message sent by the third device, where the sixth request message is used to request to obtain an AI service policy of the AI ​​service; the sixth request message is sent by the third device after receiving the second request message sent by the first device;

[0440] The sixth request message includes at least one of the following:

[0441] The type of data obtained is AI service strategy;

[0442] ASP logo;

[0443] AI service logo;

[0444] Slice identification;

[0445] DNN.

[0446] Optionally, the method further comprises:

[0447] The fourth device receives a fifth request message sent by the third device, where the fifth request message is used to request updating or deleting the second target AI service policy; the fifth request message is sent when the third device receives the fourth request message of the fifth device; the fifth request message includes at least one of the following:

[0448] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0449] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0450] Optionally, the fourth device sends a third response message to the first device; the third response message is used to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0451] Optionally, the fourth device sends a fifth response message to the third device; the second response message is used to indicate whether the update or deletion of the second target AI service policy is successfully executed.

[0452] Optionally, the fourth device sends a sixth response message to the third device, where the sixth response message is used to provide an AI service strategy.

[0453] The specific implementation process and technical effects of the method of this embodiment are the same as those in the first device side method embodiment. For details, please refer to the detailed introduction in the first device side embodiment, and will not be repeated here.

[0454] Please refer to Fig. 9 , an embodiment of the present application provides a policy negotiation method, the execution subject of this embodiment is a fifth device, and the policy negotiation method includes:

[0455] Step 501: The fifth device sends a fourth request message to the third device, where the fourth request message is used to instruct to update or delete the second target AI service policy;

[0456] The fourth request message includes at least one of the following:

[0457] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0458] The updated policy information includes at least one of the following: a policy reference identifier and a QoS association ID of the AI ​​service.

[0459] Optionally, the method further comprises:

[0460] The fifth device receives fourth response information sent by the third device, which is used to indicate whether the update or deletion of the second target AI service policy is successfully executed.

[0461] The specific implementation process and technical effects of the method of this embodiment are the same as those in the first device side method embodiment. For details, please refer to the detailed introduction in the first device side embodiment, and will not be repeated here.

[0462] The policy negotiation method provided in the embodiment of the present application can be executed by a policy negotiation device. The policy generation method provided in the embodiment of the present application can be executed by a policy generation device. In the embodiment of the present application, the policy negotiation device provided in the embodiment of the present application is illustrated by taking the policy negotiation method executed by the policy negotiation device as an example, and the policy generation method executed by the policy generation device as an example to illustrate the policy generation device provided in the embodiment of the present application.

[0463] Fig.10 is one of the structural diagrams of the policy negotiation device provided in the embodiment of the present application, such as Fig.10 As shown, the policy negotiation device includes:

[0464] The receiving module 110 is used to receive a first request message sent by the second device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0465] The sending module 120 is used to send a first response message to the second device, where the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0466] Optionally, when the first request message is used to request formulation of one or more AI service policies for guaranteeing AI services, the first request message includes at least one of the following:

[0467] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, target policy information, and first message type;

[0468] When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following:

[0469] ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information;

[0470] The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service;

[0471] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0472] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0473] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0474] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method;

[0475] The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service;

[0476] The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

[0477] Optionally, the device further comprises:

[0478] A processing module, used to determine whether the second device meets a preset condition;

[0479] The sending module 120 is further used for:

[0480] When it is determined that the second device meets the preset condition and the first request message is used to request the formulation of one or more AI service policies to ensure artificial intelligence AI services, a second request message is sent to a third device, where the second request message is used to request the third device to generate the AI ​​service policy to ensure the AI ​​service;

[0481] In a case where it is determined that the second device does not meet the preset condition, the first response message is used to indicate a rejection of the first request message of the second device, and the first response message includes a first rejection reason;

[0482] The first rejection reason includes at least one of the following:

[0483] The second device is not allowed to request an AI service policy, the second device is not allowed to initiate an AI service, the number of AI services associated with the first device reaches the upper limit, and the number of AI services associated with the second device reaches the upper limit;

[0484] The preset condition includes at least one of the following:

[0485] The second device is allowed to request an AI service policy, the second device is allowed to initiate an AI service, and the number of AI services associated with the second device does not exceed an upper limit.

[0486] Optionally, the second request message includes at least one parameter in the first request message.

[0487] Optionally, before the first device sends the first response message to the second device, the receiving module 110 is further configured to:

[0488] A second response message sent by the third device is received, where the second response message is used to indicate acceptance or rejection of the second request message of the first device.

[0489] Optionally, when the second response message indicates acceptance of the second request message of the first device, the second response message includes at least one of the following:

[0490] ASP logo;

[0491] Slice identification;

[0492] DNN;

[0493] One or more AI service identifiers;

[0494] A candidate strategy set corresponding to the AI ​​service identifier one by one;

[0495] The candidate policy set includes information of at least one candidate policy, and the information of the candidate policy includes at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of an AI service; a policy applicable time; and a policy applicable area.

[0496] Optionally, when the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, the first response message includes at least one of the following:

[0497] ASP logo;

[0498] Slice identification;

[0499] DNN;

[0500] One or more AI service identifiers;

[0501] A candidate strategy set corresponding to the AI ​​service identifier one by one;

[0502] The candidate policy set includes information of at least one candidate policy, and the information of the candidate policy includes at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of an AI service; a policy applicable time; and a policy applicable area.

[0503] Optionally, when the second response message is used to indicate rejection of the second request message of the first device, the second response message includes a second rejection reason; and the second rejection reason includes at least one of the following:

[0504] The model training requirements of the second device are not supported, the model inference requirements of the second device are not supported, the QoS requirements of the AI ​​service of the second device are not supported, and the AI ​​service of the second device is not supported.

[0505] Optionally, the QoS association ID of the AI ​​service in the first response message is represented by first information, and the first information includes:

[0506] QoS association ID and first parameter;

[0507] The first parameter includes at least one of the following: a parameter value characterizing the time dimension of the service experience; a parameter value characterizing the quality dimension of the service experience; a parameter value characterizing the data density dimension of the service experience; a parameter value characterizing the priority dimension of the service experience; a parameter value characterizing the resource assurance dimension of the service experience; a parameter value characterizing the stability dimension of the service experience; and a parameter value characterizing the resource overhead dimension of the service experience.

[0508] Optionally, the parameter value of the time dimension characterizing the service experience includes: service response delay, the service response delay is obtained based on at least one of the following information, and the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay;

[0509] The parameter value characterizing the quality dimension of the service experience includes: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value;

[0510] The parameter value of the data density dimension characterizing the service experience includes: the amount of data transmitted and processed, and the amount of data transmitted and processed includes at least one of the following: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time;

[0511] The parameter values ​​of the priority dimension characterizing the service experience include: priority level;

[0512] The parameter value of the resource assurance dimension characterizing the service experience includes: resource type, the resource type includes at least one of the following: a guaranteed resource type; a non-guaranteed resource type, a delay-sensitive guaranteed resource type;

[0513] The parameter values ​​representing the stability dimension of the service experience include: time window information;

[0514] The parameter value of the resource overhead characterizing the service experience includes at least one of the following:

[0515] The maximum data burst transfer amount, the maximum computing amount for each computing resource type, and the maximum storage amount.

[0516] Optionally, the sending module 120 is further configured to:

[0517] When it is determined that the second device meets the preset condition and the first request message is used to request generation, update or deletion of the first target AI service policy, a third request message is sent to the fourth device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy;

[0518] The receiving module 110 is also used for:

[0519] Receiving a third response message sent by the fourth device; the third response message is used to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed;

[0520] The first device sends the third response message to the second device;

[0521] The third request message includes at least one parameter in the first request message.

[0522] Optionally, the receiving module 110 is further configured to:

[0523] A first notification message from the second device is received, where the first notification message is used to indicate information about the AI ​​service policy selected by the second device.

[0524] Optionally, the first notification message occurs after the first response message;

[0525] The first notification message includes at least one of the following:

[0526] ASP logo;

[0527] Slice identification;

[0528] DNN;

[0529] AI service logo;

[0530] A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policy includes:

[0531] Policy reference identifier;

[0532] QoS association ID of the AI ​​service.

[0533] Optionally, the sending module 120 is further configured to:

[0534] The first notification message is sent to the third device, where the first notification message is used to notify the third device of information about the AI ​​service policy selected by the second device.

[0535] Optionally, the receiving module 110 is further configured to:

[0536] receiving a second notification message from the third device, where the second notification message is used to notify that the second target AI service policy has been updated or deleted;

[0537] Optionally, the sending module 120 is further configured to:

[0538] Send a first notification response message to the third device.

[0539] Optionally, the sending module 120 is further configured to:

[0540] After receiving the second notification message from the third device, sending the second notification message to the second device;

[0541] Optionally, the receiving module 110 is further configured to:

[0542] Receive a second notification response message sent by the second device.

[0543] Optionally, the second notification message includes at least one of the following:

[0544] A third message type, wherein the third message type includes one of the following: updating the second target AI service policy, deleting the second target service policy;

[0545] Slice identification;

[0546] DNN;

[0547] One or more AI service identifiers;

[0548] A policy set, wherein the policy set corresponds to the AI ​​service identifier one by one, each of the policy sets includes at least one AI service policy, and the AI ​​service policy includes: a policy reference identifier; and a QoS association ID of the AI ​​service.

[0549] The device of this embodiment can be used to execute the method of any one of the embodiments in the aforementioned first device side method embodiment. Its specific implementation process and technical effects are the same as those in the first device side method embodiment. For details, please refer to the detailed introduction in the first device side embodiment, which will not be repeated here.

[0550] Fig.11 This is a second structural diagram of the policy negotiation device provided in the embodiment of the present application, such as Fig.11 As shown, the policy negotiation device includes:

[0551] A sending module 210 is used to send a first request message to a first device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy;

[0552] The receiving module 220 is used to receive a first response message sent by the first device; the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

[0553] Optionally, when the first request message is used to request formulation of one or more AI service policies for guaranteeing AI services, the first request message includes at least one of the following:

[0554] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type;

[0555] When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following:

[0556] ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information;

[0557] The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service;

[0558] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0559] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0560] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0561] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method;

[0562] The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service;

[0563] The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

[0564] Optionally, when the first device determines that the second device does not meet a preset condition, the first response message is used to indicate rejection of the first request message of the second device, and the first response message includes a first rejection reason;

[0565] The first rejection reason includes at least one of the following:

[0566] The second device is not allowed to request an AI service policy, the second device is not allowed to initiate an AI service, the number of AI services associated with the first device reaches the upper limit, and the number of AI services associated with the second device reaches the upper limit;

[0567] The preset condition includes at least one of the following:

[0568] The second device is allowed to request an AI service policy, the second device is allowed to initiate an AI service, and the number of AI services associated with the second device does not exceed an upper limit.

[0569] Optionally, when the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, the first response message includes at least one of the following:

[0570] ASP logo;

[0571] Slice identification;

[0572] DNN;

[0573] One or more AI service identifiers;

[0574] A candidate strategy set corresponding to the AI ​​service identifier one by one;

[0575] The candidate policy set includes information of at least one candidate policy, and the information of the candidate policy includes at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of an AI service; a policy applicable time; and a policy applicable area.

[0576] Optionally, the QoS association ID of the AI ​​service in the first response message is represented by first information, and the first information includes:

[0577] QoS association ID and first parameter;

[0578] The first parameter includes at least one of the following: a parameter value characterizing the time dimension of the service experience; a parameter value characterizing the quality dimension of the service experience; a parameter value characterizing the data density dimension of the service experience; a parameter value characterizing the priority dimension of the service experience; a parameter value characterizing the resource assurance dimension of the service experience; a parameter value characterizing the stability dimension of the service experience; and a parameter value characterizing the resource overhead dimension of the service experience.

[0579] Optionally, the parameter value of the time dimension characterizing the service experience includes: service response delay, the service response delay is obtained based on at least one of the following information, and the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay;

[0580] The parameter value characterizing the quality dimension of the service experience includes: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value;

[0581] The parameter value of the data density dimension characterizing the service experience includes: the amount of data transmitted and processed, and the amount of data transmitted and processed includes at least one of the following: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time;

[0582] The parameter values ​​of the priority dimension characterizing the service experience include: priority level;

[0583] The parameter value of the resource assurance dimension characterizing the service experience includes: resource type, the resource type includes at least one of the following: a guaranteed resource type; a non-guaranteed resource type, a delay-sensitive guaranteed resource type;

[0584] The parameter values ​​representing the stability dimension of the service experience include: time window information;

[0585] The parameter value of the resource overhead characterizing the service experience includes at least one of the following:

[0586] The maximum data burst transfer amount, the maximum computing amount for each computing resource type, and the maximum storage amount.

[0587] Optionally, the device further comprises:

[0588] A processing module, configured to evaluate the candidate policy set corresponding to the AI ​​service based on the second information;

[0589] The sending module 210 is further used to send a first notification message to the first device based on the evaluation result, where the first notification message is used to indicate information about the AI ​​service strategy selected by the second device.

[0590] Optionally, the first notification message occurs after the first response message;

[0591] The first notification message includes at least one of the following:

[0592] ASP logo;

[0593] Slice identification;

[0594] DNN;

[0595] AI service logo;

[0596] A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policies include:

[0597] Policy reference identifier;

[0598] QoS association ID of the AI ​​service.

[0599] Optionally, the receiving module 220 is further configured to:

[0600] receiving a second notification message sent by the first device; the second notification message is used to notify the second target AI service policy that the update or deletion has occurred; the second notification message is sent by the third device to the first device;

[0601] The sending module 210 is further configured to send a second notification response message to the first device.

[0602] Optionally, the second notification message includes at least one of the following:

[0603] A third message type, wherein the third message type includes one of the following: updating the second target AI service policy, deleting the second target service policy;

[0604] Slice identification;

[0605] DNN;

[0606] One or more AI service identifiers;

[0607] A policy set, wherein the policy set corresponds to the AI ​​service identifier one by one, each of the policy sets includes at least one AI service policy, and the AI ​​service policy includes: a policy reference identifier; and a QoS association ID of the AI ​​service.

[0608] The device of this embodiment can be used to execute the method of any one of the embodiments in the aforementioned second device side method embodiment. Its specific implementation process and technical effects are the same as those in the second device side method embodiment. For details, please refer to the detailed introduction in the second device side embodiment, which will not be repeated here.

[0609] Fig.12 is one of the structural diagrams of the strategy generation device provided in the embodiment of the present application, such as Fig.12 As shown, the strategy generating device comprises:

[0610] The receiving module 310 is used to receive a second request message sent by the first device, where the second request message is used to request the generation of one or more AI service policies for guaranteeing artificial intelligence AI services;

[0611] The processing module 320 is used to generate a candidate policy set corresponding to the AI ​​service based on the second request message.

[0612] Optionally, the processing module 320 is specifically configured to:

[0613] Based on the second request message and the local policy of the third device, a candidate policy set corresponding to the AI ​​service is generated.

[0614] Optionally, the second request message includes at least one of the following:

[0615] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device;

[0616] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0617] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0618] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0619] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method.

[0620] Optionally, the device further comprises:

[0621] A sending module is used to send a second response message to the first device, where the second response message is used to indicate acceptance or rejection of a second request message of the first device.

[0622] Optionally, when the second response message indicates acceptance of the second request message of the first device, the second response message includes at least one of the following:

[0623] ASP logo;

[0624] Slice identification;

[0625] DNN;

[0626] AI service logo;

[0627] A candidate policy set corresponding to an AI service identifier one by one, the candidate policy set including information of at least one candidate policy, the information of the candidate policy including at least one of the following: a policy reference identifier; a model identifier; an achieved model training accuracy; an achieved model reasoning accuracy; a QoS association ID of the AI ​​service; a policy applicable time; and a policy applicable area.

[0628] Optionally, in a case where the second response message is used to indicate rejection of a second request message of the first device, the second response message includes a second rejection reason;

[0629] The second rejection reason includes at least one of the following:

[0630] The model training requirements of the second device are not supported, the model inference requirements of the second device are not supported, the QoS requirements of the AI ​​service of the second device are not supported, and the AI ​​service of the second device is not supported.

[0631] Optionally, the receiving module 310 is further configured to:

[0632] receiving a first notification message sent by the first device, where the first notification message is used to notify the third device of information about the AI ​​service policy selected by the second device;

[0633] The first notification message includes at least one of the following:

[0634] ASP logo;

[0635] Slice identification;

[0636] DNN;

[0637] AI service logo;

[0638] A selected policy set, each of the selected policy sets including at least one selected AI service policy, wherein the selected AI service policy includes:

[0639] Policy reference identifier;

[0640] QoS association ID of the AI ​​service.

[0641] Optionally, the receiving module 310 is further configured to:

[0642] receiving a fourth request message from a fifth device, where the fourth request message is used to request updating or deleting the second target AI service policy;

[0643] The fourth request message includes at least one of the following:

[0644] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0645] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0646] Optionally, the sending module is further used for:

[0647] After updating or deleting the second target AI service policy, sending a second notification message to the first device, where the second notification message is used to notify the first device that the second target AI service policy has been updated or deleted;

[0648] The receiving module 310 is further used to: receive a first notification response message sent by the first device.

[0649] Optionally, the sending module is further used for:

[0650] After receiving the first notification response message sent by the first device, a fourth response message is sent to the fifth device to indicate whether the update or deletion of the second target AI service policy is successfully executed.

[0651] Optionally, the sending module is further used for:

[0652] When the third device receives the fourth request message of the fifth device, sending a fifth request message to the fourth device, where the fifth request message is used to request to update or delete the second target AI service policy;

[0653] The fifth request message includes at least one of the following:

[0654] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0655] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0656] Optionally, the sending module is further used for:

[0657] When the third device receives the second request message sent by the first device, sending a sixth request message to the fourth device, where the sixth request message is used to request to obtain an AI service policy for the AI ​​service;

[0658] The sixth request message includes at least one of the following:

[0659] The type of data obtained is AI service strategy;

[0660] ASP logo;

[0661] AI service logo;

[0662] Slice identification;

[0663] DNN.

[0664] Optionally, the sending module is further used for:

[0665] Sending a seventh request message to the fourth device, where the seventh request message is used to instruct the fourth device to store information about the AI ​​service policy selected by the second device;

[0666] The seventh request message includes at least one of the following:

[0667] ASP logo;

[0668] Slice identification;

[0669] DNN;

[0670] One or more AI service identifiers;

[0671] A selected policy set, wherein the selected policy set corresponds to the AI ​​service identifier in a one-to-one manner, each of the selected policy sets includes at least one selected AI service policy, and the selected AI service policy includes:

[0672] Policy reference identifier;

[0673] QoS association ID of the AI ​​service.

[0674] The device of this embodiment can be used to execute the method of any one of the embodiments in the aforementioned third device side method embodiment. Its specific implementation process and technical effects are the same as those in the third device side method embodiment. For details, please refer to the detailed introduction in the third device side embodiment, which will not be repeated here.

[0675] Fig.13 This is a third structural diagram of the policy negotiation device provided in the embodiment of the present application, such as Fig.13 As shown, the policy negotiation device includes:

[0676] The receiving module 410 is used to receive a third request message sent by the first device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy.

[0677] Optionally, the third request message includes at least one of the following:

[0678] Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type;

[0679] When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following:

[0680] ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information;

[0681] The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service;

[0682] The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated;

[0683] The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and the total number of data sources; the second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: inference business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and external data source;

[0684] The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method;

[0685] The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method;

[0686] The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service;

[0687] The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

[0688] The device of this embodiment can be used to execute the method of any one of the embodiments in the aforementioned fourth device side method embodiment. Its specific implementation process and technical effects are the same as those in the fourth device side method embodiment. For details, please refer to the detailed introduction in the fourth device side embodiment, which will not be repeated here.

[0689] Fig.14 This is a fourth structural diagram of the policy negotiation device provided in the embodiment of the present application, such as Fig.14 As shown, the policy negotiation device includes:

[0690] A sending module 510 is used to send a fourth request message to the third device, where the fourth request message is used to instruct to update or delete the second target AI service policy;

[0691] The fourth request message includes at least one of the following:

[0692] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0693] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0694] Optionally, the device further comprises:

[0695] A receiving module is used to receive fourth response information sent by the third device, which is used to indicate whether the update or deletion of the second target AI service policy is successfully executed.

[0696] The device of this embodiment can be used to execute the method of any one of the embodiments in the aforementioned fifth device side method embodiment. Its specific implementation process and technical effects are the same as those in the fifth device side method embodiment. For details, please refer to the detailed introduction in the fifth device side embodiment, which will not be repeated here.

[0697] The policy negotiation device and policy generation device in the embodiment of the present application can be an electronic device, such as an electronic device with an operating system, or a component in an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal, or it can be other devices other than a terminal. Exemplarily, the terminal can include but is not limited to the types of terminals 11 listed above, and other devices can be servers, network attached storage (NAS), etc., which are not specifically limited in the embodiment of the present application.

[0698] The policy negotiation device and policy generation device provided in the embodiments of the present application can achieve Figures 2 to 9 The various processes implemented by the method embodiment and achieving the same technical effect are not described here to avoid repetition.

[0699] like Fig.15 As shown, the embodiment of the present application also provides a communication device 1500, including a processor 1501 and a memory 1502, and the memory 1502 stores a program or instruction that can be run on the processor 1501. For example, when the communication device 1500 is a terminal, the program or instruction is executed by the processor 1501 to implement the various steps of the above-mentioned policy negotiation method embodiment, and can achieve the same technical effect. When the communication device 1500 is a network side device, the program or instruction is executed by the processor 1501 to implement the various steps of the above-mentioned policy negotiation or policy generation method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0700] The embodiment of the present application also provides a terminal, including a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the following Fig. 9 The steps in the method embodiment shown. This terminal embodiment corresponds to the above-mentioned terminal side method embodiment, and each implementation process and implementation method of the above-mentioned method embodiment can be applied to this terminal embodiment and can achieve the same technical effect. Specifically, Fig.16 A schematic diagram of the hardware structure of a terminal for implementing an embodiment of the present application.

[0701] The terminal 1600 includes but is not limited to: a radio frequency unit 1601, a network module 1602, an audio output unit 1603, an input unit 1604, a sensor 1605, a display unit 1606, a user input unit 1607, an interface unit 1608, a memory 1609 and at least some of the components of the processor 1610.

[0702] Those skilled in the art will appreciate that the terminal 1600 may also include a power source (such as a battery) for supplying power to each component, and the power source may be logically connected to the processor 1610 through a power management system, thereby implementing functions such as managing charging, discharging, and power consumption management through the power management system. Fig.16 The terminal structure shown in the figure does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently, which will not be described in detail here.

[0703] It should be understood that in the embodiment of the present application, the input unit 1604 may include a graphics processing unit (GPU) 16041 and a microphone 16042, and the graphics processor 16041 processes the image data of a static picture or video obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The display unit 1606 may include a display panel 16061, and the display panel 16061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 1607 includes a touch panel 16071 and at least one of other input devices 16072. The touch panel 16071 is also called a touch screen. The touch panel 16071 may include two parts: a touch detection device and a touch controller. Other input devices 16072 may include, but are not limited to, a physical keyboard, function keys (such as a volume control button, a switch button, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

[0704] In the embodiment of the present application, after receiving downlink data from the network side device, the RF unit 1601 can transmit the data to the processor 1610 for processing; in addition, the RF unit 1601 can send uplink data to the network side device. Generally, the RF unit 1601 includes but is not limited to an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

[0705] The memory 1609 can be used to store software programs or instructions and various data. The memory 1609 may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instruction required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory 1609 may include a volatile memory or a non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), an enhanced synchronous dynamic random access memory (ESDRAM), a synchronous link dynamic random access memory (SLDRAM) and a direct memory bus random access memory (DRRAM). The memory 1609 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

[0706] The processor 1610 may include one or more processing units; optionally, the processor 1610 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor 1610.

[0707] The radio frequency unit 1601 is used to send a fourth request message to the third device, where the fourth request message is used to instruct to update or delete the second target AI service policy;

[0708] The fourth request message includes at least one of the following:

[0709] The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner.

[0710] The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

[0711] Optionally, the radio frequency unit 1601 is further configured to:

[0712] Receive fourth response information sent by the third device, which is used to indicate whether the update or deletion of the second target AI service policy is successfully executed.

[0713] It can be understood that the implementation process of each implementation method mentioned in this embodiment can refer to the relevant description of the fifth device side embodiment of the method, and achieve the same or corresponding technical effect. To avoid repetition, it will not be repeated here.

[0714] The embodiment of the present application also provides a network side device, including a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the following Figure 2-Figure 9 The steps of any method embodiment shown. This network side device embodiment corresponds to the above network side device method embodiment, and each implementation process and implementation mode of the above method embodiment can be applied to this network side device embodiment and can achieve the same technical effect.

[0715] Specifically, the embodiment of the present application also provides a network side device. Fig.17 As shown, the network side device 1700 includes: a processor 1701, a network interface 1702 and a memory 1703. The network interface 1702 is, for example, a common public radio interface (CPRI).

[0716] Specifically, the network side device 1700 of the embodiment of the present application further includes: instructions or programs stored in the memory 1703 and executable on the processor 1701, and the processor 1701 calls the instructions or programs in the memory 1703 to execute. Figure 10-Figure 14 The methods executed by the modules shown achieve the same technical effects, and therefore will not be described here in detail to avoid repetition.

[0717] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, each process of the above-mentioned policy negotiation or policy generation method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0718] The processor is the processor in the terminal described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk. In some examples, the readable storage medium may be a non-transient readable storage medium.

[0719] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned policy negotiation or policy generation method embodiments, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0720] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0721] The embodiments of the present application further provide a computer program / program product, which is stored in a storage medium and is executed by at least one processor to implement the various processes of the above-mentioned policy negotiation or policy generation method embodiments, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0722] An embodiment of the present application also provides a communication system, including: a first device, a second device, a third device, a fourth device and a fifth device, wherein the first device can be used to execute the steps of the policy negotiation method as described above, the second device can be used to execute the steps of the policy negotiation method as described above, the third device can be used to execute the steps of the policy generation method as described above, the fourth device can be used to execute the steps of the policy negotiation method as described above, and the fifth device can be used to execute the steps of the policy negotiation method as described above.

[0723] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the method and device in the embodiment of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0724] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of a computer software product plus a necessary general hardware platform, and of course, can also be implemented by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, disk, CD, etc.), including several instructions to enable a terminal or a network-side device to execute the methods described in each embodiment of the present application.

[0725] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms of implementation methods without departing from the purpose of the present application and the scope of protection of the claims, and these implementation methods are all within the protection of the present application.

Claims

1. A policy negotiation method, characterized in that: include: The first device receives a first request message sent by the second device, where the first request message is used to request the formulation of one or more AI service policies to ensure artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy; The first device sends a first response message to the second device, where the first response message is used to provide one or more AI service policies that guarantee the AI ​​service, or to indicate whether generation, update, or deletion of the first target AI service policy is successfully executed.

2. The method according to claim 1, characterized in that In the case where the first request message is used to request the formulation of one or more AI service policies to guarantee the AI ​​service, the first request message includes at least one of the following: Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type; When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following: ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information; The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service; The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated; The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and total number of data sources; The second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: reasoning business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and off-network data source; The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method; The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method; The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service; The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

3. The method according to claim 1 or 2, characterized in that: The method further comprises: The first device determines whether the second device meets a preset condition; When it is determined that the second device meets the preset condition and the first request message is used to request the formulation of one or more AI service policies to ensure artificial intelligence AI services, the first device sends a second request message to the third device, where the second request message is used to request the third device to generate the AI ​​service policy to ensure the AI ​​service; In a case where it is determined that the second device does not meet the preset condition, the first response message is used to indicate a rejection of the first request message of the second device, and the first response message includes a first rejection reason; The first rejection reason includes at least one of the following: The second device is not allowed to request an AI service policy, the second device is not allowed to initiate an AI service, the number of AI services associated with the first device reaches the upper limit, and the number of AI services associated with the second device reaches the upper limit; The preset condition includes at least one of the following: The second device is allowed to request an AI service policy, the second device is allowed to initiate an AI service, and the number of AI services associated with the second device does not exceed an upper limit.

4. The method according to claim 3, characterized in that The second request message includes at least one parameter in the first request message.

5. The method according to claim 3, characterized in that: Before the first device sends a first response message to the second device, the method further includes: The first device receives a second response message sent by the third device, where the second response message is used to indicate acceptance or rejection of a second request message of the first device.

6. The method according to claim 5, characterized in that In a case where the second response message indicates acceptance of the second request message of the first device, the second response message includes at least one of the following: ASP logo; Slice identification; DNN; One or more AI service identifiers; A candidate strategy set corresponding to the AI ​​service identifier one by one; The candidate strategy set includes information of at least one candidate strategy, and the information of the candidate strategy includes at least one of the following: a strategy reference identifier; a model identifier; an achieved model training accuracy; The achieved model reasoning accuracy; QoS association ID of the AI ​​service; policy applicable time; policy applicable area.

7. The method according to any one of claims 1 to 6, characterized in that: In the case where the first response message is used to provide one or more AI service policies for guaranteeing AI services, the first response message includes at least one of the following: ASP logo; Slice identification; DNN; One or more AI service identifiers; A candidate strategy set corresponding to the AI ​​service identifier one by one; The candidate strategy set includes information of at least one candidate strategy, and the information of the candidate strategy includes at least one of the following: a strategy reference identifier; a model identifier; an achieved model training accuracy; The achieved model reasoning accuracy; QoS association ID of the AI ​​service; policy applicable time; policy applicable area.

8. The method according to claim 5, characterized in that In the case where the second response message is used to indicate rejection of the second request message of the first device, the second response message includes a second rejection reason; and the second rejection reason includes at least one of the following: The model training requirements of the second device are not supported, the model inference requirements of the second device are not supported, the QoS requirements of the AI ​​service of the second device are not supported, and the AI ​​service of the second device is not supported.

9. The method according to claim 7, characterized in that: The QoS association ID of the AI ​​service in the first response message is represented by first information, and the first information includes: QoS association ID and first parameter; The first parameter includes at least one of the following: a parameter value characterizing the time dimension of the service experience; a parameter value characterizing the quality dimension of the service experience; a parameter value characterizing the data density dimension of the service experience; a parameter value characterizing the priority dimension of the service experience; a parameter value characterizing the resource assurance dimension of the service experience; a parameter value characterizing the stability dimension of the service experience; and a parameter value characterizing the resource overhead dimension of the service experience.

10. The method according to claim 9, characterized in that The parameter value of the time dimension characterizing the service experience includes: service response delay, the service response delay is obtained based on at least one of the following information, the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay; The parameter value characterizing the quality dimension of the service experience includes: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value; The parameter value of the data density dimension characterizing the service experience includes: the amount of data transmitted and processed, and the amount of data transmitted and processed includes at least one of the following: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time; The parameter values ​​of the priority dimension characterizing the service experience include: priority level; The parameter value of the resource assurance dimension characterizing the service experience includes: resource type, the resource type includes at least one of the following: a guaranteed resource type; a non-guaranteed resource type, a delay-sensitive guaranteed resource type; The parameter values ​​representing the stability dimension of the service experience include: time window information; The parameter value of the resource overhead characterizing the service experience includes at least one of the following: The maximum data burst transfer amount, the maximum computing amount for each computing resource type, and the maximum storage amount.

11. The method according to any one of claims 1 to 10, characterized in that: The method further comprises: When it is determined that the second device meets the preset condition and the first request message is used to request generation, update or deletion of the first target AI service policy, the first device sends a third request message to the fourth device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy; The first device receives a third response message sent by the fourth device; the third response message is used to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed; The first device sends the third response message to the second device; The third request message includes at least one parameter in the first request message.

12. The method according to any one of claims 1 to 11, characterized in that: The method further comprises: The first device receives a first notification message from the second device, where the first notification message is used to indicate information about an AI service policy selected by the second device.

13. The method according to claim 12, characterized in that: The first notification message occurs after the first response message; The first notification message includes at least one of the following: ASP logo; Slice identification; DNN; AI service logo; A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policy includes: Policy reference identifier; QoS association ID of the AI ​​service.

14. The method according to claim 12 or 13, characterized in that The method further comprises: The first device sends the first notification message to the third device, where the first notification message is used to notify the third device of information about the AI ​​service policy selected by the second device.

15. The method according to any one of claims 1 to 14, characterized in that: The method further comprises: The first device receives a second notification message from the third device, where the second notification message is used to notify that the second target AI service policy has been updated or deleted; The first device sends a first notification response message to the third device.

16. The method according to claim 15, characterized in that After the first device receives the second notification message from the third device, the method further includes: The first device sends the second notification message to the second device; The first device receives a second notification response message sent by the second device.

17. The method according to claim 15 or 16, characterized in that The second notification message includes at least one of the following: A third message type, wherein the third message type includes one of the following: updating the second target AI service policy, deleting the second target service policy; Slice identification; DNN; One or more AI service identifiers; A policy set, wherein the policy set corresponds to the AI ​​service identifier one by one, each of the policy sets includes at least one AI service policy, and the AI ​​service policy includes: a policy reference identifier; and a QoS association ID of the AI ​​service.

18. A policy negotiation method, characterized in that: include: The second device sends a first request message to the first device, where the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy; The second device receives a first response message sent by the first device; the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

19. The method according to claim 18, characterized in that In the case where the first request message is used to request the formulation of one or more AI service policies to guarantee the AI ​​service, the first request message includes at least one of the following: Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type; When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following: ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information; The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service; The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated; The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and total number of data sources; The second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: reasoning business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and off-network data source; The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method; The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method; The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service; The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

20. The method according to claim 18 or 19, characterized in that When the first device determines that the second device does not meet the preset condition, the first response message is used to indicate rejection of the first request message of the second device, and the first response message includes a first rejection reason; The first rejection reason includes at least one of the following: The second device is not allowed to request an AI service policy, the second device is not allowed to initiate an AI service, the number of AI services associated with the first device reaches the upper limit, and the number of AI services associated with the second device reaches the upper limit; The preset condition includes at least one of the following: The second device is allowed to request an AI service policy, the second device is allowed to initiate an AI service, and the number of AI services associated with the second device does not exceed an upper limit.

21. The method according to any one of claims 18 to 20, characterized in that: In the case where the first response message is used to provide one or more AI service policies for guaranteeing AI services, the first response message includes at least one of the following: ASP logo; Slice identification; DNN; One or more AI service identifiers; A candidate strategy set corresponding to the AI ​​service identifier one by one; The candidate strategy set includes information of at least one candidate strategy, and the information of the candidate strategy includes at least one of the following: a strategy reference identifier; a model identifier; an achieved model training accuracy; The achieved model reasoning accuracy; QoS association ID of the AI ​​service; policy applicable time; policy applicable area.

22. The method according to claim 21, characterized in that The QoS association ID of the AI ​​service in the first response message is represented by first information, and the first information includes: QoS association ID and first parameter; The first parameter includes at least one of the following: a parameter value characterizing the time dimension of the service experience; a parameter value characterizing the quality dimension of the service experience; a parameter value characterizing the data density dimension of the service experience; a parameter value characterizing the priority dimension of the service experience; a parameter value characterizing the resource assurance dimension of the service experience; a parameter value characterizing the stability dimension of the service experience; and a parameter value characterizing the resource overhead dimension of the service experience.

23. The method according to claim 22, characterized in that The parameter value of the time dimension characterizing the service experience includes: service response delay, the service response delay is obtained based on at least one of the following information, the at least one of the information includes: communication delay, AI processing delay, pure data collection and transmission delay; The parameter value characterizing the quality dimension of the service experience includes: service accuracy, which is obtained based on at least one of the following information, wherein the at least one of the information includes: communication reliability value, AI processing accuracy, and data integrity, accuracy or timeliness value; The parameter value of the data density dimension characterizing the service experience includes: the amount of data transmitted and processed, and the amount of data transmitted and processed includes at least one of the following: data transmission rate, amount of calculation per unit time, and amount of data collected and transmitted per unit time; The parameter values ​​of the priority dimension characterizing the service experience include: priority level; The parameter value of the resource assurance dimension characterizing the service experience includes: resource type, the resource type includes at least one of the following: a guaranteed resource type; a non-guaranteed resource type, a delay-sensitive guaranteed resource type; The parameter values ​​representing the stability dimension of the service experience include: time window information; The parameter value of the resource overhead characterizing the service experience includes at least one of the following: The maximum data burst transfer amount, the maximum computing amount for each computing resource type, and the maximum storage amount.

24. The method according to any one of claims 18 to 23, characterized in that: The method further comprises: The second device evaluates the candidate policy set corresponding to the AI ​​service based on the second information; The second device sends a first notification message to the first device based on the evaluation result, where the first notification message is used to indicate information about the AI ​​service policy selected by the second device.

25. The method according to claim 24, characterized in that: The first notification message occurs after the first response message; The first notification message includes at least one of the following: ASP logo; Slice identification; DNN; AI service logo; A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policies include: Policy reference identifier; QoS association ID of the AI ​​service.

26. The method according to any one of claims 18 to 25, characterized in that The method further comprises: The second device receives a second notification message sent by the first device; the second notification message is used to notify the second target AI service policy that the update or deletion has occurred; the second notification message is sent by the third device to the first device; The second device sends a second notification response message to the first device.

27. The method according to claim 26, characterized in that The second notification message includes at least one of the following: A third message type, wherein the third message type includes one of the following: updating the second target AI service policy, deleting the second target service policy; Slice identification; DNN; One or more AI service identifiers; A policy set, wherein the policy set corresponds to the AI ​​service identifier one by one, each of the policy sets includes at least one AI service policy, and the AI ​​service policy includes: a policy reference identifier; QoS association ID of the AI ​​service.

28. A strategy generation method, characterized in that: include: The third device receives a second request message sent by the first device, where the second request message is used to request generation of one or more AI service policies to ensure artificial intelligence AI services; The third device generates a candidate policy set corresponding to the AI ​​service based on the second request message.

29. The method according to claim 28, characterized in that The third device generates a candidate policy set corresponding to the AI ​​service based on the second request message, including: The third device generates a candidate policy set corresponding to the AI ​​service based on the second request message and the local policy of the third device.

30. The method according to claim 28 or 29, characterized in that The second request message includes at least one of the following: Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device; The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated; The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and total number of data sources; The second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: reasoning business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and off-network data source; The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method; The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method.

31. The method according to any one of claims 28 to 30, characterized in that: The method further comprises: The third device sends a second response message to the first device, where the second response message is used to indicate acceptance or rejection of a second request message of the first device.

32. The method according to claim 31, characterized in that In a case where the second response message indicates acceptance of the second request message of the first device, the second response message includes at least one of the following: ASP logo; Slice identification; DNN; AI service logo; A candidate strategy set corresponding to the AI ​​service identifier one by one, the candidate strategy set including information of at least one candidate strategy, the information of the candidate strategy including at least one of the following: a strategy reference identifier; a model identifier; and the achieved model training accuracy; The achieved model reasoning accuracy; QoS association ID of the AI ​​service; policy applicable time; policy applicable area.

33. The method according to claim 31, characterized in that In a case where the second response message is used to indicate rejection of a second request message of the first device, the second response message includes a second rejection reason; The second rejection reason includes at least one of the following: The model training requirements of the second device are not supported, the model inference requirements of the second device are not supported, the QoS requirements of the AI ​​service of the second device are not supported, and the AI ​​service of the second device is not supported.

34. The method according to any one of claims 28 to 33, characterized in that The method further comprises: The third device receives a first notification message sent by the first device, where the first notification message is used to notify the third device of information about the AI ​​service policy selected by the second device; The first notification message includes at least one of the following: ASP logo; Slice identification; DNN; AI service logo; A selected policy set, each of which includes at least one selected AI service policy, and the selected AI service policy includes: Policy reference identifier; QoS association ID of the AI ​​service.

35. The method according to any one of claims 28 to 34, characterized in that The method further comprises: The third device receives a fourth request message from the fifth device, where the fourth request message is used to request to update or delete the second target AI service policy; The fourth request message includes at least one of the following: The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner. The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

36. The method according to claim 35, characterized in that After the third device updates or deletes the second target AI service policy, the method further includes: The third device sends a second notification message to the first device, where the second notification message is used to notify the first device that the second target AI service policy has been updated or deleted; The third device receives the first notification response message sent by the first device.

37. The method according to claim 36, characterized in that After the third device receives the first notification response message sent by the first device, the method further includes: The third device sends a fourth response message to the fifth device, indicating whether the update or deletion of the second target AI service policy is successfully executed.

38. The method according to any one of claims 35 to 37, characterized in that The method further comprises: When the third device receives the fourth request message of the fifth device, the third device sends a fifth request message to the fourth device, where the fifth request message is used to request to update or delete the second target AI service policy; The fifth request message includes at least one of the following: The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner. The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

39. The method according to any one of claims 28 to 38, characterized in that The method further comprises: When the third device receives the second request message sent by the first device, the third device sends a sixth request message to the fourth device, where the sixth request message is used to request to obtain an AI service policy for the AI ​​service; The sixth request message includes at least one of the following: The type of data obtained is AI service strategy; ASP logo; AI service logo; Slice identification; DNN.

40. The method according to any one of claims 28 to 39, characterized in that The method further comprises: The third device sends a seventh request message to the fourth device, where the seventh request message is used to instruct the fourth device to store information about the AI ​​service policy selected by the second device; The seventh request message includes at least one of the following: ASP logo; Slice identification; DNN; One or more AI service identifiers; A selected policy set, wherein the selected policy set corresponds to the AI ​​service identifier in a one-to-one manner, each of the selected policy sets includes at least one selected AI service policy, and the selected AI service policy includes: Policy reference identifier; QoS association ID of the AI ​​service.

41. A policy negotiation method, characterized in that: include: The fourth device receives a third request message sent by the first device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy.

42. The method according to claim 41, characterized in that The third request message includes at least one of the following: Application service provider ASP identifier, AI service identifier, AI service type, data source information, model training requirements, model reasoning requirements, quality of service QoS requirements for AI services, slice identifier, data network name DNN, expected policy application time, expected policy application area, first indication information, and parameter information defined by the second device, first message type; When the first request message is used to request generation, update, or deletion of a first target AI service policy, the first request message includes at least one of the following: ASP identifier, second message type, slice identifier, data network name DNN, AI service identifier, and target policy information; The AI ​​service type includes at least one of the following: reasoning service, training service, reasoning service and training service; The first indication information is used to indicate whether the first device notifies the second device when the AI ​​service policy is updated; The data source information includes at least one of the following: second indication information, third indication information, terminal data format, one or more terminal identifiers, network data format, external data source format, external data source identifier, and total number of data sources; The second indication information is used to indicate the business scope applicable to the data source, and the business scope includes at least one of the following: reasoning business, training business; the third indication information is used to indicate whether the data source information includes a specific data source, and the specific data source includes at least one of the following: terminal data source, network data source, and off-network data source; The model training requirement includes at least one of the following: whether the second device carries a model, a model identifier, a model size, a model training accuracy, and a training accuracy evaluation method; The model reasoning requirements include at least one of the following: reasoning delay, generalization scope, effective area, effective target, reasoning accuracy, and reasoning accuracy evaluation method; The target policy information includes at least one of the following: a policy reference identifier, a QoS association ID of the AI ​​service; The second message type includes one of the following: generating a first target AI service policy, updating a first target AI service policy, and deleting a first target AI service policy.

43. A policy negotiation method, characterized in that: include: The fifth device sends a fourth request message to the third device, where the fourth request message is used to instruct to update or delete the second target AI service policy; The fourth request message includes at least one of the following: The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner. The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

44. The method according to claim 43, characterized in that The method further comprises: The fifth device receives fourth response information sent by the third device, which is used to indicate whether the update or deletion of the second target AI service policy is successfully executed.

45. A policy negotiation device, characterized in that: include: A receiving module, configured to receive a first request message sent by a second device, wherein the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy; A sending module is used to send a first response message to the second device, where the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

46. ​​A policy negotiation device, characterized in that: include: A sending module, configured to send a first request message to a first device, wherein the first request message is used to request the formulation of one or more AI service policies for guaranteeing artificial intelligence AI services, or to request the generation, update or deletion of a first target AI service policy; A receiving module is used to receive a first response message sent by the first device; the first response message is used to provide one or more AI service policies for guaranteeing the AI ​​service, or to indicate whether the generation, update or deletion of the first target AI service policy is successfully executed.

47. A strategy generation device, characterized in that: include: A receiving module, configured to receive a second request message sent by the first device, where the second request message is used to request generation of one or more AI service policies to ensure artificial intelligence AI services; A processing module is used to generate a candidate policy set corresponding to the AI ​​service based on the second request message.

48. A policy negotiation device, characterized in that: include: A receiving module is used to receive a third request message sent by the first device, where the third request message is used to request the fourth device to generate, update or delete the first target AI service policy.

49. A policy negotiation device, characterized in that: include: A sending module, configured to send a fourth request message to a third device, where the fourth request message is used to instruct to update or delete the second target AI service policy; The fourth request message includes at least one of the following: The third message type includes a slice identifier, a DNN, one or more AI service identifiers, and updated policy information, where the updated policy information corresponds to the AI ​​service identifier in a one-to-one manner. The updated policy information includes at least one of the following: a policy reference identifier, and a QoS association ID of the AI ​​service.

50. A first device, characterized in that: It comprises a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the policy negotiation method as described in any one of claims 1 to 17 are implemented.

51. A second device, characterized in that: It includes a processor and a memory, the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the policy negotiation method as described in any one of claims 18 to 27 are implemented.

52. A third device, characterized in that: It comprises a processor and a memory, wherein the memory stores a program or instruction that can be run on the processor, and when the program or instruction is executed by the processor, the steps of the strategy generation method as described in any one of claims 28 to 40 are implemented.

53. A fourth device, characterized in that: It includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the policy negotiation method as described in any one of claims 41 to 42 are implemented.

54. A fifth device, characterized in that: It includes a processor and a memory, wherein the memory stores programs or instructions that can be run on the processor, and when the programs or instructions are executed by the processor, the steps of the policy negotiation method as described in any one of claims 43 to 44 are implemented.

55. A readable storage medium, characterized in that: The readable storage medium stores programs or instructions, and when the programs or instructions are executed by the processor, they implement the policy negotiation method as described in any one of claims 1 to 17, or implement the steps of the policy negotiation method as described in any one of claims 18 to 27, or implement the steps of the policy generation method as described in any one of claims 28 to 40, or implement the steps of the policy negotiation method as described in any one of claims 41 to 42, or implement the steps of the policy negotiation method as described in any one of claims 43 to 44.

Images