AI service connection establishment method and device, terminal and network side equipment

Abstract

The invention discloses an AI service connection establishment method and device, a terminal and network side equipment, and belongs to the technical field of communication, and the AI service connection establishment method comprises the steps that the terminal sends a first request message to a network side, and the first request message is used for requesting the network side to allocate network resources for an AI service; and the terminal receives a first request response message returned by the network side, wherein the first request response message comprises at least one of the following items: information of a first network element allocated for the AI service by the network side, information of a second network element allocated for the AI service by the network side, and information of a model. In the technical scheme of the invention, the network side can allocate the communication resource and the AI resource for the AI service according to the request of the terminal, and the problem that a wireless mobile communication system cannot provide the AI service for the terminal is solved.

Description

Technical Field

[0001] The present application belongs to the field of communication technology, and specifically relates to a method, device, terminal and network-side equipment for establishing an AI service connection. Background Art

[0002] With the development of wireless mobile communication technology, artificial intelligence (AI) will become one of the pillar technologies of wireless mobile communication, and wireless mobile communication systems need to provide end-to-end support for AI-related services and applications. However, existing wireless mobile communication systems do not provide AI as a general service to terminals. Summary of the invention

[0003] The embodiments of the present application provide a method, apparatus, terminal and network-side equipment for establishing an AI service connection, which can solve the problem that the wireless mobile communication system cannot provide AI services for the terminal.

[0004] In a first aspect, a method for establishing an AI service connection is provided, which is executed by a terminal, and the method includes:

[0005] The terminal sends a first request message to the network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service;

[0006] The terminal receives a first request response message returned by the network side, where the first request response message includes at least one of the following: information about a first network element allocated by the network side to the AI ​​service, information about a second network element allocated by the network side to the AI ​​service, and information about the model.

[0007] In a second aspect, a method for establishing an AI service connection is provided, which is executed by a network side, and the method includes:

[0008] The fourth network element receives a communication resource request message, where the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service;

[0009] The fourth network element allocates communication resources according to the communication resource request message, and returns a communication resource establishment response message, wherein the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[0010] In a third aspect, a method for establishing an AI service connection is provided, which is executed by a network side, and the method includes:

[0011] The fifth network element receives an AI resource establishment request message, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service;

[0012] The fifth network element allocates AI resources according to the AI ​​resource establishment request message, and returns an AI resource establishment response message, where the AI ​​resource establishment response message includes at least one of the following: information about the first network element and information about the second network element.

[0013] In a fourth aspect, a method for establishing an AI service connection is provided, which is executed by a network side, and the method includes:

[0014] The sixth network element receives the first request message, where the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal;

[0015] The sixth network element selects the fifth network element and the fourth network element, sends an AI resource establishment request message to the fifth network element, and sends a communication resource request message to the fourth network element;

[0016] The sixth network element receives the AI ​​resource establishment response message returned by the fifth network element, where the AI ​​resource establishment response message includes information of the first network element allocated for the AI ​​service, and receives the communication resource establishment response message returned by the fourth network element, where the communication resource establishment response message includes information of the second network element allocated for the AI ​​service;

[0017] The sixth network element returns a first request response message, where the first request response message includes information about the first network element and information about the second network element.

[0018] In a fifth aspect, an AI service connection establishment device is provided, which is applied to a terminal, including:

[0019] A first sending module, configured to send a first request message to a network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service;

[0020] The first receiving module is used to receive a first request response message returned by the network side, wherein the first request response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[0021] In a sixth aspect, an AI service connection establishment device is provided, which is applied to a network side, including:

[0022] A second receiving module is used to receive a communication resource request message, where the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service;

[0023] A first processing module, configured to allocate communication resources according to the communication resource request message;

[0024] The second sending module is used to return a communication resource establishment response message, wherein the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[0025] In a seventh aspect, an AI service connection establishment device is provided, which is applied to a network side, including:

[0026] A third receiving module is used to receive an AI resource establishment request message, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service;

[0027] A second processing module, configured to allocate AI resources according to the AI ​​resource establishment request message;

[0028] The third sending module is used to return an AI resource establishment response message, where the AI ​​resource establishment response message includes at least one of the following: information about the first network element and information about the second network element.

[0029] In an eighth aspect, an AI service connection establishment device is provided, which is applied to a network side, including:

[0030] A fourth receiving module, configured to receive a first request message, where the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal;

[0031] A third processing module, used for selecting a fifth network element and a fourth network element;

[0032] a fourth sending module, configured to send an AI resource establishment request message to the fifth network element, and send a communication resource request message to the fourth network element;

[0033] The fourth receiving module is further configured to receive an AI resource establishment response message returned by the fifth network element, the AI ​​resource establishment response message including information of a first network element allocated for the AI ​​service, and receive a communication resource establishment response message returned by the fourth network element, the communication resource establishment response message including information of a second network element allocated for the AI ​​service;

[0034] The fourth sending module is further used to return a first request response message, where the first request response message includes information about the first network element and information about the second network element.

[0035] In a ninth aspect, a terminal is provided, comprising 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 method described in the first aspect are implemented.

[0036] In the tenth aspect, a terminal is provided, comprising a processor and a communication interface, wherein the communication interface is used to send a first request message to the network side, the first request message is used to request the network side to allocate network resources for the AI ​​service; receive a first request response message returned by the network side, the first request response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[0037] In the eleventh aspect, a network side 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 second aspect, the third aspect, or the fourth aspect are implemented.

[0038] In the twelfth aspect, a network side device is provided, including a processor and a communication interface, wherein the communication interface is used to receive a communication resource request message, and the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service; the processor is used to allocate communication resources according to the communication resource request message, and the communication interface is also used to return a communication resource establishment response message, and the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[0039] In a thirteenth aspect, a network side device is provided, including a processor and a communication interface, wherein the communication interface is used to receive an AI resource establishment request message, and the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service; the processor is used to allocate AI resources according to the AI ​​resource establishment request message, and the communication interface is also used to return an AI resource establishment response message, and the AI ​​resource establishment response message includes at least one of the following: information of the first network element and information of the second network element.

[0040] In a fourteenth aspect, a network side device is provided, including a processor and a communication interface, wherein the communication interface is used to receive a first request message, the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal; the processor is used to select a fifth network element and a fourth network element, and the communication interface is also used to send an AI resource establishment request message to the fifth network element and send a communication resource request message to the fourth network element; receive an AI resource establishment response message returned by the fifth network element, the AI ​​resource establishment response message includes information of the first network element allocated for the AI ​​service, and receive a communication resource establishment response message returned by the fourth network element, the communication resource establishment response message includes information of the second network element allocated for the AI ​​service; the sixth network element returns a first request response message, the first request response message includes information of the first network element and information of the second network element.

[0041] In the fifteenth 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, the third aspect, or the fourth aspect are implemented.

[0042] In the sixteenth aspect, a wireless communication system is provided, including: a terminal and a network side device, wherein the terminal can be used to execute the steps of the method described in the first aspect, and the network side device can be used to execute the steps of the method described in the second aspect, the third aspect, or the fourth aspect.

[0043] In the seventeenth 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 steps of the method described in the first aspect, or to implement the steps of the method described in the second aspect, the third aspect, or the fourth aspect.

[0044] In the eighteenth aspect, a computer program / program product is provided, wherein the computer program / program product is stored in a storage medium, and the program / program product is executed by at least one processor to implement the steps of the method described in the first aspect, or to implement the steps of the method described in the second aspect, the third aspect, or the fourth aspect.

[0045] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services according to the request of the terminal, thereby solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also ensures the security of terminal data flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a block diagram of a wireless communication system to which the embodiments of the present application can be applied;

[0047] Figure 2 It is a schematic diagram of the process of establishing a communication transmission channel of the control plane;

[0048] Figure 3 It is a schematic diagram of the process of establishing a communication transmission channel on the user plane;

[0049] Figure 4-Figure 7 This is a flowchart of a method for establishing an AI service connection provided in an embodiment of the present application;

[0050] Figure 8 It is a schematic diagram of the protocol stack provided in Example 1 of the present application;

[0051] Fig. 9 This is a flowchart of the AI ​​service connection establishment method provided in Example 1 of the present application;

[0052] Fig.10 This is a schematic diagram of the protocol stack provided in Example 2 of the present application;

[0053] Fig.11 This is a flowchart of the method for establishing an AI service connection provided in Embodiment 2 of the present application;

[0054] Fig.12 This is a schematic diagram of the protocol stack provided in Example 3 of the present application;

[0055] Fig.13 This is a flowchart of the AI ​​service connection establishment method provided in Embodiment 3 of the present application;

[0056] Fig.14 This is a flowchart of the AI ​​service connection establishment method provided in Embodiment 4 of the present application;

[0057] Fig.15 This is a schematic diagram of the protocol stack provided in Example 5 of the present application;

[0058] Fig.16 This is a flowchart of the AI ​​service connection establishment method provided in Embodiment 5 of the present application;

[0059] Fig.17 This is a schematic diagram of the protocol stack provided in Example 6 of the present application;

[0060] Fig.18 This is a flowchart of the AI ​​service connection establishment method provided in Example 6 of the present application;

[0061] Fig.19 This is a schematic diagram of the protocol stack provided in Example 7 of the present application;

[0062] Figure 20-23It is a structural diagram of an AI service connection establishment device provided in an embodiment of the present application;

[0063] Fig.24 It is a structural diagram of a communication device provided in an embodiment of the present application;

[0064] Fig.25 is a schematic diagram of the structure of a terminal provided in an embodiment of the present application;

[0065] Fig.26 It is a structural diagram of a network side device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0066] 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.

[0067] 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.

[0068] The term "indication" in this application can be either 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.

[0069] 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.

[0070] 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 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 12 may include access network equipment or core network equipment, wherein the access network equipment may also be referred to as a radio access network (Radio Access Network, RAN) equipment, a radio access network function or a radio access network unit. The access network equipment may include a base station, a wireless local area network (Wireless Local Area Network, WLAN) access point (Access Point, AP) 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 (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.

[0071] 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), NAS interface function, communication management function, communication resource node, policy charging function, AI service management function, external service management function, service management function, computing service management function, AI resource management function, resource management function, computing resource management function, AI model management function, model management function, AI resource node, resource node, computing resource node, service resource node, data plane management function, data source, application server, etc. It should be noted that in the embodiment of the present application, only the core network device in the NR system is taken as an example for introduction, and the specific type of the core network device is not limited.

[0072] During UE data transmission, Figure 2 As shown, firstly, a communication transmission channel of the control plane is established. The process of establishing the communication transmission channel of the control plane includes the following steps:

[0073] Step 1. The UE sends a protocol data unit (PDU) session establishment request; the UE non-access-stratum session management (NAS-SM) layer generates a PDU session establishment request message, encapsulates it in an N1 SM container, and forwards it to the non-access-stratum mobility management (NAS-MM) layer, which is then forwarded to the SMF via the AMF. The AMF does not parse the PDU session establishment request;

[0074] Step 2. User plane establishment.

[0075] SMF selects UPF and establishes the user plane. SMF sends the core network channel information (CN tunnel information) to the Next Generation Radio Access Network (NG-RAN).

[0076] Then establish the data transmission channel of the user plane, such as Figure 3 As shown, the data transmission channel establishment process of the data plane includes the following steps:

[0077] Step 1. A data packet comes down from the application layer;

[0078] Step 2. The UE performs a Domain Name System (DNS) query to obtain the server address. The UE sets the destination address of the data packet to the server address and the source address to the UE IP address.

[0079] Step 3. NG-RAN forwards the data packet to UPF, which continues routing according to the destination address.

[0080] The existing wireless mobile communication system only uses AI (such as the Network Data Analytics Function (NWDAF) network element) to perform some operation optimizations, and does not provide AI as a general service to the terminal. In addition, the operator network can only provide a data transmission pipeline with a certain quality of service (transmission delay, packet error rate, etc.). The pipeline is currently only used for data transmission between UE and target nodes, and cannot perform corresponding AI business processing, such as using trained models for rendering tasks, and feeding back reconstructed environmental data to the terminal in the autonomous driving scenario.

[0081] In conjunction with the accompanying drawings, the following describes in detail the AI ​​service connection establishment method, device, terminal, and network side device provided in the embodiments of the present application through some embodiments and their application scenarios.

[0082] It is worth noting that in this embodiment, an AI session refers to a special session between a user terminal and an AI resource node. After the AI ​​session is established, a data transmission channel between the UE and the AI ​​resource node is established. Therefore, in the prior art, if the UE wants to use AI services (such as AI data transmission), an AI session must first be established. In other words, an AI session is an AI resource established between a UE and an AI resource node in a wireless mobile communication network. An AI session may also be referred to as an AI service connection or an AI connection.

[0083] In this embodiment, the sixth network element can specifically be a non-access layer (Non-Access-Stratum, NAS) interface function or AMF, which has the following functions: the termination point of the access network (Access Network, AN) signaling interface (N2), the termination point of NAS (N1) signaling (MM message); responsible for the encryption and integrity protection of NAS messages, and responsible for registration, access, mobility management, authentication, SMS and other functions.

[0084] The fourth network element can be specifically a communication management function or SMF, which has the following functions: termination point of SM message of NAS message; establishment, modification and release of session; allocation management of UE IP; UPF selection and control; collection of billing data and support of billing interface; determination of the secure terminal connection service (Session and Service Continuity, SSC) mode of a session; and downlink data indication.

[0085] The second network element can specifically be a communication resource node or UPF, which has the following functions: an anchor point for intra-Radio Access Technology (RAT) / inter-RAT mobility (when applicable); a session point for interconnection between external PDUs and data networks; packet routing and forwarding data packet inspection; and implementation of some user plane policy rules, such as gating, redirection, and traffic steering.

[0086] The seventh network element may specifically be a policy charging function or PCF, which has the following functions: application and service data flow detection, and policy (UE policy, session policy, etc.) generation and provision.

[0087] The fifth network element can specifically be an AI service management function, an external service management function, a service management function or a computing power service management function, and has the following functions: used for service management, service identifier resolution, specifically, converting external AI service requests into internal AI resource management functions and / or communication resource management processes; determining the required AI resources considering the AI ​​service category and the quality of service (QoS) requirements of the AI ​​service.

[0088] The ninth network element can specifically be an AI model management function or a model management function, and has the following functions: used for model management, specifically, determining a suitable model based on an external model request.

[0089] The eighth network element may specifically be an AI resource management function, a resource management function or a computing power resource management function, and has the following functions: used for business resource management, specifically, allocation, update and release of AI and / or computing power resources on demand; status update and maintenance of AI and / or computing power resource nodes.

[0090] The first network element may specifically be an AI resource node, a resource node, a computing resource node or a business resource node, and has the following functions: used for business processing and business policy rule implementation, specifically, performing computational processing of AI business according to the command of the AI ​​resource management function (user plane execution process); and updating the AI ​​resource status to the AI ​​resource management function.

[0091] The tenth network element may specifically be a data plane management function, having the following functions: used to perform data management, including determining a data source according to a data request.

[0092] The third network element may specifically be a data source, and has the following functions: used for generating, storing and forwarding data.

[0093] The eleventh network element may specifically be an application server (AS), which is used to execute external service processing, specifically to process and feed back the data forwarded by the resource node.

[0094] See also Figure 4 , Figure 4 is a flowchart of a method for establishing an AI service connection provided in an embodiment of the present application, the method being executed by a terminal, such as Figure 4 As shown, the method comprises the following steps:

[0095] Step 101: The terminal sends a first request message to the network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service;

[0096] Step 102: The terminal receives a first request response message returned by the network side, where the first request response message includes at least one of the following: information about a first network element allocated by the network side for the AI ​​service, information about a second network element allocated by the network side for the AI ​​service, and information about the model.

[0097] In an embodiment of the present application, the terminal requests the network side to allocate network resources for the AI ​​service. The network side can allocate communication resources and AI resources for the AI ​​service according to the request of the terminal, thereby solving the problem that the wireless mobile communication system cannot provide AI services for the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[0098] It should be noted that network resources are a total resource set including communication resources, AI resources and / or policy resources, and are a general term for various resources that provide reliable guarantee links for a certain service. Communication resources are a set of resources provided by the existing 3GPP to terminals or third parties for communication services, including but not limited to communication links, allocated and scheduled frequency bands, memory of base stations or network elements, signaling forwarding and other resources. AI resources are a set of resources to ensure AI services, including but not limited to the computing power of high-performance computing machines (CPU / GPU, etc.), storage capacity for large amounts of training data, etc. Mapped to the actual link, it can be understood as a PDU session and an AI session. The former belongs to a session on the communication resource link, and the latter belongs to a session on the AI ​​resource link.

[0099] In some embodiments, the first request message includes at least one of the following:

[0100] AI business requirements description information;

[0101] Data network name DNN;

[0102] Single Network Slice Selection Assistance Information S-NSSAI;

[0103] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[0104] PDU session information;

[0105] Policy reference identifier, including the AI ​​service policy indication negotiated between the UE and the policy charging function;

[0106] Address of the application server AS.

[0107] Based on the information included in the above-mentioned first request message, the central control node on the network side can convert the business requirements of the AI ​​service into internal policies that can be executed on the network side, including communication policies or AI policies, etc., wherein the communication policy includes the QoS rules corresponding to the existing PDU session, and the AI ​​policy includes the AI ​​service processing delay, AI service processing accuracy, etc. For example, when the central control node is an AI service management function, it decides on demand whether to interact with other AI network elements according to the AI ​​policy to obtain relevant configuration information (for example, if the model is not provided, the AI ​​service management function needs to interact with the AI ​​model management function to obtain the model; if the first request message contains an indication that the data source is the data plane, the AI ​​service management function needs to interact with the data plane to obtain the data source); the configuration information may include at least one of the following: model, model download information (such as model download address), data source information (data source address), AI service computing capacity (computing power size, type).

[0108] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network assigns the AI ​​service quality requirements to the user.

[0109] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0110] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0111] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0112] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0113] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0114] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0115] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0116] In some embodiments, the type of the first request message may include at least one of the following:

[0117] Communication resource request message, used to request the network to allocate communication resources for AI services;

[0118] AI resource request message, used to request the network side to allocate AI resources for AI services;

[0119] Non-access layer NAS message, used to request the network side to allocate network resources for AI services.

[0120] The terminal can request the network to allocate resources for the AI ​​service through the above message types.

[0121] In some embodiments, the communication resource request message may include at least one of the following:

[0122] A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[0123] PDU session modification request message, used to modify the PDU session;

[0124] Communication channel establishment request message, used to establish a communication bearer;

[0125] Communication channel modification request message, used to modify the communication bearer;

[0126] The AI ​​resource request message includes at least one of the following:

[0127] An AI resource establishment request message is used to establish an AI session, where the AI ​​session is a data transmission channel between the first network element and the terminal;

[0128] AI session establishment request message;

[0129] The NAS message includes at least one of the following:

[0130] Registration message;

[0131] Service request message;

[0132] Uplink NAS transmission message;

[0133] The first control message is used to establish or modify a control signaling connection.

[0134] In this embodiment, the terminal can request the network side to allocate resources for the AI ​​service through any one of the above messages or any multiple messages.

[0135] In some embodiments, after the terminal receives the first request response message returned by the network side, the method further includes:

[0136] The terminal sends AI service data to the first network element, where the AI ​​service data is carried by the AI ​​session and is routed to the first network element via the second network element;

[0137] The terminal receives the AI ​​service data processing result returned by the first network element.

[0138] In some embodiments, the AI ​​service data processing result is obtained by the first network element based on at least one of a model, terminal data and in-network data, and the in-network data is obtained by the first network element from a third network element.

[0139] In this embodiment, the network side can be responsible for the overall processing of an AI service, and the user plane path of the AI ​​service data is: UE→communication resource node→AI resource node. After the UE obtains the information of the AI ​​resource node and the communication resource node, the processing of the AI ​​service data includes the following two methods:

[0140] (1) The UE sets the destination address of the AI ​​service data to the AI ​​resource node.

[0141] The UE encapsulates the AI ​​service data in a data packet, and the data packet is routed to the communication resource node. The communication resource node finds that the destination address is not itself, and then encapsulates the data packet and routes it to the AI ​​resource node.

[0142] (2) The UE sets the destination address of the AI ​​service data to the communication resource node.

[0143] The UE obtains the communication resource node information, and the communication resource node obtains the AI ​​resource node information through the communication management function. The UE encapsulates the address of the communication resource node in a data packet, the data packet is routed to the communication resource node, and the communication resource node uses the information of the AI ​​resource node to forward it to the corresponding AI resource node.

[0144] Alternatively, the network side can be responsible for partial processing of an AI service. In this case, the participation of an external AS is required to process the remaining part of an AI service. The user plane path of AI service data is: UE→communication resource node→AI resource node→AS. After the UE obtains the information of the AI ​​resource node and the communication resource node, the processing of AI service data includes the following three methods:

[0145] (1) The UE sets the destination address of the AI ​​service data to the AS address. The UE can use DNS query to obtain the AS address; the communication resource node obtains the AI ​​resource node information through the communication management function.

[0146] The UE encapsulates the AS address in the data packet, and the data packet is routed to the communication resource node. The communication resource node encapsulates the data packet according to the destination address and the AI ​​resource node address and routes it to the AI ​​resource node. The AI ​​resource node encapsulates the data packet according to the destination address and routes it to the AS.

[0147] (2) The UE sets the destination address of the AI ​​service data to the AI ​​resource node. The UE can use DNS query to obtain the AS address; the AI ​​resource node obtains the AS address through the AI ​​resource management function.

[0148] The UE encapsulates the AI ​​resource node address in the data packet, and the data packet is routed to the communication resource node. The communication resource node encapsulates the data packet according to the destination address and routes it to the AI ​​resource node. The AI ​​resource node encapsulates the data packet according to the AS address and routes it to the AS.

[0149] (3) The UE sets the destination address of the AI ​​service data to the communication resource node. The UE can use DNS query to obtain the AS address; the AI ​​resource node obtains the AS address through the AI ​​resource management function; and the communication resource node obtains the AI ​​resource node information through the communication management function.

[0150] The UE encapsulates the communication resource node address in the data packet, and the data packet is routed to the communication resource node. The communication resource node encapsulates the data packet according to the AI ​​resource node address and routes it to the AI ​​resource node. The AI ​​resource node encapsulates the data packet according to the AS address and routes it to the AS.

[0151] See also Figure 5 , Figure 5 is a flowchart of a method for establishing an AI service connection provided in an embodiment of the present application, which is executed by the network side, such as Figure 5 As shown, the method comprises the following steps:

[0152] Step 201: The fourth network element receives a communication resource request message, where the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service;

[0153] Step 202: The fourth network element allocates communication resources according to the communication resource request message, and returns a communication resource establishment response message, wherein the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[0154] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services, solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[0155] In some embodiments, the fourth network element needs to satisfy at least one of the following:

[0156] The fourth network element supports forwarding the AI ​​resource request message;

[0157] The fourth network element supports establishing a bearer for the AI ​​service;

[0158] The fourth network element supports communication with the AI ​​network function;

[0159] The fourth network element supports selection of AI network functions.

[0160] In some embodiments, the communication resource request message includes at least one of the following:

[0161] Terminal identification;

[0162] Protocol data unit PDU session identifier;

[0163] The associated AI session identifier;

[0164] Data network name DNN;

[0165] Single Network Slice Selection Assistance Information S-NSSAI;

[0166] Session type indication, used to indicate whether the session is a PDU session or an AI session;

[0167] PDU session information, used to indicate the attributes of the PDU session;

[0168] AI business requirements description information;

[0169] the address of the fifth network element;

[0170] Address of the application server AS.

[0171] Based on the information included in the above communication resource request message, the central control node on the network side can convert the business requirements of the AI ​​service into internal policies executable on the network side, including communication policies, wherein the communication policies include QoS rules corresponding to existing PDU sessions.

[0172] In some embodiments, the type of the communication resource request message includes at least one of the following:

[0173] A PDU session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[0174] PDU session modification request message, used to modify the PDU session;

[0175] Communication channel establishment request message, used to establish a communication bearer;

[0176] The communication channel modification request message is used to modify the communication bearer.

[0177] In this embodiment, resources may be requested for the AI ​​service through any of the above message types.

[0178] In some embodiments, before the fourth network element allocates communication resources according to the communication resource request message, the method further includes:

[0179] The fourth network element selects a fifth network element, and sends an AI resource establishment request message to the fifth network element, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service, and the AI ​​resource establishment request message is generated by the terminal or the fourth network element;

[0180] The fourth network element receives the AI ​​resource establishment response message returned by the fifth network element, where the AI ​​resource establishment response message includes information of the first network element allocated to the AI ​​service.

[0181] In this embodiment, the fourth network element may first obtain information about the first network element, that is, the AI ​​resource node, and select the second network element (communication resource node) according to the communication strategy and the information about the AI ​​resource node.

[0182] In some embodiments, when the type of the communication resource request message is the PDU session establishment request message or the communication channel establishment request, there is a many-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier;

[0183] In the case where the type of the communication resource request message is the PDU session modification request message or the communication channel modification request, there is a one-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier.

[0184] In this way, the PDU session identifier can be determined according to the mapping relationship and the AI ​​session identifier, or the AI ​​session identifier can be determined according to the mapping relationship and the PDU session identifier.

[0185] In some embodiments, the AI ​​resource establishment request message may include at least one of the following: an AI policy, information of the second network element, an address of the fourth network element, and an information feedback indication of the first network element.

[0186] In some embodiments, the fourth network element may select the fifth network element according to at least one of the following:

[0187] Area of ​​interest;

[0188] Contract information for terminal AI services;

[0189] DNN;

[0190] S-NSSAI;

[0191] AI business logo;

[0192] Supported AI services;

[0193] The capabilities of the fifth network element.

[0194] It can be understood that the area of ​​interest is the area of ​​interest of the terminal; DNN is the DNN supported by the fifth network element, S-NSSAI is the S-NSSAI supported by the fifth network element, and "supported AI services" refers to the AI ​​services supported by the fifth network element.

[0195] In some embodiments, the fourth network element receiving the communication resource request message includes at least one of the following:

[0196] Get it from the terminal;

[0197] Obtained from the sixth network element;

[0198] Obtained from the fifth network element.

[0199] In some embodiments, the method further comprises:

[0200] The fourth network element acquires the communication strategy, the AI ​​strategy, the address of the application server AS and / or the information of the first network element for determining the second network element, and the acquisition method includes at least one of the following:

[0201] Determining according to the communication resource request message;

[0202] Obtained from the fifth network element;

[0203] Obtained from the sixth network element;

[0204] Obtained from the seventh network element.

[0205] In this embodiment, the fourth network element can determine the second network element based on the communication strategy and the information of the first network element; or, determine the second network element based on the communication strategy, the AI ​​strategy and the information of the first network element; or, determine the second network element based on the communication strategy, the address of the AS and the information of the first network element; or, determine the second network element based on the communication strategy, the AI ​​strategy, the address of the AS and the information of the first network element.

[0206] In some embodiments, when the communication strategy and the AI ​​strategy are obtained from the seventh network element, the fourth network element obtaining the communication strategy and the AI ​​strategy includes:

[0207] In the case of deploying a dynamic policy control and charging PCC architecture, the fourth network element sends a policy request message to the seventh network element, requesting the seventh network element to allocate a policy, and receiving the communication policy and the AI ​​policy returned by the seventh network element, where the policy request message includes at least one of the following: a terminal identifier, a PDU session identifier, an associated AI session identifier, a DNN, an S-NSSAI, and an AI service requirement description information;

[0208] When the dynamic PCC architecture is not deployed, the fourth network element determines the communication strategy and the AI ​​strategy according to at least one of the following: local configuration, AI service identifier, and policy reference identifier.

[0209] In this embodiment, when a dynamic policy control and charging PCC architecture is deployed, the seventh network element may allocate the AI ​​policy and the communication policy; when a dynamic PCC architecture is not deployed, the fourth network element itself may determine the communication policy and the AI ​​policy.

[0210] See also Figure 6 , Figure 6 is a flowchart of a method for establishing an AI service connection provided in an embodiment of the present application, which is executed by the network side, such as Figure 6 As shown, the method comprises the following steps:

[0211] Step 301: The fifth network element receives an AI resource establishment request message, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service;

[0212] Step 302: The fifth network element allocates AI resources according to the AI ​​resource establishment request message, and returns an AI resource establishment response message, where the AI ​​resource establishment response message includes at least one of the following: information of the first network element and information of the second network element.

[0213] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services, solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[0214] In some embodiments, the AI ​​resource establishment request message includes at least one of the following:

[0215] AI business requirements description information;

[0216] Data network name DNN;

[0217] Single Network Slice Selection Assistance Information S-NSSAI;

[0218] Session type indication, used to indicate whether the session is a PDU session or an AI session;

[0219] PDU session information;

[0220] Policy reference identifier;

[0221] Address of the application server AS.

[0222] In some embodiments, before the fifth network element allocates the AI ​​resource according to the AI ​​resource establishment request message, the method further includes:

[0223] The fifth network element determines the AI ​​strategy and / or communication strategy according to the AI ​​service demand description information. Based on the AI ​​service demand description information, the fifth network element can convert the service demand of the AI ​​service into an internal strategy executable on the network side, wherein the internal strategy includes a communication strategy or an AI strategy, etc., wherein the communication strategy includes the QoS rules corresponding to the existing PDU session, and the AI ​​strategy includes the AI ​​service processing delay, the AI ​​service processing accuracy, etc. For example, when the central control node is an AI service management function, it is determined on demand whether to interact with other AI network elements according to the AI ​​strategy to obtain relevant configuration information (for example, if the model is not provided, the AI ​​service management function needs to interact with the AI ​​model management function to obtain the model; if the first request message contains information indicating that the data source is the data plane, the AI ​​service management function needs to interact with the data plane to obtain the data source); the configuration information may include at least one of the following: model, model download information (such as model download address), data source information (data source address), AI service computing capacity (computing power size, type).

[0224] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network allocates the AI ​​service quality requirements to the user.

[0225] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0226] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0227] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0228] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0229] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0230] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0231] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0232] In some embodiments, the method further comprises:

[0233] The fifth network element acquires the AI ​​strategy and / or the communication strategy, and the acquisition method includes at least one of the following:

[0234] Get it based on local configuration;

[0235] Obtained from the seventh network element;

[0236] Obtained from the fourth network element;

[0237] Obtained from the sixth network element.

[0238] In some embodiments, when the information is obtained from the fourth network element, the method further includes:

[0239] The fifth network element sends a communication resource request message to the fourth network element, requesting the fourth network element to select a second network element, wherein the communication resource request message includes at least one of the following: a communication strategy, information of the first network element, an address of the fifth network element, and an information feedback indication of the first network element.

[0240] In some embodiments, the fifth network element allocating AI resources according to the AI ​​resource establishment request message includes:

[0241] The fifth network element sends an AI resource request message to the eighth network element, so that the eighth network element selects the first network element according to the AI ​​resource request message, where the AI ​​resource request message includes at least one of the following: model information, data source information, AI service computing capacity, address of application server AS, and AI policy;

[0242] The fifth network element receives an AI resource request response message returned by the eighth network element, where the AI ​​resource request response message includes information of the first network element, and the information of the first network element includes at least one of the following:

[0243] Internet protocol IP address;

[0244] Identification information;

[0245] Fully qualified domain name FQDN.

[0246] In some embodiments, before the fifth network element allocates the AI ​​resource according to the AI ​​resource establishment request message, the method further includes:

[0247] When the first condition is met, the fifth network element sends a model request message to the ninth network element, where the model request message is used to request model information;

[0248] The fifth network element receives a model request response message sent by the ninth network element, where the model request response message is used to indicate model information;

[0249] The model information includes at least one of the following: model identification, model download address, model size, model training algorithm, model precision, accuracy, model factory, model function identification, model architecture, and weight parameters;

[0250] The first condition includes at least one of the following:

[0251] The AI ​​resource establishment request message does not include model information;

[0252] The AI ​​resource establishment request message includes a first indication indicating that interaction with the ninth network element is required;

[0253] The fifth network element cannot obtain the model information.

[0254] In this embodiment, the AI ​​business management function can send a model request message to the AI ​​model management function to determine the model information.

[0255] In some embodiments, the fifth network element may select the ninth network element according to at least one of the following:

[0256] DNN;

[0257] S-NSSAI;

[0258] AI business logo;

[0259] Supported AI services;

[0260] Service area information;

[0261] Supported model information.

[0262] It can be understood that DNN is the DNN supported by the ninth network element, S-NSSAI is the S-NSSAI supported by the ninth network element, “supported AI services” refers to the AI ​​services supported by the ninth network element, “service area information” refers to the service area information of the ninth network element, and “supported model information” refers to the model information supported by the ninth network element.

[0263] In some embodiments, before the fifth network element allocates the AI ​​resource according to the AI ​​resource establishment request message, the method further includes:

[0264] When the second condition is met, the fifth network element sends a data source request message to the tenth network element, where the data source request message is used to request data source information;

[0265] The fifth network element receives a data source response message sent by the tenth network element, where the data source response message is used to indicate data source information;

[0266] The data source information includes at least one of the following: data source IP address, data source FQDN, data source identifier;

[0267] The second condition includes at least one of the following:

[0268] The AI ​​resource establishment request message does not include data source information;

[0269] The AI ​​resource establishment request message includes a second indication, indicating that interaction with the tenth network element is required;

[0270] The fifth network element cannot obtain data source information.

[0271] In this embodiment, the AI ​​service management function may send a data source request message to the data plane management function to determine the data source information.

[0272] In some embodiments, the fifth network element may select the tenth network element according to at least one of the following:

[0273] DNN;

[0274] S-NSSAI;

[0275] AI business logo;

[0276] Supported AI services;

[0277] Service area information;

[0278] Data source information.

[0279] It can be understood that DNN is the DNN supported by the tenth network element, S-NSSAI is the S-NSSAI supported by the tenth network element, “supported AI services” refers to the AI ​​services supported by the tenth network element, “service area information” refers to the service area information of the tenth network element, and “supported model information” refers to the model information supported by the tenth network element.

[0280] In some embodiments, the communication resource request message may include at least one of the following:

[0281] A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[0282] PDU session modification request message, used to modify the PDU session;

[0283] Communication channel establishment request message, used to establish a communication bearer;

[0284] The communication channel modification request message is used to modify the communication bearer.

[0285] In this embodiment, any one of the above messages or any multiple messages may be used to request the network side to allocate resources for the AI ​​service.

[0286] In some embodiments, when the type of the communication resource request message is a PDU session establishment request message or a communication channel establishment request, there is a many-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier;

[0287] In the case where the type of the communication resource request message is a PDU session modification request message or a communication channel modification request, there is a one-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier.

[0288] In this way, the PDU session identifier can be determined according to the mapping relationship and the AI ​​session identifier, or the AI ​​session identifier can be determined according to the mapping relationship and the PDU session identifier.

[0289] In some embodiments, the fifth network element receiving the AI ​​resource establishment request message includes at least one of the following:

[0290] Acquired from a fourth network element, where the AI ​​resource establishment request message is generated by the terminal or the fourth network element;

[0291] The AI ​​resource establishment request message is obtained from the sixth network element and is generated by the terminal.

[0292] In some embodiments, when the information is obtained from the sixth network element, the method further includes:

[0293] The fifth network element sends an AI resource establishment request response to the sixth network element to feed back an AI resource establishment result, wherein the AI ​​resource establishment request response includes at least one of the following: information of the first network element and an AI strategy.

[0294] In some embodiments, the information of the first network element includes at least one of the following: an IP address of the first network element, identification information of the first network element, and a FQDN of the first network element;

[0295] The information of the second network element includes at least one of the following: the second network element IP address, the second network element identification information, and the second network element FQDN.

[0296] See also Figure 7 , Figure 7 is a flowchart of a method for establishing an AI service connection provided in an embodiment of the present application, which is executed by the network side, such as Figure 7 As shown, the method comprises the following steps:

[0297] Step 401: The sixth network element receives a first request message, where the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal;

[0298] Step 402: The sixth network element selects a fifth network element and a fourth network element, sends an AI resource establishment request message to the fifth network element, and sends a communication resource request message to the fourth network element;

[0299] Step 403: the sixth network element receives an AI resource establishment response message returned by the fifth network element, the AI ​​resource establishment response message including information of the first network element allocated for the AI ​​service, and receives a communication resource establishment response message returned by the fourth network element, the communication resource establishment response message including information of the second network element allocated for the AI ​​service;

[0300] Step 404: The sixth network element returns a first request response message, where the first request response message includes information about the first network element and information about the second network element.

[0301] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services, solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[0302] It should be noted that network resources are a total resource set including communication resources, AI resources and / or policy resources, and are a general term for various resources that provide reliable guarantee links for a certain service. Communication resources are a set of resources provided by the existing 3GPP to terminals or third parties for communication services, including but not limited to communication links, allocated and scheduled frequency bands, memory of base stations or network elements, signaling forwarding and other resources. AI resources are a set of resources to ensure AI services, including but not limited to the computing power of high-performance computing machines (CPU / GPU, etc.), storage capacity for large amounts of training data, etc. Mapped to the actual link, it can be understood as a PDU session and an AI session. The former belongs to a session on the communication resource link, and the latter belongs to a session on the AI ​​resource link.

[0303] In some embodiments, the method further comprises:

[0304] The sixth network element decides to first send the communication resource request message or the AI ​​resource establishment request message according to at least one of the following:

[0305] Local configuration;

[0306] The resource determination indication is used to indicate whether the AI ​​resources or the communication resources should be determined first.

[0307] In some embodiments, the first request message includes at least one of the following:

[0308] AI business requirements description information;

[0309] Data network name DNN;

[0310] Single Network Slice Selection Assistance Information S-NSSAI;

[0311] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[0312] PDU session information;

[0313] Policy reference identifier;

[0314] Address of the application server AS.

[0315] Based on the information included in the above-mentioned first request message, the network side can convert the business requirements of the AI ​​service into internal policies executable by the network side, including communication policies and AI policies, etc., wherein the communication policy includes the QoS rules corresponding to the existing PDU session, and the AI ​​policy includes the AI ​​service processing delay, AI service processing accuracy, etc. For example, when the central control node is an AI service management function, it is determined on demand whether to interact with other AI network elements according to the AI ​​policy to obtain relevant configuration information (for example, if the model is not provided, the AI ​​service management function needs to interact with the AI ​​model management function to obtain the model; if the first request message contains an indication that the data source is the data plane, the AI ​​service management function needs to interact with the data plane to obtain the data source); the configuration information may include at least one of the following: model, model download information (such as model download address), data source information (data source address), AI service computing capacity (computing power size, type).

[0316] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network allocates the AI ​​service quality requirements to the user.

[0317] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0318] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0319] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0320] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0321] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0322] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0323] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0324] In some embodiments, the communication resource request message includes at least one of the following:

[0325] A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[0326] PDU session modification request message, used to modify the PDU session;

[0327] Communication channel establishment request message, used to establish a communication bearer;

[0328] Communication channel modification request message, used to modify the communication bearer;

[0329] The AI ​​resource request message includes at least one of the following:

[0330] An AI resource establishment request message is used to establish an AI session, where the AI ​​session is a data transmission channel between the first network element and the terminal;

[0331] AI session establishment request message;

[0332] In this embodiment, any one of the above messages or any multiple messages may be used to request the network side to allocate resources for the AI ​​service.

[0333] In some embodiments, the sixth network element selects the fifth network element according to at least one of the following:

[0334] Area of ​​interest;

[0335] Contract information for terminal AI services;

[0336] DNN;

[0337] S-NSSAI;

[0338] AI business logo;

[0339] Supported AI services;

[0340] The capabilities of the fifth network element.

[0341] In some embodiments, the fourth network element satisfies at least one of the following:

[0342] The fourth network element supports forwarding the AI ​​resource request message;

[0343] The fourth network element supports establishing a bearer for the AI ​​service;

[0344] The fourth network element supports communication with the AI ​​network function;

[0345] The fourth network element supports selection of AI network functions.

[0346] The technical solution of the present application is further introduced below in conjunction with the accompanying drawings and specific embodiments.

[0347] Embodiment 1

[0348] In this embodiment, the communication management function is used as the central control node. Figure 8 Based on the protocol stack shown in the figure, a new protocol layer is added in the UE, such as the AI ​​management layer, which is responsible for generating AI service messages, such as AI resource establishment request messages, which are used to trigger the allocation of AI resources. The message encapsulation method is: encapsulating the AI ​​resource request message in the communication resource request message, and encapsulating the communication resource request message in the NAS message. Among them, the NAS-SM layer of the terminal generates the communication resource request message, and the AI ​​management layer of the UE generates the AI ​​resource request message.

[0349] like Fig. 9 As shown, this embodiment includes the following steps:

[0350] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0351] Among them, the first request message can be a communication resource request message, specifically, it can be at least one of the following: a PDU session establishment request message, a PDU session modification request message, a communication channel establishment request message (used to establish a data transmission channel); the first request message is encapsulated in a NAS message, specifically, the encapsulation method can be at least one of the following: encapsulated in an N1 SM container, encapsulated in a newly added container (used to encapsulate session management, channel establishment and other messages).

[0352] The first request message includes at least one of the following:

[0353] AI business requirements description information;

[0354] Data network name DNN;

[0355] Single Network Slice Selection Assistance Information S-NSSAI;

[0356] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[0357] PDU session information;

[0358] Policy reference identifier, including the AI ​​service policy indication negotiated between the UE and the policy charging function;

[0359] Address of the application server AS.

[0360] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network allocates the AI ​​service quality requirements to the user.

[0361] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0362] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0363] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0364] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0365] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0366] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0367] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0368] In this embodiment, the NAS interface function has similar functions to AMF, or the NAS interface function is AMF.

[0369] Step 2: The NAS interface function selects the communication management function.

[0370] The NAS interface function can discover the communication management function instance using NRF or based on local configuration information, similar to AMF selecting SMF when establishing a PDU session. The NAS interface function requires additional considerations to select the communication management function, including at least one of the following: the communication management function supports forwarding AI-related messages, the communication management function supports establishing a bearer for AI services, and the communication management function supports communication with the AI ​​function (i.e., has an interface with the AI ​​function).

[0371] Step 3: The NAS interface function sends a first request message to the communication management function to request allocation of corresponding network resources for the AI ​​service.

[0372] The NAS interface function determines to send the first request message to the communication management function according to at least one of the following: session type indication, first request message, container type.

[0373] The first request message in this step may include at least one of the following: UE ID, PDU session ID, associated AI session ID, session type indication, DNN, S-NSSAI, PDU session information, AI service requirement description information, AI service management function instance address, and AS address.

[0374] Among them, when the first request message is the PDU session establishment request message, there is a many-to-one binding relationship between the AI ​​session ID and the PDU session ID; when the first request message is the PDU session modification message, there is a one-to-one binding relationship between the AI ​​session ID and the PDU session ID.

[0375] Step 4: The communication management function determines the communication strategy and AI strategy based on the first request message.

[0376] Implementation method 1: If a dynamic PCC architecture is deployed, the communication management function sends a policy request message to the policy charging function, and the policy charging function allocates a policy according to the policy request message. The request message includes at least one of the following: UE ID, PDU session ID, associated AI session ID, DNN, S-NSSAI, and one or more of AI service requirement description information.

[0377] Implementation method 2: If the dynamic PCC architecture is not deployed, the communication management function allocates a policy based on at least one of the following: local configuration, AI service identifier, and policy ID.

[0378] Step 5: The communication management function selects the communication resource node according to the communication strategy, and establishes or modifies the corresponding communication bearer, where the communication bearer is used to transmit AI service data.

[0379] The communication resource node information includes at least one of the following: IP address; ID information; FQDN;

[0380] Step 6: The communication management function selects the AI ​​service management function instance.

[0381] Optionally, before selecting an AI service management function instance, the communication management function may request the UE's contract data from the unified data management function (UDM) and / or the unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, proceed with the subsequent process; if not, the NAS interface function rejects the UE's AI service request.

[0382] The communication management function discovers a matching AI service management function instance using NRF or based on local configuration information according to at least one of the following:

[0383] Area of ​​interest;

[0384] UE AI task-related contract information;

[0385] DNN;

[0386] S-NSSAI;

[0387] AI business logo;

[0388] Supported AI services;

[0389] AI business management function instance capabilities, including AI business processing latency, AI business processing accuracy, AI business computing capacity and other determination capabilities.

[0390] Step 7: The communication management function sends an AI resource establishment request message to the AI ​​service management function to request the network side to allocate corresponding AI resources. The AI ​​resource establishment request message includes but is not limited to at least one of the following: AI policy, communication resource node information, the communication management function instance address, and AI resource node information feedback indication.

[0391] The AI ​​resource establishment request message is generated by a newly added protocol layer (such as an AI management layer) in the UE and is encapsulated in a first request message. The encapsulation method can be at least one of the following: encapsulated in a newly added container (used to encapsulate AI-related messages, such as AI session management, AI channel establishment, etc.).

[0392] Step 8a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0393] Step 8a is an optional step. Step 8a needs to be performed under at least one of the following conditions:

[0394] The AI ​​resource establishment request message does not include model information;

[0395] The AI ​​resource establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0396] The AI ​​business management function cannot obtain model information;

[0397] The AI ​​business management function selects the AI ​​model management function instance according to at least one of the following:

[0398] DNN;

[0399] S-NSSAI;

[0400] AI business logo;

[0401] Supported AI services;

[0402] Service area information;

[0403] Supported model information;

[0404] Step 8b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate the model information. In this way, the AI ​​business management function can determine the data source information based on the AI ​​model information. If it has been determined, step 9 is not executed.

[0405] Step 9a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0406] Step 9a is an optional step. Step 9a needs to be performed under at least one of the following conditions:

[0407] The AI ​​resource establishment request message does not include data source information;

[0408] The AI ​​resource establishment request message includes an indication that interaction with the data plane management function is required;

[0409] The AI ​​business management function cannot obtain data source information;

[0410] The AI ​​service management function selects the data plane management function instance according to at least one of the following:

[0411] DNN;

[0412] S-NSSAI;

[0413] AI business logo;

[0414] Supported AI services;

[0415] Service area information;

[0416] Data source information.

[0417] Step 9b: The data plane management function sends a model request response to the AI ​​service management function to indicate data source information.

[0418] In this embodiment, the AI ​​business management function, AI model management function and AI resource management function can be combined, that is, the AI ​​business management function, AI model management function and AI resource management function can be integrated into one. If the AI ​​business management function, AI model management function and AI resource management function are combined, steps 8-9 are omitted.

[0419] Step 10: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0420] The AI ​​resource request message includes at least one of the following:

[0421] Model information (model or model download address);

[0422] Data source information (data source address);

[0423] AI business computing volume (computing power size and type);

[0424] AS address.

[0425] Step 11: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0426] Step 12: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to the AI ​​resource node to guide the AI ​​resource node to process and forward AI service data.

[0427] The AI ​​resource configuration information includes at least one of the following:

[0428] Data source information (data source address);

[0429] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0430] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0431] AS address.

[0432] Step 13: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0433] IP address;

[0434] ID information;

[0435] FQDN.

[0436] Step 14: The AI ​​service management function sends an AI resource establishment response message to the communication management function, which includes AI resource node information.

[0437] Step 15: The communication management function sends a first request response message to the NAS interface function to notify the network side of the resource establishment result, which includes communication resource node information and AI resource node information.

[0438] Step 16: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[0439] Step 17: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0440] The AI ​​service data includes at least one of the following: UE data; model information;

[0441] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0442] Step 18: Optionally, the AI ​​resource node obtains network data from a data source.

[0443] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0444] Step 19: Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0445] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0446] Step 20: The AI ​​resource node processes the AI ​​service data according to the network model, UE data and network data and feeds back the processing result to the UE.

[0447] Embodiment 2

[0448] In this embodiment, the AI ​​service management function is used as the central control node. Fig.10 Based on the protocol stack shown in the figure, a new protocol layer is added in the UE, such as the AI ​​management layer, which is responsible for generating AI service messages, such as AI resource establishment request messages, which are used to trigger the allocation of AI resources. The newly added protocol layer is below the existing NAS-SM layer. The message encapsulation method is: encapsulate the communication resource request message in the AI ​​resource request message, and encapsulate the AI ​​resource request message in the NAS message. Among them, the UE NAS-SM layer generates the communication resource request message, and the UE AI management layer generates the AI ​​resource request message.

[0449] like Fig.11 As shown, this embodiment includes the following steps:

[0450] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0451] Among them, the first request message can be a communication resource request message, specifically, it can be at least one of the following: a PDU session establishment request message, a PDU session modification request message, and a communication channel establishment request message (used to establish a data transmission channel); the first request message is generated by a newly added protocol layer in the UE (such as an AI management layer), and the first request message is encapsulated in a NAS message. Specifically, the encapsulation method can be at least one of the following: encapsulated in a newly added container (used to encapsulate AI-related messages, such as AI session management, AI channel establishment, etc.).

[0452] The first request message includes at least one of the following:

[0453] AI business requirements description information;

[0454] Data network name DNN;

[0455] Single Network Slice Selection Assistance Information S-NSSAI;

[0456] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[0457] PDU session information;

[0458] Policy reference identifier, including the AI ​​service policy indication negotiated between the UE and the policy charging function;

[0459] Address of the application server AS.

[0460] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network allocates the AI ​​service quality requirements to the user.

[0461] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0462] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0463] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0464] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0465] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0466] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0467] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0468] Step 2: The NAS interface function selects the AI ​​service management function instance.

[0469] Optionally, before selecting an AI service management function instance, the UE's contract data may be requested from a unified data management function (UDM) and / or a unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, the subsequent process is performed; if not, the NAS interface function rejects the UE's AI service request.

[0470] The matching AI service management function instance can be discovered by using NRF or local configuration information according to at least one of the following:

[0471] Area of ​​interest;

[0472] UE AI task-related contract information;

[0473] DNN;

[0474] S-NSSAI;

[0475] AI business logo;

[0476] Supported AI services;

[0477] AI business management function instance capabilities, including AI business processing latency, AI business processing accuracy, AI business computing capacity and other determination capabilities.

[0478] Step 3: The NAS interface function sends a first request message to the AI ​​service management function.

[0479] Step 4: The AI ​​service management function determines the communication strategy and the AI ​​strategy according to the first request message.

[0480] Implementation method 1: If a dynamic PCC architecture is deployed, the AI ​​service management function sends a policy request message to the policy charging function, and the policy charging function allocates a policy according to the policy request message. The request message includes at least one of the following: UEID, PDU session ID, associated AI session ID, DNN, S-NSSAI, and one or more of the AI ​​service requirement description information.

[0481] Implementation method 2: If the dynamic PCC architecture is not deployed, the AI ​​service management function allocates a policy based on at least one of the following: local configuration, AI service identifier, and policy ID.

[0482] Step 5a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0483] Step 5a is an optional step. Step 5a needs to be performed under at least one of the following conditions:

[0484] The AI ​​resource establishment request message does not include model information;

[0485] The AI ​​resource establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0486] The AI ​​business management function cannot obtain model information;

[0487] The AI ​​business management function selects the AI ​​model management function instance according to at least one of the following:

[0488] DNN;

[0489] S-NSSAI;

[0490] AI business logo;

[0491] Supported AI services;

[0492] Service area information;

[0493] Supported model information;

[0494] Step 5b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate the model information. In this way, the AI ​​business management function can determine the data source information based on the AI ​​model information. If it has been determined, step 6 is not executed.

[0495] Step 6a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0496] Step 6a is an optional step. Step 6a needs to be performed under at least one of the following conditions:

[0497] The AI ​​resource establishment request message does not include data source information;

[0498] The AI ​​resource establishment request message includes an indication that interaction with the data plane management function is required;

[0499] The AI ​​business management function cannot obtain data source information;

[0500] The AI ​​service management function selects the data plane management function instance according to at least one of the following:

[0501] DNN;

[0502] S-NSSAI;

[0503] AI business logo;

[0504] Supported AI services;

[0505] Service area information;

[0506] Data source information.

[0507] Step 6b: The data plane management function sends a model request response to the AI ​​service management function to indicate data source information.

[0508] In this embodiment, the AI ​​business management function, AI model management function and AI resource management function can be combined, that is, the AI ​​business management function, AI model management function and AI resource management function can be integrated into one. If the AI ​​business management function, AI model management function and AI resource management function are combined, steps 5-6 are omitted.

[0509] Step 7: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0510] The AI ​​resource request message includes at least one of the following:

[0511] Model information (model or model download address);

[0512] Data source information (data source address);

[0513] AI business computing volume (computing power size and type);

[0514] AS address.

[0515] Step 8: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0516] Step 9: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to the AI ​​resource node to guide the AI ​​resource node to process and forward AI service data.

[0517] The AI ​​resource configuration information includes at least one of the following:

[0518] Data source information (data source address);

[0519] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0520] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0521] AS address.

[0522] Step 10: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0523] IP address;

[0524] ID information;

[0525] FQDN.

[0526] Step 11. The AI ​​service management function selects the communication management function instance. The AI ​​service management function can use NRF or discover the communication management function instance according to local configuration information. The AI ​​service management function selects the communication management function according to at least one of the following: the communication management function supports forwarding AI-related messages, the communication management function supports establishing a bearer for the AI ​​service, and the communication management function supports communication with the AI ​​function (i.e., has an interface with the AI ​​function).

[0527] Step 11 is an optional step and needs to be executed if the AI ​​service management function does not receive the address of the communication management function instance.

[0528] Step 12: The AI ​​service management function sends a communication resource request message to the communication management function, requesting the communication management function to select a suitable communication resource node, which includes at least one of the following: communication strategy, AI resource node information, the AI ​​service management function instance address and AI resource node information feedback indication.

[0529] Specifically, the communication resource request message may be at least one of the following: a PDU session establishment request message, a PDU session modification message, a communication channel establishment request message (for establishing a data transmission channel);

[0530] When the communication resource request message is the PDU session establishment request message, there is a many-to-one binding relationship between the AI ​​session ID and the PDU session ID; when the communication resource request message is the PDU session modification message, there is a one-to-one binding relationship between the AI ​​session ID and the PDU session ID.

[0531] Step 13: The communication management function instance selects a suitable communication resource node according to the communication strategy and AI resource node information.

[0532] Step 14: The communication management function establishes or modifies a corresponding communication bearer according to a communication strategy, where the communication bearer is used to transmit AI service data.

[0533] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[0534] Step 15a: The communication management function sends a communication resource establishment response message to the AI ​​service management function, which includes communication resource node information.

[0535] Step 15a is an optional step, and is executed if the communication management function receives the AI ​​service management function instance address and / or AI resource node information feedback indication. Otherwise, the communication management function executes step 16.

[0536] Step 15b: The AI ​​service management function sends a first request response message to the NAS interface function, which is used to notify the network resource establishment result and includes AI resource node information and communication resource node information.

[0537] Step 16: The communication management function sends a network resource establishment response message to the NAS interface function to notify the network resource establishment result, which includes AI resource node information and communication resource node information.

[0538] Step 16 is an optional step.

[0539] Step 17: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[0540] The first request response message is encapsulated in a container in the NAS message, for example, in an N1 AI management container.

[0541] Step 18: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0542] The AI ​​service data includes at least one of the following: UE data; model information;

[0543] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0544] Step 19: Optionally, the AI ​​resource node obtains network data from a data source.

[0545] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0546] Step 20: Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0547] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0548] Step 21: The AI ​​resource node processes the AI ​​service data according to the network model, UE data and network data and feeds back the processing result to the UE.

[0549] Embodiment 3

[0550] In this embodiment, the AI ​​service management function is used as the central control node. Fig.12 Based on the protocol stack shown in FIG. 1 , a new protocol layer is added in the UE, such as the AI ​​management layer, which is responsible for generating AI service messages, such as AI resource establishment request messages, for triggering the allocation of AI resources. The newly added protocol layer is below the existing NAS-SM layer. The message encapsulation method is: encapsulating the communication resource request message in the AI ​​resource request message, and encapsulating the AI ​​resource request message in the NAS message. Among them, the AI ​​service management function generates the communication resource request message, and the UE AI management layer generates the AI ​​resource request message.

[0551] like Fig.13 As shown, this embodiment includes the following steps:

[0552] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0553] Among them, the first request message can be a communication resource request message, specifically, it can be at least one of the following: a PDU session establishment request message, a PDU session modification request message, and a communication channel establishment request message (used to establish a data transmission channel); the first request message is generated by a newly added protocol layer (such as an AI management layer) in the UE, and the first request message is encapsulated in a NAS message. Specifically, the encapsulation method can be at least one of the following: a newly added container (such as an N1 AIM container) is used to encapsulate AI-related messages, such as AI session management, AI channel establishment and other messages.

[0554] The first request message includes at least one of the following:

[0555] AI business requirements description information;

[0556] Data network name DNN;

[0557] Single Network Slice Selection Assistance Information S-NSSAI;

[0558] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[0559] PDU session information;

[0560] Policy reference identifier, including the AI ​​service policy indication negotiated between the UE and the policy charging function;

[0561] Address of the application server AS.

[0562] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network allocates the AI ​​service quality requirements to the user.

[0563] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0564] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0565] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0566] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0567] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0568] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0569] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0570] Step 2: The NAS interface function selects the AI ​​service management function instance.

[0571] Optionally, before selecting an AI service management function instance, the UE's contract data may be requested from a unified data management function (UDM) and / or a unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, the subsequent process is performed; if not, the NAS interface function rejects the UE's AI service request.

[0572] The NAS interface function discovers a matching AI service management function instance using NRF or based on local configuration information according to at least one of the following:

[0573] Area of ​​interest;

[0574] UE AI task-related contract information;

[0575] DNN;

[0576] S-NSSAI;

[0577] AI business logo;

[0578] Supported AI services;

[0579] AI business management function instance capabilities, including AI business processing latency, AI business processing accuracy, AI business computing capacity and other determination capabilities.

[0580] Step 3: The NAS interface function sends a first request message to the AI ​​service management function.

[0581] Step 4: The AI ​​service management function determines the communication strategy and the AI ​​strategy according to the first request message.

[0582] Implementation method 1: If a dynamic PCC architecture is deployed, the AI ​​service management function sends a policy request message to the policy charging function, and the policy charging function allocates a policy according to the policy request message. The request message includes at least one of the following: UEID, PDU session ID, associated AI session ID, DNN, S-NSSAI, and one or more of the AI ​​service requirement description information.

[0583] Implementation method 2: If the dynamic PCC architecture is not deployed, the AI ​​service management function allocates a policy based on at least one of the following: local configuration, AI service identifier, and policy ID.

[0584] Step 5a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0585] Step 5a is an optional step. Step 5a needs to be performed under at least one of the following conditions:

[0586] The AI ​​resource establishment request message does not include model information;

[0587] The AI ​​resource establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0588] The AI ​​business management function cannot obtain model information;

[0589] The AI ​​business management function selects the AI ​​model management function instance according to at least one of the following:

[0590] DNN;

[0591] S-NSSAI;

[0592] AI business logo;

[0593] Supported AI services;

[0594] Service area information;

[0595] Supported model information;

[0596] Step 5b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate the model information. In this way, the AI ​​business management function can determine the data source information based on the AI ​​model information. If it has been determined, step 6 is not executed.

[0597] Step 6a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0598] Step 6a is an optional step. Step 6a needs to be performed under at least one of the following conditions:

[0599] The AI ​​resource establishment request message does not include data source information;

[0600] The AI ​​resource establishment request message includes an indication that interaction with the data plane management function is required;

[0601] The AI ​​business management function cannot obtain data source information;

[0602] The AI ​​service management function selects the data plane management function instance according to at least one of the following:

[0603] DNN;

[0604] S-NSSAI;

[0605] AI business logo;

[0606] Supported AI services;

[0607] Service area information;

[0608] Data source information.

[0609] Step 6b: The data plane management function sends a model request response to the AI ​​service management function to indicate data source information.

[0610] In this embodiment, the AI ​​business management function, AI model management function and AI resource management function can be combined, that is, the AI ​​business management function, AI model management function and AI resource management function can be integrated into one. If the AI ​​business management function, AI model management function and AI resource management function are combined, steps 5-6 are omitted.

[0611] Step 7: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0612] The AI ​​resource request message includes at least one of the following:

[0613] Model information (model or model download address);

[0614] Data source information (data source address);

[0615] AI business computing volume (computing power size and type);

[0616] AS address.

[0617] Step 8: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0618] Step 9: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to the AI ​​resource node to guide the AI ​​resource node to process and forward AI service data.

[0619] The AI ​​resource configuration information includes at least one of the following:

[0620] Data source information (data source address);

[0621] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0622] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0623] AS address.

[0624] Step 10: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0625] IP address;

[0626] ID information;

[0627] FQDN.

[0628] Step 11. The AI ​​service management function selects the communication management function instance. The AI ​​service management function can use NRF or discover the communication management function instance according to local configuration information. The AI ​​service management function selects the communication management function according to at least one of the following: the communication management function supports forwarding AI-related messages, the communication management function supports establishing a bearer for the AI ​​service, and the communication management function supports communication with the AI ​​function (i.e., has an interface with the AI ​​function).

[0629] Step 11 is an optional step and needs to be executed if the AI ​​service management function does not receive the address of the communication management function instance.

[0630] Step 12: The AI ​​service management function generates and sends a communication resource request message to the communication management function, requesting the communication management function to select a suitable communication resource node, which includes at least one of the following: communication strategy, AI resource node information, the AI ​​service management function instance address, AI resource node information feedback indication, and PDU session information.

[0631] The communication resource request message triggers the communication management function to send a PDU session modification message and a communication channel modification message (used to modify a data transmission channel);

[0632] At this time, there is a one-to-one binding relationship between the AI ​​session ID and the PDU session ID.

[0633] The AI ​​service management function determines to generate and send a communication resource request message according to at least one of the following items in the first request message: session type indication, AI description information, and PDU session information.

[0634] Step 13: The communication management function instance selects a suitable communication resource node according to the communication strategy and AI resource node information.

[0635] Step 14: The communication management function establishes or modifies a corresponding communication bearer according to a communication strategy, where the communication bearer is used to transmit AI service data.

[0636] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[0637] Step 15a: The communication management function sends a communication resource establishment response message to the AI ​​service management function, which includes communication resource node information.

[0638] Step 15a is an optional step, and is executed if the communication management function receives the AI ​​service management function instance address and / or AI resource node information feedback indication. Otherwise, the communication management function executes step 16.

[0639] Step 15b: The AI ​​service management function sends a first request response message to the NAS interface function, which is used to notify the network resource establishment result and includes AI resource node information and communication resource node information.

[0640] Step 16: The communication management function sends a network resource establishment response message to the NAS interface function to notify the network resource establishment result, which includes AI resource node information and communication resource node information.

[0641] Step 16 is an optional step.

[0642] Step 17: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[0643] The first request response message is encapsulated in a container in the NAS message, for example, in an N1 AI management container.

[0644] Step 18: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0645] The AI ​​service data includes at least one of the following: UE data; model information;

[0646] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0647] Step 19: Optionally, the AI ​​resource node obtains network data from a data source.

[0648] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0649] Step 20: Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0650] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0651] Step 21: The AI ​​resource node processes the AI ​​service data according to the network model, UE data and network data and feeds back the processing result to the UE.

[0652] Embodiment 4

[0653] In this embodiment, the AI ​​service management function is used as the central control node. Fig.12 Based on the protocol stack shown in FIG. 1 , a new protocol layer is added in the UE, such as the AI ​​management layer, which is responsible for generating AI service messages, such as AI resource establishment request messages, for triggering the allocation of AI resources. The newly added protocol layer is below the existing NAS-SM layer. The message encapsulation method is: encapsulating the communication resource request message in the AI ​​resource request message, and encapsulating the AI ​​resource request message in the NAS message. Among them, the AI ​​service management function generates the communication resource request message, and the UE AI management layer generates the AI ​​resource request message.

[0654] like Fig.14 As shown, this embodiment includes the following steps:

[0655] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0656] Among them, the first request message can be an AI resource request message, specifically, it can be at least one of the following: an AI resource establishment request message (used to establish an AI session, the AI ​​session is a data transmission channel between the first network element and the terminal), an AI session establishment request message; the first request message is generated by a newly added protocol layer (such as an AI management layer) in the UE, and the first request message is encapsulated in a NAS message. Specifically, the encapsulation method can be at least one of the following: encapsulated in a newly added container (used to encapsulate AI-related messages, such as AI session management, AI channel establishment, etc.).

[0657] The first request message includes at least one of the following:

[0658] AI business requirements description information;

[0659] Data network name DNN;

[0660] Single Network Slice Selection Assistance Information S-NSSAI;

[0661] Policy reference identifier, including the AI ​​service policy indication negotiated between the UE and the policy charging function;

[0662] Address of the application server AS.

[0663] In this embodiment, the AI ​​service requirement description information is used to indicate the expected AI service quality requirements. Specifically, the user can sign up for an AI service with the operator network, and the operator network allocates the AI ​​service quality requirements to the user.

[0664] In some embodiments, the AI ​​business requirement description information may include at least one of the following:

[0665] AI service identifier, used to identify the requested AI service. The network side determines the specific AI service requested by the UE based on the service identifier, such as environment reconstruction service, facial recognition service, etc.

[0666] AI service type, used to indicate the type of AI service requested, based on which the network determines whether the UE requests a training service or an inference service;

[0667] AI service quality of experience (QoE) requirements are used to indicate the service quality that the requested AI service terminal is expected to achieve, including the response speed, reliability, and satisfaction of the AI ​​service;

[0668] AI service quality of service (QoS) information, which is used to indicate the network indicators required for the requested AI service, including priority, latency, packet loss rate, packet error rate, peak rate, etc.

[0669] Model parameters, used to indicate the model parameters required for the requested AI service, including model size, model training algorithm, model precision / accuracy, model manufacturer, model identifier, model function identifier, model architecture, and weight parameters;

[0670] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[0671] Step 2: The NAS interface function selects the AI ​​service management function instance and the communication management function instance.

[0672] Optionally, before selecting an AI service management function instance, the NAS interface function may request the UE's contract data from the unified data management function (UDM) and / or the unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, proceed with the subsequent process; if not, the NAS interface function rejects the UE's AI service request.

[0673] The NAS interface function discovers a matching AI service management function instance using NRF or based on local configuration information according to at least one of the following:

[0674] Area of ​​interest;

[0675] UE AI task-related contract information;

[0676] DNN;

[0677] S-NSSAI;

[0678] AI business logo;

[0679] Supported AI services;

[0680] AI business management function instance capabilities, including AI business processing latency, AI business processing accuracy, AI business computing capacity and other determination capabilities.

[0681] Step 3: The NAS interface function sends an AI resource establishment request message to the AI ​​service management function.

[0682] Step 4: The AI ​​service management function establishes a request message based on the AI ​​resource and determines the communication strategy and AI strategy.

[0683] Implementation method 1: If a dynamic PCC architecture is deployed, the AI ​​service management function sends a policy request message to the policy charging function, and the policy charging function allocates a policy according to the policy request message. The request message includes at least one of the following: UEID, PDU session ID, associated AI session ID, DNN, S-NSSAI, and one or more of the AI ​​service requirement description information.

[0684] Implementation method 2: If the dynamic PCC architecture is not deployed, the AI ​​service management function allocates a policy based on at least one of the following: local configuration, AI service identifier, and policy ID.

[0685] Step 5a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0686] Step 5a is an optional step. Step 5a needs to be performed under at least one of the following conditions:

[0687] The AI ​​resource establishment request message does not include model information;

[0688] The AI ​​resource establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0689] The AI ​​business management function cannot obtain model information;

[0690] The AI ​​business management function selects the AI ​​model management function instance according to at least one of the following:

[0691] DNN;

[0692] S-NSSAI;

[0693] AI business logo;

[0694] Supported AI services;

[0695] Service area information;

[0696] Supported model information;

[0697] Step 5b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate the model information. In this way, the AI ​​business management function can determine the data source information based on the AI ​​model information. If it has been determined, step 6 is not executed.

[0698] Step 6a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0699] Step 6a is an optional step. Step 6a needs to be performed under at least one of the following conditions:

[0700] The AI ​​resource establishment request message does not include data source information;

[0701] The AI ​​resource establishment request message includes an indication that interaction with the data plane management function is required;

[0702] The AI ​​business management function cannot obtain data source information;

[0703] The AI ​​service management function selects the data plane management function instance according to at least one of the following:

[0704] DNN;

[0705] S-NSSAI;

[0706] AI business logo;

[0707] Supported AI services;

[0708] Service area information;

[0709] Data source information.

[0710] Step 6b: The data plane management function sends a model request response to the AI ​​service management function to indicate data source information.

[0711] In this embodiment, the AI ​​business management function, AI model management function and AI resource management function can be combined, that is, the AI ​​business management function, AI model management function and AI resource management function can be integrated into one. If the AI ​​business management function, AI model management function and AI resource management function are combined, steps 5-6 are omitted.

[0712] Step 7: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0713] The AI ​​resource request message includes at least one of the following:

[0714] Model information (model or model download address);

[0715] Data source information (data source address);

[0716] AI business computing volume (computing power size and type);

[0717] AS address.

[0718] Step 8: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0719] Step 9: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to the AI ​​resource node to guide the AI ​​resource node to process and forward AI service data.

[0720] The AI ​​resource configuration information includes at least one of the following:

[0721] Data source information (data source address);

[0722] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0723] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0724] AS address.

[0725] Step 10: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0726] IP address;

[0727] ID information;

[0728] FQDN.

[0729] Step 11: The AI ​​service management function sends an AI resource establishment request response message to the NAS interface function to feedback the AI ​​resource establishment result, which includes at least one of the following: AI resource node information and AI policy.

[0730] Step 12: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE.

[0731] Step 13: The UE determines that communication resources need to be established, and sends a communication resource establishment request message to the communication management function to trigger the network to allocate communication resources for the AI ​​service. The communication resource establishment request message includes the address of the AI ​​resource node.

[0732] The UE determines to send the request according to at least one of the following items: received AI resource node information and AI service requirement description information.

[0733] Step 14: The communication management function instance selects a suitable communication resource node according to the communication strategy and AI resource node information.

[0734] Step 15: The communication management function establishes or modifies a corresponding communication bearer according to a communication strategy, where the communication bearer is used to transmit AI service data.

[0735] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[0736] Step 16: The communication management function sends a communication resource establishment request response to notify the communication resource establishment result.

[0737] Step 17: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0738] The AI ​​service data includes at least one of the following: UE data; model information;

[0739] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0740] Step 18: Optionally, the AI ​​resource node obtains network data from a data source.

[0741] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0742] Step 19: Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0743] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0744] Step 20: The AI ​​resource node processes the AI ​​service data according to the network model, UE data and network data and feeds back the processing result to the UE.

[0745] Embodiment 5

[0746] In this embodiment, the NAS interface function is used as the central control node. Fig.15 Based on the protocol stack shown in FIG. 1 , a new protocol layer is added in the UE, such as the AI ​​management layer. The AI ​​management layer is at the same level as the NAS-SM layer. The message encapsulation method is: encapsulate the communication resource request message and the AI ​​resource request message in the NAS message. Among them, the NAS interface function forwards the communication resource request message and the AI ​​resource request message in parallel.

[0747] like Fig.16As shown, this embodiment includes the following steps:

[0748] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0749] The first request message may be at least one of the following: a registration message, a service request message, an uplink NAS transmission message, and a newly added control message (used to establish a control connection).

[0750] Step 2: The NAS interface function selects the communication management function and the AI ​​service management function according to the first request message sent by the UE.

[0751] The NAS interface function can discover the communication management function instance using NRF or based on local configuration information, similar to AMF selecting SMF when establishing a PDU session. The NAS interface function requires additional considerations to select the communication management function, including at least one of the following: the communication management function supports forwarding AI-related messages, the communication management function supports establishing a bearer for AI services, and the communication management function supports communication with the AI ​​function (i.e., has an interface with the AI ​​function).

[0752] Optionally, before selecting an AI service management function instance, the UE's contract data may be requested from a unified data management function (UDM) and / or a unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, the subsequent process is performed; if not, the NAS interface function rejects the UE's AI service request.

[0753] Step 3a: The NAS interface function sends a communication resource request message to the communication management function to request allocation of corresponding communication resources for the AI ​​service.

[0754] The communication resource request message may be at least one of the following: a PDU session establishment request message, a PDU session modification message, and a communication channel establishment request message (for establishing a data transmission channel).

[0755] Step 3b: The NAS interface function sends an AI resource establishment request message to the AI ​​service management function to request allocation of corresponding AI resources for the AI ​​service. The AI ​​resource establishment request message includes but is not limited to at least one of the following: AI policy, communication resource node information, the communication management function instance address, and AI resource node information feedback indication.

[0756] If the network deploys a dynamic policy architecture, this embodiment includes the following steps 4-6:

[0757] Step 4a: The communication management function sends a communication policy request message to the policy charging function to request a dynamic communication policy applicable to the PDUsession. The communication policy request message includes at least one of the following: UE ID, PDU session ID, DNN, S-NSSAI, associated AI session ID, AI service requirement description information, etc.

[0758] Step 4b: The AI ​​service management function sends an AI policy request message to the policy charging function to request a dynamic AI policy applicable to the AI ​​session. The AI ​​policy request message includes at least one of the following: UE ID, AI session ID, DNN, S-NSSAI, associated PDU session ID, service requirement description information, etc.

[0759] Step 5: The policy charging function generates the communication policy and AI policy corresponding to the AI ​​service according to steps 4a and 4b. The policy charging function shall comprehensively consider the end-to-end service requirements of the AI ​​service when allocating the communication policy and AI policy.

[0760] Step 6a: The policy charging function sends a communication policy request response message to the communication management function, which includes but is not limited to at least one of the following: communication policy and AS address.

[0761] Step 6b: The policy charging function sends an AI policy request response message to the AI ​​service management function, which includes but is not limited to at least one of the following: AI policy and AS address.

[0762] If the network does not deploy a dynamic PCC architecture, steps 4-6 are omitted, the following steps are taken, and then step 7 is executed.

[0763] The communication management function determines the communication strategy according to the local configuration, and the AI ​​service management function determines the AI ​​strategy according to the local configuration.

[0764] Step 7a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0765] Step 7a is an optional step and needs to be performed under at least one of the following conditions:

[0766] The AI ​​session establishment request message does not include model information;

[0767] The AI ​​session establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0768] The AI ​​business management function cannot obtain model information;

[0769] The AI ​​business management function selects the AI ​​model management function instance by considering at least one of the following factors:

[0770] DNN;

[0771] S-NSSAI;

[0772] AI business logo;

[0773] Supported AI services;

[0774] Service area information;

[0775] Supported model information;

[0776] Step 7b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate model information.

[0777] The AI ​​business management function can determine the data source information based on the AI ​​model information. If the data source information has been determined, step 8 is not executed.

[0778] Step 8a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0779] Step 8a is an optional step and needs to be performed under at least one of the following conditions:

[0780] The AI ​​session establishment request message does not include data source information;

[0781] The AI ​​session establishment request message includes an indication that interaction with the data plane management function is required;

[0782] The AI ​​business management function cannot obtain data source information;

[0783] The AI ​​service management function selects the data plane management function instance based on at least one of the following factors:

[0784] DNN;

[0785] S-NSSAI;

[0786] AI business logo;

[0787] Supported AI services;

[0788] Service area information;

[0789] Data source information;

[0790] Step 8b: The data plane management function sends a model request response message to the AI ​​service management function to indicate data source information.

[0791] In this embodiment, the AI ​​business management function, AI model management function and AI resource management function can be integrated. If integrated, steps 7-8 are omitted.

[0792] In this embodiment, the AI ​​service management function may first determine the AI ​​resource node and feed it back to the communication management function, and the communication management function then selects the communication resource node. The following steps are included:

[0793] Step 9: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0794] The AI ​​resource request message includes at least one of the following:

[0795] Model information (model or model download address);

[0796] Data source information (data source address);

[0797] AI business computing volume (computing power size and type);

[0798] AS address.

[0799] Step 10: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0800] Step 11: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to guide the AI ​​resource node to process and forward AI service data.

[0801] The AI ​​resource configuration information includes at least one of the following:

[0802] Data source information (data source address);

[0803] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0804] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0805] AS address.

[0806] Step 12: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0807] IP address;

[0808] ID information;

[0809] FQDN.

[0810] Step 13: The AI ​​service management function sends a resource selection information notification to the communication management function, which is used for the communication management function to comprehensively consider the impact of communication and AI processing when selecting a communication resource node to ensure the end-to-end QoS of the AI ​​service, including at least one of the following:

[0811] AI resource node information;

[0812] AI strategy.

[0813] Before the AI ​​service management function sends a resource selection information notification to the communication management function, the communication management function needs to subscribe to AI resource creation or modification notifications to learn about changes in AI resources.

[0814] Step 14: The communication management function selects a suitable communication resource node according to the communication strategy, AI resource node information and AI strategy.

[0815] The communication resource node information includes at least one of the following:

[0816] IP address;

[0817] ID information;

[0818] FQDN.

[0819] Step 15: The communication management function establishes or modifies the corresponding communication bearer according to the communication strategy, and establishes or modifies the process with the existing N4session / N3 tunnel. The communication bearer is used to transmit AI service data.

[0820] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[0821] Step 16a: The communication management function sends a communication resource request response message to the NAS interface function to notify that the communication resource allocation is successful, which may include the communication resource node information.

[0822] Step 16b: The AI ​​service management function sends an AI resource establishment response message to the NAS interface function to notify that the AI ​​resource allocation is successful, which may include the AI ​​resource node information.

[0823] Step 17: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes but is not limited to at least one of the following: communication resource node and AI resource node information.

[0824] Step 18: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[0825] The first request response message is encapsulated in a container in the NAS message, for example, in an N1 AI management container.

[0826] Step 19: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0827] The AI ​​service data includes at least one of the following: UE data; model information;

[0828] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0829] Step 20: Optionally, the AI ​​resource node obtains in-network data from a data source.

[0830] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0831] Step 21. Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0832] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0833] In this embodiment, the communication management function may first determine the communication resource node and feed it back to the AI ​​service management function, which then selects the AI ​​resource node.

[0834] In this embodiment, after step 6, the following steps are included:

[0835] Step 7: The communication management function selects a suitable communication resource node according to the communication strategy, AI resource node information and AI strategy.

[0836] The communication resource node information includes at least one of the following:

[0837] IP address;

[0838] ID information;

[0839] FQDN.

[0840] Step 8: The communication management function establishes or modifies the corresponding communication bearer according to the communication strategy, and establishes or modifies the process with the existing N4session / N3 tunnel. The communication bearer is used to transmit AI service data.

[0841] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[0842] Step 9: The communication management function sends a resource selection information notification to the AI ​​service management function, which is used for the AI ​​service management function to comprehensively consider the communication and AI processing impact when selecting AI resource nodes to ensure the end-to-end QoS of the AI ​​service, including at least one of the following:

[0843] Communication resource node information;

[0844] Communication strategy.

[0845] Before the communication management function sends a resource selection information notification to the AI ​​service management function, the AI ​​service management function needs to subscribe to a communication resource establishment or modification notification to be informed of changes in communication resources.

[0846] Step 10a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0847] Step 10a is an optional step and needs to be performed under at least one of the following conditions:

[0848] The AI ​​session establishment request message does not include model information;

[0849] The AI ​​session establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0850] The AI ​​business management function cannot obtain model information;

[0851] The AI ​​business management function selects the AI ​​model management function instance by considering at least one of the following factors:

[0852] DNN;

[0853] S-NSSAI;

[0854] AI business logo;

[0855] Supported AI services;

[0856] Service area information;

[0857] Supported model information;

[0858] Step 10b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate model information.

[0859] The AI ​​business management function can determine the data source information based on the AI ​​model information. If the data source information has been determined, step 11 is not executed.

[0860] Step 11a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0861] Step 11a is an optional step and needs to be performed under at least one of the following conditions:

[0862] The AI ​​session establishment request message does not include data source information;

[0863] The AI ​​session establishment request message includes an indication that interaction with the data plane management function is required;

[0864] The AI ​​business management function cannot obtain data source information;

[0865] The AI ​​service management function selects the data plane management function instance based on at least one of the following factors:

[0866] DNN;

[0867] S-NSSAI;

[0868] AI business logo;

[0869] Supported AI services;

[0870] Service area information;

[0871] Data source information;

[0872] Step 11b: The data plane management function sends a model request response message to the AI ​​service management function to indicate data source information.

[0873] In this embodiment, the AI ​​business management function, AI model management function and AI resource management function can be integrated. If integrated, steps 7-8 are omitted.

[0874] Step 12: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0875] The AI ​​resource request message includes at least one of the following:

[0876] Model information (model or model download address);

[0877] Data source information (data source address);

[0878] AI business computing volume (computing power size and type);

[0879] AS address.

[0880] Step 13: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0881] Step 14: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to guide the AI ​​resource node to process and forward AI service data.

[0882] The AI ​​resource configuration information includes at least one of the following:

[0883] Data source information (data source address);

[0884] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0885] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0886] AS address.

[0887] Step 15: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0888] IP address;

[0889] ID information;

[0890] FQDN.

[0891] Step 16a: The communication management function sends a communication resource request response message to the NAS interface function to notify that the communication resource allocation is successful, which may include the communication resource node information.

[0892] Step 16b: The AI ​​service management function sends an AI resource establishment response message to the NAS interface function to notify that the AI ​​resource allocation is successful, which may include the AI ​​resource node information.

[0893] Step 17: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes but is not limited to at least one of the following: communication resource node and AI resource node information.

[0894] Step 18: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[0895] The first request response message is encapsulated in a container in the NAS message, for example, in an N1 AI management container.

[0896] Step 19: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0897] The AI ​​service data includes at least one of the following: UE data; model information;

[0898] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0899] Step 20: Optionally, the AI ​​resource node obtains in-network data from a data source.

[0900] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0901] Step 21. Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0902] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0903] Embodiment 6

[0904] In this embodiment, the NAS interface function is used as the central control node. Fig.17 Based on the protocol stack shown in FIG. 1 , a new protocol layer is added in the UE, such as the AI ​​management layer. The AI ​​management layer is at the same level as the NAS-SM layer. Among them, the NAS interface function determines the order in which the NAS interface function forwards the communication resource request message and the AI ​​resource request message in parallel.

[0905] like Fig.18 As shown, this embodiment includes the following steps:

[0906] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0907] The first request message may be at least one of the following: a registration message, a service request message, an uplink NAS transmission message, and a newly added control message (used to establish a control connection).

[0908] Step 2: The NAS interface function selects the communication management function and the AI ​​service management function according to the first request message sent by the UE.

[0909] The NAS interface function can discover the communication management function instance using NRF or based on local configuration information, similar to AMF selecting SMF when establishing a PDU session. The NAS interface function requires additional considerations to select the communication management function, including at least one of the following: the communication management function supports forwarding AI-related messages, the communication management function supports establishing a bearer for AI services, and the communication management function supports communication with the AI ​​function (i.e., has an interface with the AI ​​function).

[0910] Optionally, before selecting an AI service management function instance, the UE's contract data may be requested from a unified data management function (UDM) and / or a unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, the subsequent process is performed; if not, the NAS interface function rejects the UE's AI service request.

[0911] Step 3: The NAS interface function sends an AI resource establishment request message to the AI ​​service management function to request allocation of corresponding AI resources for the AI ​​service. The AI ​​resource establishment request message includes but is not limited to at least one of the following: AI policy, communication resource node information, the communication management function instance address, and AI resource node information feedback indication.

[0912] In this embodiment, the NAS interface function decides to send the AI ​​resource establishment request message first, based on at least one of the following: local configuration, resource determination indication (used to indicate whether to determine AI resources or communication resources first).

[0913] Step 4: The AI ​​service management function sends an AI policy request message to the policy charging function to request a dynamic AI policy applicable to the AI ​​session. The AI ​​policy request message includes at least one of the following: UE ID, AI session ID, DNN, S-NSSAI, associated PDU session ID, AI service requirement description information, etc.

[0914] The policy charging function generates an AI policy corresponding to the AI ​​service. The policy charging function shall comprehensively consider the end-to-end service requirements of the AI ​​service when allocating the AI ​​policy. The policy charging function sends an AI policy request response message to the AI ​​service management function, which includes but is not limited to at least one of the following: AI policy and AS address.

[0915] Step 5a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[0916] Step 5a is an optional step and needs to be performed under at least one of the following conditions:

[0917] The AI ​​session establishment request message does not include model information;

[0918] The AI ​​session establishment request message includes an indication that interaction with the AI ​​model management function is required;

[0919] The AI ​​business management function cannot obtain model information;

[0920] The AI ​​business management function selects the AI ​​model management function instance by considering at least one of the following factors:

[0921] DNN;

[0922] S-NSSAI;

[0923] AI business logo;

[0924] Supported AI services;

[0925] Service area information;

[0926] Supported model information;

[0927] Step 5b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate model information.

[0928] The AI ​​business management function can determine the data source information based on the AI ​​model information. If the data source information has been determined, step 6 is not executed.

[0929] Step 6a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[0930] Step 6a is an optional step and needs to be performed under at least one of the following conditions:

[0931] The AI ​​session establishment request message does not include data source information;

[0932] The AI ​​session establishment request message includes an indication that interaction with the data plane management function is required;

[0933] The AI ​​business management function cannot obtain data source information;

[0934] The AI ​​service management function selects the data plane management function instance based on at least one of the following factors:

[0935] DNN;

[0936] S-NSSAI;

[0937] AI business logo;

[0938] Supported AI services;

[0939] Service area information;

[0940] Data source information;

[0941] Step 6b: The data plane management function sends a model request response message to the AI ​​service management function to indicate data source information.

[0942] Step 7: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[0943] The AI ​​resource request message includes at least one of the following:

[0944] Model information (model or model download address);

[0945] Data source information (data source address);

[0946] AI business computing volume (computing power size and type);

[0947] AS address.

[0948] Step 8: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[0949] Step 9: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to guide the AI ​​resource node to process and forward AI service data.

[0950] The AI ​​resource configuration information includes at least one of the following:

[0951] Data source information (data source address);

[0952] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[0953] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[0954] AS address.

[0955] Step 10: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, wherein the message includes AI resource node information, and the AI ​​resource node information includes at least one of the following:

[0956] IP address;

[0957] ID information;

[0958] FQDN.

[0959] Step 11: The AI ​​service management function sends an AI resource establishment response message to the NAS interface function to notify that the AI ​​resource allocation is successful, which may include the AI ​​resource node information.

[0960] Step 12: The NAS interface function sends a communication resource request message to the communication management function to request allocation of corresponding communication resources for the AI ​​service.

[0961] The communication resource request message may be at least one of the following: a PDU session establishment request message, a PDU session modification message, and a communication channel establishment request message (for establishing a data transmission channel).

[0962] Step 13: The communication management function sends a communication policy request message to the policy charging function to request a dynamic communication policy applicable to the PDUsession. The communication policy request message includes at least one of the following: UE ID, PDU session ID, DNN, S-NSSAI, associated AI session ID, AI service requirement description information, etc.

[0963] The policy charging function generates the communication policy corresponding to the AI ​​service. When allocating the communication policy, the policy charging function should comprehensively consider the end-to-end service requirements of the AI ​​service.

[0964] The policy charging function sends a communication policy request response message to the communication management function, which includes but is not limited to at least one of the following: a communication policy and an AS address.

[0965] The communication management function selects appropriate communication resource nodes according to the communication strategy, AI resource node information and AI strategy.

[0966] The communication resource node information includes at least one of the following:

[0967] IP address;

[0968] ID information;

[0969] FQDN.

[0970] Step 14: The communication management function establishes or modifies the corresponding communication bearer according to the communication strategy, and establishes or modifies the process with the existing N4session / N3 tunnel. The communication bearer is used to transmit AI service data.

[0971] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[0972] Step 15: The communication management function sends a communication resource request response message to the NAS interface function to notify that the communication resource allocation is successful, which may include the communication resource node information.

[0973] Step 16: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes but is not limited to at least one of the following: communication resource node and AI resource node information.

[0974] Step 17: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[0975] The first request response message is encapsulated in a container in the NAS message, for example, in an N1 AI management container.

[0976] Step 18: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[0977] The AI ​​service data includes at least one of the following: UE data; model information;

[0978] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[0979] Step 19: Optionally, the AI ​​resource node obtains network data from a data source.

[0980] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0981] Step 20: Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[0982] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[0983] In this embodiment, the NAS interface function may also first send a communication resource message, including the following steps:

[0984] Step 1: The UE sends a first request message to the NAS interface function, where the first request message is used to request the network side to allocate network resources for the AI ​​service.

[0985] The first request message may be at least one of the following: a registration message, a service request message, an uplink NAS transmission message, and a newly added control message (used to establish a control connection).

[0986] Step 2: The NAS interface function selects the communication management function and the AI ​​service management function according to the first request message sent by the UE.

[0987] The NAS interface function can discover the communication management function instance using NRF or based on local configuration information, similar to AMF selecting SMF when establishing a PDU session. The NAS interface function requires additional considerations to select the communication management function, including at least one of the following: the communication management function supports forwarding AI-related messages, the communication management function supports establishing a bearer for AI services, and the communication management function supports communication with the AI ​​function (i.e., has an interface with the AI ​​function).

[0988] Optionally, before selecting an AI service management function instance, the UE's contract data may be requested from a unified data management function (UDM) and / or a unified database (UDR) to verify whether the AI ​​service-related contract information exists. If so, the subsequent process is performed; if not, the NAS interface function rejects the UE's AI service request.

[0989] Step 3: The NAS interface function sends a communication resource request message to the communication management function to request allocation of corresponding communication resources for the AI ​​service.

[0990] The communication resource request message may be at least one of the following: a PDU session establishment request message, a PDU session modification message, and a communication channel establishment request message (for establishing a data transmission channel).

[0991] Step 4: The communication management function sends a communication policy request message to the policy charging function to request a dynamic communication policy applicable to the PDUsession. The communication policy request message includes at least one of the following: UE ID, PDU session ID, DNN, S-NSSAI, associated AI session ID, AI service requirement description information, etc.

[0992] The policy charging function generates the communication policy corresponding to the AI ​​service. When allocating the communication policy, the policy charging function should comprehensively consider the end-to-end service requirements of the AI ​​service.

[0993] The policy charging function sends a communication policy request response message to the communication management function, which includes but is not limited to at least one of the following: a communication policy and an AS address.

[0994] The communication management function selects appropriate communication resource nodes according to the communication strategy, AI resource node information and AI strategy.

[0995] The communication resource node information includes at least one of the following:

[0996] IP address;

[0997] ID information;

[0998] FQDN.

[0999] Step 5: The communication management function establishes or modifies the corresponding communication bearer according to the communication strategy, and establishes or modifies the process with the existing N4session / N3 tunnel. The communication bearer is used to transmit AI service data.

[1000] The communication management function may also configure the AI ​​resource node information to the communication resource node, so that the communication resource node processes and forwards the AI ​​service data packet.

[1001] Step 6: The communication management function sends a communication resource request response message to the NAS interface function to notify that the communication resource allocation is successful, which may include the communication resource node information.

[1002] Step 7: The NAS interface function sends an AI resource establishment request message to the AI ​​service management function to request allocation of corresponding AI resources for the AI ​​service. The AI ​​resource establishment request message includes but is not limited to at least one of the following: AI policy, communication resource node information, the communication management function instance address, and AI resource node information feedback indication.

[1003] Step 8: The AI ​​service management function sends an AI policy request message to the policy charging function to request a dynamic AI policy applicable to the AI ​​session. The AI ​​policy request message includes at least one of the following: UE ID, AI session ID, DNN, S-NSSAI, associated PDU session ID, service requirement description information, etc.

[1004] The policy charging function generates AI policies corresponding to AI services. When allocating AI policies, the policy charging function should comprehensively consider the end-to-end service requirements of AI services.

[1005] The policy charging function sends an AI policy request response message to the AI ​​service management function, which includes but is not limited to at least one of the following: AI policy and AS address.

[1006] Step 9a: The AI ​​business management function sends a model request message to the AI ​​model management function to determine model information.

[1007] Step 9a is an optional step and needs to be performed under at least one of the following conditions:

[1008] The AI ​​session establishment request message does not include model information;

[1009] The AI ​​session establishment request message includes an indication that interaction with the AI ​​model management function is required;

[1010] The AI ​​business management function cannot obtain model information;

[1011] The AI ​​business management function selects the AI ​​model management function instance by considering at least one of the following factors:

[1012] DNN;

[1013] S-NSSAI;

[1014] AI business logo;

[1015] Supported AI services;

[1016] Service area information;

[1017] Supported model information;

[1018] Step 9b: The AI ​​model management function sends a model request response message to the AI ​​business management function to indicate model information.

[1019] The AI ​​business management function can determine the data source information based on the AI ​​model information. If the data source information has been determined, step 10 is not executed.

[1020] Step 10a: The AI ​​service management function sends a data source request message to the data plane management function to determine data source information;

[1021] Step 10a is an optional step and needs to be performed under at least one of the following conditions:

[1022] The AI ​​session establishment request message does not include data source information;

[1023] The AI ​​session establishment request message includes an indication that interaction with the data plane management function is required;

[1024] The AI ​​business management function cannot obtain data source information;

[1025] The AI ​​service management function selects the data plane management function instance based on at least one of the following factors:

[1026] DNN;

[1027] S-NSSAI;

[1028] AI business logo;

[1029] Supported AI services;

[1030] Service area information;

[1031] Data source information;

[1032] Step 10b: The data plane management function sends a model request response message to the AI ​​service management function to indicate data source information.

[1033] Step 11: The AI ​​service management function sends an AI resource request message to the AI ​​resource management function, and the AI ​​resource management function determines a suitable AI resource node according to the AI ​​resource request message and the AI ​​strategy.

[1034] The AI ​​resource request message includes at least one of the following:

[1035] Model information (model or model download address);

[1036] Data source information (data source address);

[1037] AI business computing volume (computing power size and type);

[1038] AS address.

[1039] Step 12: The AI ​​resource management node selects a suitable AI resource node according to the AI ​​resource request message.

[1040] Step 13: The AI ​​resource management function establishes an AI bearer and provides AI resource configuration information to guide the AI ​​resource node to process and forward AI service data.

[1041] The AI ​​resource configuration information includes at least one of the following:

[1042] Data source information (data source address);

[1043] AI service detection information, used for AI resource node detection and identification of AI service traffic;

[1044] AI service processing rules are used by AI resource nodes to process and forward AI service traffic;

[1045] AS address.

[1046] Step 14: The AI ​​resource management function sends an AI resource request response message to the AI ​​service management function, which includes AI resource node information. The AI ​​resource node information includes at least one of the following:

[1047] IP address;

[1048] ID information;

[1049] FQDN.

[1050] Step 15: The AI ​​service management function sends an AI resource establishment response message to the NAS interface function to notify that the AI ​​resource allocation is successful, which may include the AI ​​resource node information.

[1051] Step 16: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes but is not limited to at least one of the following: communication resource node and AI resource node information.

[1052] Step 17: The NAS interface function sends a first request response message corresponding to the request message in step 1 to the UE, to notify the UE of the result of the AI ​​service request, which includes AI resource node information. Optionally, the first request response message may also include at least one of the following: communication resource node information, model information.

[1053] The first request response message is encapsulated in a container in the NAS message, for example, in an N1 AI management container.

[1054] Step 18: The application APP triggers the transmission of AI service data. The AI ​​service data is carried on the PDU session and routed to the AI ​​resource node via the communication resource node.

[1055] The AI ​​service data includes at least one of the following: UE data; model information;

[1056] In this step, the communication management function does not know the AI ​​resource node information, and the UE can only set the AI ​​resource node information as the destination address.

[1057] Step 19: Optionally, the AI ​​resource node obtains network data from a data source.

[1058] The AI ​​resource node determines the data source information, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[1059] Step 20: Optionally, the AI ​​resource node interacts with the AS according to a service level agreement (SLA) to execute the complete AI service logic. The SLA indicates that the AI ​​resource node only performs partial service processing.

[1060] The AI ​​resource node determines the AS address, including but not limited to at least one of the following methods: obtaining from AI resource configuration information; obtaining from local configuration.

[1061] Embodiment 7

[1062] In this embodiment, the communication management function is used as the central control node. Fig.19 The protocol stack shown in FIG. 1 is used as the basis to enhance the NAS-SM layer. When the NAS-SM layer generates a communication resource request message, it includes the AI ​​service description information, which is used to instruct the network to allocate AI resources. The message encapsulation method is to encapsulate the AI ​​service description information in the communication resource request message, and encapsulate the communication resource request message in the NAS message.

[1063] In this embodiment, one implementation method is that the communication management function and the AI ​​service management function are separated. At this time, the steps of this embodiment are similar to those of the first embodiment, except that: the communication management function generates an AI resource establishment request message and sends it to the AI ​​service management function, which is an interaction between network elements; the communication management function can parse the AI ​​resource node information.

[1064] Another implementation method is: the communication management function and the AI ​​business management function are jointly set up, and the communication management function is responsible for allocating AI resources.

[1065] The AI ​​service connection establishment method provided in the embodiment of the present application can be executed by an AI service connection establishment device. In the embodiment of the present application, the AI ​​service connection establishment method executed by the AI ​​service connection establishment device is taken as an example to illustrate the AI ​​service connection establishment device provided in the embodiment of the present application.

[1066] See also Fig. 20 , Fig. 20 is a structural diagram of an AI service connection establishment device provided in an embodiment of the present application, and the device is applied to a terminal 1, such as Fig. 20 As shown, the AI ​​service connection establishment device includes:

[1067] A first sending module 11 is used to send a first request message to the network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service;

[1068] The first receiving module 12 is used to receive a first request response message returned by the network side, wherein the first request response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[1069] In an embodiment of the present application, the terminal requests the network side to allocate network resources for the AI ​​service. The network side can allocate communication resources and AI resources for the AI ​​service according to the request of the terminal, thereby solving the problem that the wireless mobile communication system cannot provide AI services for the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[1070] In some embodiments, the first request message includes at least one of the following:

[1071] AI business requirements description information;

[1072] Data network name DNN;

[1073] Single Network Slice Selection Assistance Information S-NSSAI;

[1074] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[1075] PDU session information;

[1076] Policy reference identifier;

[1077] Address of the application server AS.

[1078] In some embodiments, the AI ​​business requirement description information includes at least one of the following:

[1079] AI business logo;

[1080] AI business type;

[1081] AI service quality of experience (QoE) requirements;

[1082] AI business service quality QoS information;

[1083] Model parameters;

[1084] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[1085] In some embodiments, the type of the first request message includes at least one of the following:

[1086] Communication resource request message, used to request the network to allocate communication resources for AI services;

[1087] AI resource request message, used to request the network side to allocate AI resources for AI services;

[1088] Non-access layer NAS message, used to request the network side to allocate network resources for AI services.

[1089] In some embodiments, the communication resource request message includes at least one of the following:

[1090] A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[1091] PDU session modification request message, used to modify the PDU session;

[1092] Communication channel establishment request message, used to establish a communication bearer;

[1093] Communication channel modification request message, used to modify the communication bearer;

[1094] The AI ​​resource request message includes at least one of the following:

[1095] An AI resource establishment request message is used to establish an AI session, where the AI ​​session is a data transmission channel between the first network element and the terminal;

[1096] AI session establishment request message;

[1097] The NAS message includes at least one of the following:

[1098] Registration message;

[1099] Service request message;

[1100] Uplink NAS transmission message;

[1101] The first control message is used to establish or modify a control signaling connection.

[1102] In some embodiments, the first sending module 11 is further used to send AI service data to the first network element, where the AI ​​service data is carried by the AI ​​session and routed to the first network element via the second network element;

[1103] The first receiving module 12 is further used to receive the AI ​​service data processing result returned by the first network element.

[1104] In some embodiments, the AI ​​service data processing result is obtained by the first network element based on at least one of a model, terminal data and in-network data, and the in-network data is obtained by the first network element from a third network element.

[1105] The AI ​​service connection establishment 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 a device 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.

[1106] See also Fig.21 , Fig.21 is a structural diagram of an AI service connection establishment device provided in an embodiment of the present application, and the device is applied to a fourth network element 2, such as Fig.21 As shown, the AI ​​service connection establishment device includes:

[1107] A second receiving module 21 is used to receive a communication resource request message, where the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service;

[1108] A first processing module 22, configured to allocate communication resources according to the communication resource request message;

[1109] The second sending module 23 is used to return a communication resource establishment response message, and the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[1110] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services, solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[1111] In some embodiments, the fourth network element satisfies at least one of the following:

[1112] The fourth network element supports forwarding the AI ​​resource request message;

[1113] The fourth network element supports establishing a bearer for the AI ​​service;

[1114] The fourth network element supports communication with the AI ​​network function;

[1115] The fourth network element supports selection of AI network functions.

[1116] In some embodiments, the communication resource request message includes at least one of the following:

[1117] Terminal identification;

[1118] Protocol data unit PDU session identifier;

[1119] The associated AI session identifier;

[1120] Data network name DNN;

[1121] Single Network Slice Selection Assistance Information S-NSSAI;

[1122] Session type indication, used to indicate whether the session is a PDU session or an AI session;

[1123] PDU session information, used to indicate the attributes of the PDU session;

[1124] AI business requirements description information;

[1125] the address of the fifth network element;

[1126] Address of the application server AS.

[1127] In some embodiments, the type of the communication resource request message includes at least one of the following:

[1128] A PDU session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[1129] PDU session modification request message, used to modify the PDU session;

[1130] Communication channel establishment request message, used to establish a communication bearer;

[1131] The communication channel modification request message is used to modify the communication bearer.

[1132] In some embodiments, the first processing module 22 is further used to select a fifth network element, and the second sending module 21 is further used to send an AI resource establishment request message to the fifth network element, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service, and the AI ​​resource establishment request message is generated by the terminal or the fourth network element;

[1133] The second receiving module 23 is further configured to receive an AI resource establishment response message returned by the fifth network element, where the AI ​​resource establishment response message includes information about the first network element allocated to the AI ​​service.

[1134] In some embodiments, when the type of the communication resource request message is the PDU session establishment request message or the communication channel establishment request, there is a many-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier;

[1135] In the case where the type of the communication resource request message is the PDU session modification request message or the communication channel modification request, there is a one-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier.

[1136] In some embodiments, the AI ​​resource establishment request message includes at least one of the following: an AI policy, information of the second network element, an address of the fourth network element, and an information feedback indication of the first network element.

[1137] In some embodiments, the first processing module 22 is specifically configured to select the fifth network element according to at least one of the following:

[1138] Area of ​​interest;

[1139] Contract information for terminal AI services;

[1140] DNN;

[1141] S-NSSAI;

[1142] AI business logo;

[1143] Supported AI services;

[1144] The capabilities of the fifth network element.

[1145] In some embodiments, the second receiving module 23 receiving the communication resource request message includes at least one of the following:

[1146] Get it from the terminal;

[1147] Obtained from the sixth network element;

[1148] Obtained from the fifth network element.

[1149] In some embodiments, the second receiving module 23 is further used to obtain a communication strategy, an AI strategy, an address of an application server AS and / or information of the first network element, so as to determine the second network element, and the obtaining method includes at least one of the following:

[1150] Determining according to the communication resource request message;

[1151] Obtained from the fifth network element;

[1152] Obtained from the sixth network element;

[1153] Obtained from the seventh network element.

[1154] In some embodiments, when the communication strategy and the AI ​​strategy are obtained from the seventh network element,

[1155] In the case of deploying a dynamic policy control and charging PCC architecture, the second sending module 21 sends a policy request message to the seventh network element, requesting the seventh network element to allocate a policy, and receiving the communication policy and the AI ​​policy returned by the seventh network element, wherein the policy request message includes at least one of the following: a terminal identifier, a PDU session identifier, an associated AI session identifier, a DNN, an S-NSSAI, and an AI service requirement description information;

[1156] In the case where the dynamic PCC architecture is not deployed, the first processing module 22 determines the communication strategy and the AI ​​strategy according to at least one of the following: local configuration, AI service identifier, and policy reference identifier.

[1157] See also Fig. 22 , Fig. 22 is a structural diagram of an AI service connection establishment device provided in an embodiment of the present application, and the device is applied to the fifth network element 3, such as Fig.23 As shown, the AI ​​service connection establishment device includes:

[1158] The third receiving module 31 is used to receive an AI resource establishment request message, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service;

[1159] A second processing module 32, configured to allocate AI resources according to the AI ​​resource establishment request message;

[1160] The third sending module 33 is used to return an AI resource establishment response message, where the AI ​​resource establishment response message includes at least one of the following: information about the first network element and information about the second network element.

[1161] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services, solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[1162] In some embodiments, the AI ​​resource establishment request message includes at least one of the following:

[1163] AI business requirements description information;

[1164] Data network name DNN;

[1165] Single Network Slice Selection Assistance Information S-NSSAI;

[1166] Session type indication, used to indicate whether the session is a PDU session or an AI session;

[1167] PDU session information;

[1168] Policy reference identifier;

[1169] Address of the application server AS.

[1170] In some embodiments, the second processing module 32 is used to determine the AI ​​strategy and / or communication strategy according to the AI ​​business requirement description information.

[1171] In some embodiments, the third receiving module 33 is further used to obtain the AI ​​strategy and / or the communication strategy, and the obtaining method includes at least one of the following:

[1172] Get it based on local configuration;

[1173] Obtained from the seventh network element;

[1174] Obtained from the fourth network element;

[1175] Obtained from the sixth network element.

[1176] In some embodiments, the third sending module 31 is also used to send a communication resource request message to the fourth network element, requesting the fourth network element to select a second network element, and the communication resource request message includes at least one of the following: a communication strategy, information of the first network element, an address of the fifth network element, and an information feedback indication of the first network element.

[1177] In some embodiments, the third sending module 31 is further used to send an AI resource request message to the eighth network element, so that the eighth network element selects the first network element according to the AI ​​resource request message, and the AI ​​resource request message includes at least one of the following: model information, data source information, AI service computing amount, address of the application server AS, and AI strategy;

[1178] The third receiving module 33 is further configured to receive an AI resource request response message returned by the eighth network element, wherein the AI ​​resource request response message includes information about the first network element, and the information about the first network element includes at least one of the following:

[1179] Internet protocol IP address;

[1180] Identification information;

[1181] Fully qualified domain name FQDN.

[1182] In some embodiments, the third sending module 31 is further used to send a model request message to the ninth network element when the first condition is met, where the model request message is used to request model information;

[1183] The third receiving module 33 is further used to receive a model request response message sent by the ninth network element, where the model request response message is used to indicate model information;

[1184] The model information includes at least one of the following: model identification, model download address, model size, model training algorithm, model precision, accuracy, model factory, model function identification, model architecture, and weight parameters; the first condition includes at least one of the following:

[1185] The AI ​​resource establishment request message does not include model information;

[1186] The AI ​​resource establishment request message includes a first indication, indicating that interaction with the ninth network element is required;

[1187] The fifth network element cannot obtain the model information.

[1188] In some embodiments, the second processing module 32 is configured to select the ninth network element according to at least one of the following:

[1189] DNN;

[1190] S-NSSAI;

[1191] AI business logo;

[1192] Supported AI services;

[1193] Service area information;

[1194] Supported model information.

[1195] In some embodiments, the third sending module 31 is further configured to send a data source request message to the tenth network element when the second condition is met, wherein the data source request message is used to request data source information;

[1196] The third receiving module 33 is further used to receive a data source response message sent by the tenth network element, where the data source response message is used to indicate data source information;

[1197] The data source information includes at least one of the following: data source IP address, data source FQDN, data source identifier;

[1198] The second condition includes at least one of the following:

[1199] The AI ​​resource establishment request message does not include data source information;

[1200] The AI ​​resource establishment request message includes a second indication, indicating that interaction with the tenth network element is required;

[1201] The fifth network element cannot obtain data source information.

[1202] In some embodiments, the second processing module 32 is configured to select the tenth network element according to at least one of the following:

[1203] DNN;

[1204] S-NSSAI;

[1205] AI business logo;

[1206] Supported AI services;

[1207] Service area information;

[1208] Data source information.

[1209] In some embodiments, the communication resource request message includes at least one of the following:

[1210] A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[1211] PDU session modification request message, used to modify the PDU session;

[1212] Communication channel establishment request message, used to establish a communication bearer;

[1213] The communication channel modification request message is used to modify the communication bearer.

[1214] In some embodiments, when the type of the communication resource request message is a PDU session establishment request message or a communication channel establishment request, there is a many-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier;

[1215] In the case where the type of the communication resource request message is a PDU session modification request message or a communication channel modification request, there is a one-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier.

[1216] In some embodiments, receiving the AI ​​resource establishment request message includes at least one of the following:

[1217] Acquired from a fourth network element, where the AI ​​resource establishment request message is generated by the terminal or the fourth network element;

[1218] The AI ​​resource establishment request message is obtained from the sixth network element and is generated by the terminal.

[1219] In some embodiments, the third sending module 31 is further used to send an AI resource establishment request response to the sixth network element to feedback the AI ​​resource establishment result, and the AI ​​resource establishment request response includes at least one of the following: information of the first network element and an AI strategy.

[1220] In some embodiments, the AI ​​service requirement description information includes at least one of the following information:

[1221] AI business logo;

[1222] AI service quality of experience (QoE) requirements;

[1223] AI business service quality QoS information;

[1224] Model parameters;

[1225] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[1226] In some embodiments, the information of the first network element includes at least one of the following: an IP address of the first network element, identification information of the first network element, and a FQDN of the first network element;

[1227] The information of the second network element includes at least one of the following: the second network element IP address, the second network element identification information, and the second network element FQDN.

[1228] See also Fig.23 , Fig.23 is a structural diagram of an AI service connection establishment device provided in an embodiment of the present application, and the device is applied to the sixth network element 4, such as Fig.23 As shown, the AI ​​service connection establishment device includes:

[1229] The fourth receiving module 41 is used to receive a first request message, where the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal;

[1230] A third processing module 42, configured to select a fifth network element and a fourth network element;

[1231] The fourth sending module 43 is used to send an AI resource establishment request message to the fifth network element and send a communication resource request message to the fourth network element;

[1232] The fourth receiving module 41 is further configured to receive an AI resource establishment response message returned by the fifth network element, the AI ​​resource establishment response message including information of the first network element allocated for the AI ​​service, and receive a communication resource establishment response message returned by the fourth network element, the communication resource establishment response message including information of the second network element allocated for the AI ​​service;

[1233] The fourth sending module 43 is further configured to return a first request response message, where the first request response message includes information about the first network element and information about the second network element.

[1234] In an embodiment of the present application, the network side can allocate communication resources and AI resources for AI services, solving the problem that the wireless mobile communication system cannot provide AI services to the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[1235] In some embodiments, the third processing module 42 is configured to determine whether to send the communication resource request message or the AI ​​resource establishment request message first according to at least one of the following:

[1236] Local configuration;

[1237] The resource determination indication is used to indicate whether the AI ​​resources or the communication resources should be determined first.

[1238] In some embodiments, the first request message includes at least one of the following:

[1239] AI business requirements description information;

[1240] Data network name DNN;

[1241] Single Network Slice Selection Assistance Information S-NSSAI;

[1242] Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session;

[1243] PDU session information;

[1244] Policy reference identifier;

[1245] Address of the application server AS.

[1246] In some embodiments, the AI ​​business requirement description information includes at least one of the following:

[1247] AI business logo;

[1248] AI business type;

[1249] AI service quality of experience (QoE) requirements;

[1250] AI business service quality QoS information;

[1251] Model parameters;

[1252] The first information indicates whether network internal data and / or data source are needed. Optionally, the first information is used to indicate whether network internal data needs to be obtained from a data source.

[1253] In some embodiments, the communication resource request message includes at least one of the following:

[1254] A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal;

[1255] PDU session modification request message, used to modify the PDU session;

[1256] Communication channel establishment request message, used to establish a communication bearer;

[1257] Communication channel modification request message, used to modify the communication bearer;

[1258] The AI ​​resource request message includes at least one of the following:

[1259] An AI resource establishment request message is used to establish an AI session, where the AI ​​session is a data transmission channel between the first network element and the terminal;

[1260] AI session establishment request message;

[1261] In some embodiments, the third processing module 42 selects the fifth network element according to at least one of the following:

[1262] Area of ​​interest;

[1263] Contract information for terminal AI services;

[1264] DNN;

[1265] S-NSSAI;

[1266] AI business logo;

[1267] Supported AI services;

[1268] The capabilities of the fifth network element.

[1269] In some embodiments, the fourth network element satisfies at least one of the following:

[1270] The fourth network element supports forwarding the AI ​​resource request message;

[1271] The fourth network element supports establishing a bearer for the AI ​​service;

[1272] The fourth network element supports communication with the AI ​​network function;

[1273] The fourth network element supports selection of AI network functions.

[1274] The AI ​​service connection establishment device provided in the embodiment of the present application can achieve Figures 4 to 19 The various processes implemented by the method embodiment and achieving the same technical effect are not described here to avoid repetition.

[1275] like Fig.24 As shown, the embodiment of the present application further provides a communication device 60, including a processor 61 and a memory 62, and the memory 62 stores a program or instruction that can be run on the processor 61. For example, when the communication device 60 is a terminal, the program or instruction is executed by the processor 61 to implement the various steps of the above-mentioned AI service connection establishment method embodiment, and can achieve the same technical effect. When the communication device 60 is a network side, the program or instruction is executed by the processor 61 to implement the various steps of the above-mentioned AI service connection establishment method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[1276] 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 Figure 4 , Figure 8-Figure 19 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 mode of the above-mentioned method embodiment can be applied to this terminal embodiment and can achieve the same technical effect.

[1277] Specifically, Fig.25 A schematic diagram of the hardware structure of a terminal for implementing an embodiment of the present application.

[1278] The terminal 500 includes but is not limited to: a radio frequency unit 501, a network module 502, an audio output unit 503, an input unit 504, a sensor 505, a display unit 506, a user input unit 507, an interface unit 508, a memory 509 and at least some of the components of a processor 510.

[1279] Those skilled in the art will appreciate that the terminal 500 may also include a power source (such as a battery) for supplying power to various components, and the power source may be logically connected to the processor 510 through a power management system, thereby implementing functions such as managing charging, discharging, and power consumption management through the power management system. Fig.25 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.

[1280] It should be understood that in the embodiment of the present application, the input unit 504 may include a graphics processing unit (GPU) 5041 and a microphone 5042, and the graphics processor 5041 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 506 may include a display panel 5061, and the display panel 5061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 507 includes a touch panel 5071 and at least one of other input devices 5072. The touch panel 5071 is also called a touch screen. The touch panel 5071 may include two parts: a touch detection device and a touch controller. Other input devices 5072 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.

[1281] In the embodiment of the present application, after receiving downlink data from the network side, the RF unit 501 can transmit it to the processor 510 for processing; in addition, the RF unit 501 can send uplink data to the network side. Generally, the RF unit 501 includes but is not limited to an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

[1282] The memory 509 can be used to store software programs or instructions and various data. The memory 509 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 509 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 509 in the embodiment of the present application includes but is not limited to these and any other suitable types of memory.

[1283] The processor 510 may include one or more processing units; optionally, the processor 510 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 510.

[1284] Among them, the radio frequency unit 501 is used to send a first request message to the network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service; receive a first request response message returned by the network side, where the first request response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

[1285] In an embodiment of the present application, the terminal requests the network side to allocate network resources for the AI ​​service. The network side can allocate communication resources and AI resources for the AI ​​service according to the request of the terminal, thereby solving the problem that the wireless mobile communication system cannot provide AI services for the terminal. In this way, the operator network can not only provide basic communication guarantees, but also provide AI business processing functions, which not only expands the operator's business scenarios, but also can ensure the security of terminal data flow.

[1286] It can be understood that the implementation process of each implementation method mentioned in this embodiment can refer to the relevant description of the above-mentioned method embodiment, and achieve the same or corresponding technical effect. To avoid repetition, it will not be repeated here.

[1287] 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 5-Figure 7 The steps of the method embodiment shown. This network side embodiment corresponds to the above network side method embodiment, and each implementation process and implementation mode of the above method embodiment can be applied to this network side embodiment and can achieve the same technical effect.

[1288] Specifically, the embodiment of the present application also provides a network side device. Fig.26 As shown, the network side device 70 includes: an antenna 71, a radio frequency device 72, a baseband device 73, a processor 74 and a memory 75. The antenna 71 is connected to the radio frequency device 72. In the uplink direction, the radio frequency device 72 receives information through the antenna 71 and sends the received information to the baseband device 73 for processing. In the downlink direction, the baseband device 73 processes the information to be sent and sends it to the radio frequency device 72. The radio frequency device 72 processes the received information and sends it out through the antenna 71.

[1289] The method executed by the network side in the above embodiment may be implemented in the baseband device 73, which includes a baseband processor.

[1290] The baseband device 73 may include, for example, at least one baseband board on which a plurality of chips are arranged. Fig.26 As shown, one of the chips is, for example, a baseband processor, which is connected to the memory 75 via a bus interface to call the program in the memory 75 to execute the network device operations shown in the above method embodiment.

[1291] The network side may also include a network interface 76, which is, for example, a Common Public Radio Interface (CPRI).

[1292] Specifically, the network side device 70 of the embodiment of the present application further includes: instructions or programs stored in the memory 75 and executable on the processor 74, and the processor 74 calls the instructions or programs in the memory 75 to execute. Figure 21-23 The methods executed by the modules shown achieve the same technical effects, and therefore will not be described here in detail to avoid repetition.

[1293] The 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 the processor, each process of the above-mentioned AI service connection establishment method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[1294] 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.

[1295] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement each process of the above-mentioned AI service connection establishment method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[1296] 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.

[1297] The embodiment of the present application further provides 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 AI service connection establishment method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[1298] An embodiment of the present application also provides a wireless communication system, including: a terminal and a network side device, wherein the terminal can be used to execute the steps of the AI ​​service connection establishment method as described above, and the network side device can be used to execute the steps of the AI ​​service connection establishment method as described above.

[1299] 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 the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including one..." do 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 methods and devices in the embodiments 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.

[1300] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment method can be implemented by means of a computer software product plus a necessary general hardware platform, and of course, it 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 the terminal or network side to execute the methods described in each embodiment of the present application.

[1301] 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 method for establishing an AI service connection, characterized in that: include: The terminal sends a first request message to the network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service; The terminal receives a first request response message returned by the network side, where the first request response message includes at least one of the following: information about a first network element allocated by the network side to the AI ​​service, information about a second network element allocated by the network side to the AI ​​service, and information about the model.

2. The AI ​​service connection establishment method according to claim 1, characterized in that: The first request message includes at least one of the following: AI business requirements description information; Data network name DNN; Single Network Slice Selection Assistance Information S-NSSAI; Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session; PDU session information; Policy reference identifier; Address of the application server AS.

3. The AI ​​service connection establishment method according to claim 2, characterized in that: The AI ​​business requirement description information includes at least one of the following: AI business logo; AI business type; AI service quality of experience (QoE) requirements; AI business service quality QoS information; Model parameters; The first information indicates whether network internal data and data source are needed.

4. The AI ​​service connection establishment method according to claim 1, characterized in that: The type of the first request message includes at least one of the following: Communication resource request message, used to request the network to allocate communication resources for AI services; AI resource request message, used to request the network side to allocate AI resources for AI services; Non-access layer NAS message, used to request the network side to allocate network resources for AI services.

5. The AI ​​service connection establishment method according to claim 4, characterized in that: The communication resource request message includes at least one of the following: A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal; PDU session modification request message, used to modify the PDU session; Communication channel establishment request message, used to establish a communication bearer; Communication channel modification request message, used to modify the communication bearer; The AI ​​resource request message includes at least one of the following: An AI resource establishment request message is used to establish an AI session, where the AI ​​session is a data transmission channel between the first network element and the terminal; AI session establishment request message; The NAS message includes at least one of the following: Registration message; Service request message; Uplink NAS transmission message; The first control message is used to establish or modify a control signaling connection.

6. The AI ​​service connection establishment method according to claim 1, characterized in that: After the terminal receives the first request response message returned by the network side, the method further includes: The terminal sends AI service data to the first network element, where the AI ​​service data is carried by the AI ​​session and is routed to the first network element via the second network element; The terminal receives the AI ​​service data processing result returned by the first network element.

7. The AI ​​service connection establishment method according to claim 6, characterized in that: The AI ​​service data processing result is obtained by the first network element according to at least one of a model, terminal data and in-network data, and the in-network data is obtained by the first network element from a third network element.

8. A method for establishing an AI service connection, characterized in that: include: The fourth network element receives a communication resource request message, where the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service; The fourth network element allocates communication resources according to the communication resource request message, and returns a communication resource establishment response message, wherein the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

9. The method for establishing an AI service connection according to claim 8, characterized in that: The fourth network element satisfies at least one of the following: The fourth network element supports forwarding the AI ​​resource request message; The fourth network element supports establishing a bearer for the AI ​​service; The fourth network element supports communication with the AI ​​network function; The fourth network element supports selection of AI network functions.

10. The AI ​​service connection establishment method according to claim 8, characterized in that: The communication resource request message includes at least one of the following: Terminal identification; Protocol data unit PDU session identifier; The associated AI session identifier; Data network name DNN; Single Network Slice Selection Assistance Information S-NSSAI; Session type indication, used to indicate whether the session is a PDU session or an AI session; PDU session information, used to indicate the attributes of the PDU session; AI business requirements description information; the address of the fifth network element; Address of the application server AS.

11. The method for establishing an AI service connection according to claim 10, characterized in that: The type of the communication resource request message includes at least one of the following: A PDU session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal; PDU session modification request message, used to modify the PDU session; Communication channel establishment request message, used to establish a communication bearer; The communication channel modification request message is used to modify the communication bearer.

12. The AI ​​service connection establishment method according to claim 8, characterized in that: Before the fourth network element allocates communication resources according to the communication resource request message, the method further includes: The fourth network element selects a fifth network element, and sends an AI resource establishment request message to the fifth network element, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service, and the AI ​​resource establishment request message is generated by the terminal or the fourth network element; The fourth network element receives the AI ​​resource establishment response message returned by the fifth network element, where the AI ​​resource establishment response message includes information of the first network element allocated to the AI ​​service.

13. The AI ​​service connection establishment method according to claim 11, characterized in that: When the type of the communication resource request message is the PDU session establishment request message or the communication channel establishment request, there is a many-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier; In the case where the type of the communication resource request message is the PDU session modification request message or the communication channel modification request, there is a one-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier.

14. The method for establishing an AI service connection according to claim 12, characterized in that: The AI ​​resource establishment request message includes at least one of the following: an AI policy, information about the second network element, an address of the fourth network element, and an information feedback indication of the first network element.

15. The method for establishing an AI service connection according to claim 12, characterized in that: The fourth network element selects the fifth network element according to at least one of the following: Area of ​​interest; Contract information for terminal AI services; DNN; S-NSSAI; AI business logo; Supported AI services; The capabilities of the fifth network element.

16. The method for establishing an AI service connection according to claim 8, characterized in that: The fourth network element receiving the communication resource request message includes at least one of the following: Get it from the terminal; Obtained from the sixth network element; Obtained from the fifth network element.

17. The AI ​​service connection establishment method according to claim 8, characterized in that: The method further comprises: The fourth network element acquires the communication strategy, the AI ​​strategy, the address of the application server AS and / or the information of the first network element for determining the second network element, and the acquisition method includes at least one of the following: Determining according to the communication resource request message; Obtained from the fifth network element; Obtained from the sixth network element; Obtained from the seventh network element.

18. The AI ​​service connection establishment method according to claim 17, characterized in that: In a case where the communication strategy and the AI ​​strategy are acquired from the seventh network element, the fourth network element acquiring the communication strategy and the AI ​​strategy includes: In the case of deploying a dynamic policy control and charging PCC architecture, the fourth network element sends a policy request message to the seventh network element, requesting the seventh network element to allocate a policy, and receiving the communication policy and the AI ​​policy returned by the seventh network element, where the policy request message includes at least one of the following: a terminal identifier, a PDU session identifier, an associated AI session identifier, a DNN, an S-NSSAI, and an AI service requirement description information; When the dynamic PCC architecture is not deployed, the fourth network element determines the communication strategy and the AI ​​strategy according to at least one of the following: local configuration, AI service identifier, and policy reference identifier.

19. A method for establishing an AI service connection, characterized in that: include: The fifth network element receives an AI resource establishment request message, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service; The fifth network element allocates AI resources according to the AI ​​resource establishment request message, and returns an AI resource establishment response message, where the AI ​​resource establishment response message includes at least one of the following: information about the first network element and information about the second network element.

20. The AI ​​service connection establishment method according to claim 19, characterized in that: The AI ​​resource establishment request message includes at least one of the following: AI business requirements description information; Data network name DNN; Single Network Slice Selection Assistance Information S-NSSAI; Session type indication, used to indicate whether the session is a PDU session or an AI session; PDU session information; Policy reference identifier; Address of the application server AS.

21. The method for establishing an AI service connection according to claim 20, characterized in that: Before the fifth network element allocates the AI ​​resource according to the AI ​​resource establishment request message, the method further includes: The fifth network element determines an AI strategy and / or a communication strategy according to the AI ​​service requirement description information.

22. The AI ​​service connection establishment method according to claim 21, characterized in that: The method further comprises: The fifth network element acquires the AI ​​strategy and / or the communication strategy, and the acquisition method includes at least one of the following: Get it based on local configuration; Obtained from the seventh network element; Obtained from the fourth network element; Obtained from the sixth network element.

23. The AI ​​service connection establishment method according to claim 22, characterized in that: In the case of obtaining from the fourth network element, the method further includes: The fifth network element sends a communication resource request message to the fourth network element, requesting the fourth network element to select a second network element, wherein the communication resource request message includes at least one of the following: a communication strategy, information of the first network element, an address of the fifth network element, and an information feedback indication of the first network element.

24. The AI ​​service connection establishment method according to claim 19, characterized in that: The fifth network element allocating AI resources according to the AI ​​resource establishment request message includes: The fifth network element sends an AI resource request message to the eighth network element, so that the eighth network element selects the first network element according to the AI ​​resource request message, where the AI ​​resource request message includes at least one of the following: model information, data source information, AI service computing capacity, address of application server AS, and AI policy; The fifth network element receives an AI resource request response message returned by the eighth network element, where the AI ​​resource request response message includes information of the first network element, and the information of the first network element includes at least one of the following: Internet protocol IP address; Identification information; Fully qualified domain name FQDN.

25. The method for establishing an AI service connection according to claim 24, characterized in that: Before the fifth network element allocates the AI ​​resource according to the AI ​​resource establishment request message, the method further includes: When the first condition is met, the fifth network element sends a model request message to the ninth network element, where the model request message is used to request model information; The fifth network element receives a model request response message sent by the ninth network element, where the model request response message is used to indicate model information; The model information includes at least one of the following: model identification, model download address, model size, model training algorithm, model precision, accuracy, model factory, model function identification, model architecture, and weight parameters; the first condition includes at least one of the following: The AI ​​resource establishment request message does not include model information; The AI ​​resource establishment request message includes a first indication, indicating that interaction with the ninth network element is required; The fifth network element cannot obtain the model information.

26. The AI ​​service connection establishment method according to claim 25, characterized in that: The fifth network element selects the ninth network element according to at least one of the following: DNN; S-NSSAI; AI business logo; Supported AI services; Service area information; Supported model information.

27. The method for establishing an AI service connection according to claim 24, characterized in that: Before the fifth network element allocates the AI ​​resource according to the AI ​​resource establishment request message, the method further includes: When the second condition is met, the fifth network element sends a data source request message to the tenth network element, where the data source request message is used to request data source information; The fifth network element receives a data source response message sent by the tenth network element, where the data source response message is used to indicate data source information; The data source information includes at least one of the following: data source IP address, data source FQDN, data source identifier; The second condition includes at least one of the following: The AI ​​resource establishment request message does not include data source information; The AI ​​resource establishment request message includes a second indication, indicating that interaction with the tenth network element is required; The fifth network element cannot obtain data source information.

28. The method for establishing an AI service connection according to claim 27, characterized in that: The fifth network element selects the tenth network element according to at least one of the following: DNN; S-NSSAI; AI business logo; Supported AI services; Service area information; Data source information.

29. The method for establishing an AI service connection according to claim 23, characterized in that: The communication resource request message includes at least one of the following: A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and a terminal. PDU session modification request message, used to modify the PDU session; Communication channel establishment request message, used to establish a communication bearer; The communication channel modification request message is used to modify the communication bearer.

30. The method for establishing an AI service connection according to claim 29, characterized in that: When the type of the communication resource request message is a PDU session establishment request message or a communication channel establishment request, there is a many-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier; In the case where the type of the communication resource request message is a PDU session modification request message or a communication channel modification request, there is a one-to-one mapping relationship between the AI ​​session identifier and the PDU session identifier.

31. The method for establishing an AI service connection according to claim 19, characterized in that: The fifth network element receives the AI ​​resource establishment request message including at least one of the following: Acquired from a fourth network element, where the AI ​​resource establishment request message is generated by the terminal or the fourth network element; The AI ​​resource establishment request message is obtained from the sixth network element and is generated by the terminal.

32. The AI ​​service connection establishment method according to claim 31, characterized in that: In the case of obtaining from the sixth network element, the method further includes: The fifth network element sends an AI resource establishment request response to the sixth network element to feed back an AI resource establishment result, wherein the AI ​​resource establishment request response includes at least one of the following: information of the first network element and an AI strategy.

33. The method for establishing an AI service connection according to claim 20, characterized in that: The AI ​​business requirement description information includes at least one of the following information: AI business logo; AI service quality of experience (QoE) requirements; AI business service quality QoS information; Model parameters; The first information indicates whether network internal data and data source are needed.

34. The method for establishing an AI service connection according to claim 19, characterized in that: The information of the first network element includes at least one of the following: the first network element IP address, the first network element identification information, and the first network element FQDN; The information of the second network element includes at least one of the following: the second network element IP address, the second network element identification information, and the second network element FQDN.

35. A method for establishing an AI service connection, characterized in that: include: The sixth network element receives the first request message, where the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal; The sixth network element selects the fifth network element and the fourth network element, sends an AI resource establishment request message to the fifth network element, and sends a communication resource request message to the fourth network element; The sixth network element receives the AI ​​resource establishment response message returned by the fifth network element, where the AI ​​resource establishment response message includes information of the first network element allocated for the AI ​​service, and receives the communication resource establishment response message returned by the fourth network element, where the communication resource establishment response message includes information of the second network element allocated for the AI ​​service; The sixth network element returns a first request response message, where the first request response message includes information about the first network element and information about the second network element.

36. The method for establishing an AI service connection according to claim 35, characterized in that: The method further comprises: The sixth network element decides to first send the communication resource request message or the AI ​​resource establishment request message according to at least one of the following: Local configuration; The resource determination indication is used to indicate whether the AI ​​resources or the communication resources should be determined first.

37. The method for establishing an AI service connection according to claim 35, characterized in that: The first request message includes at least one of the following: AI business requirements description information; Data network name DNN; Single Network Slice Selection Assistance Information S-NSSAI; Session type indication, used to indicate whether the session is a protocol data unit PDU session or an AI session; PDU session information; Policy reference identifier; Address of the application server AS.

38. The method for establishing an AI service connection according to claim 37, characterized in that: The AI ​​business requirement description information includes at least one of the following: AI business logo; AI business type; AI service quality of experience (QoE) requirements; AI business service quality QoS information; Model parameters; The first information indicates whether network internal data and data source are needed.

39. The method for establishing an AI service connection according to claim 35, characterized in that: The communication resource request message includes at least one of the following: A protocol data unit (PDU) session establishment request message is used to establish a PDU session, where the PDU session is a data transmission channel between a communication resource node and the terminal; PDU session modification request message, used to modify the PDU session; Communication channel establishment request message, used to establish a communication bearer; Communication channel modification request message, used to modify the communication bearer; The AI ​​resource request message includes at least one of the following: An AI resource establishment request message is used to establish an AI session, where the AI ​​session is a data transmission channel between the first network element and the terminal; AI session establishment request message.

40. The method for establishing an AI service connection according to claim 35, characterized in that: The sixth network element selects the fifth network element according to at least one of the following: Area of ​​interest; Contract information for terminal AI services; DNN; S-NSSAI; AI business logo; Supported AI services; The capabilities of the fifth network element.

41. The method for establishing an AI service connection according to claim 35, characterized in that: The fourth network element satisfies at least one of the following: The fourth network element supports forwarding the AI ​​resource request message; The fourth network element supports establishing a bearer for the AI ​​service; The fourth network element supports communication with the AI ​​network function; The fourth network element supports selection of AI network functions.

42. An AI service connection establishment device, characterized in that: include: A first sending module, configured to send a first request message to a network side, where the first request message is used to request the network side to allocate network resources for the AI ​​service; The first receiving module is used to receive a first request response message returned by the network side, wherein the first request response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

43. An AI service connection establishment device, characterized in that: include: A second receiving module is used to receive a communication resource request message, where the communication resource request message is used to request the network side to allocate communication resources for the AI ​​service; A first processing module, configured to allocate communication resources according to the communication resource request message; The second sending module is used to return a communication resource establishment response message, wherein the communication resource establishment response message includes at least one of the following: information about the first network element allocated by the network side for the AI ​​service, information about the second network element allocated by the network side for the AI ​​service, and information about the model.

44. An AI service connection establishment device, characterized in that: include: A third receiving module is used to receive an AI resource establishment request message, where the AI ​​resource establishment request message is used to request the network side to allocate AI resources for the AI ​​service; A second processing module, configured to allocate AI resources according to the AI ​​resource establishment request message; The third sending module is used to return an AI resource establishment response message, where the AI ​​resource establishment response message includes at least one of the following: information about the first network element and information about the second network element.

45. An AI service connection establishment device, characterized in that: include: A fourth receiving module, configured to receive a first request message, where the first request message is used to request the network side to allocate network resources for the AI ​​service of the terminal; A third processing module, used for selecting a fifth network element and a fourth network element; a fourth sending module, configured to send an AI resource establishment request message to the fifth network element, and send a communication resource request message to the fourth network element; The fourth receiving module is further configured to receive an AI resource establishment response message returned by the fifth network element, the AI ​​resource establishment response message including information of a first network element allocated for the AI ​​service, and receive a communication resource establishment response message returned by the fourth network element, the communication resource establishment response message including information of a second network element allocated for the AI ​​service; The fourth sending module is further used to return a first request response message, where the first request response message includes information about the first network element and information about the second network element.

46. ​​A terminal, 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 AI ​​service connection establishment method according to any one of claims 1 to 7 are implemented.

47. A network side 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 AI ​​service connection establishment method according to any one of claims 8 to 41 are implemented.

48. A readable storage medium, characterized in that The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps of the method for establishing an AI service connection according to any one of claims 1 to 7 are implemented, or the steps of the method for establishing an AI service connection according to any one of claims 8 to 41 are implemented.

Images