Communication method and apparatus

Abstract

The present application relates to the field of communications, and provides a communication method and apparatus, which adapt to more complex service scenarios of AI / ML models. The method comprises: a network service consumption entity sends a first message to a network service production entity, wherein the first message is used for requesting the network service production entity to provide an AI / ML model used for completing a service between a network and a terminal; and the network service consumption entity receives a second message from the network service production entity, wherein the second message comprises information of the AI / ML model, and the AI / ML model comprises a first AI / ML model and a second AI / ML model; and when the first AI / ML model is deployed to the network and the second AI / ML model is deployed to the terminal, the first AI / ML model and the second AI / ML model are used for completing the service by means of interaction.

Description

Communication methods and devices

[0001] This application claims priority to Chinese Patent Application No. 202411616121.9, filed with the State Intellectual Property Office of China on November 12, 2024, entitled "Communication Method and Apparatus", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communications, and more particularly to a communication method and apparatus. Background Technology

[0003] Artificial intelligence (AI) enables machines to possess human-like intelligence, such as allowing machines to use computer hardware and software to simulate certain intelligent human behaviors. To achieve AI, machine learning (ML) methods can be used to acquire AI / ML models. AI / ML models can be used for reasoning, meaning they can be used to obtain output data corresponding to given input data. Therefore, AI / ML models have rich application scenarios in many industries. By licensing AI / ML models to model consumers in the network, these consumers can implement functions such as dialogue, writing, code generation, and image generation based on the AI / ML models.

[0004] However, under the current circumstances, how to deal with more complex AI / ML model application scenarios in the future is a hot research topic. Summary of the Invention

[0005] This application provides a communication method and apparatus to further expand the application scenarios of AI / ML models.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] In a first aspect, a communication method is provided, applied to a network service consuming entity. The method includes: sending a first message to a network service producing entity, the first message being used to request the network service producing entity to provide an artificial intelligence (AI) / machine learning (ML) model for completing business between the network and a terminal; and receiving a second message from the network service producing entity, the second message including information about the AI / ML model, the AI / ML model including a first AI / ML model and a second AI / ML model, wherein, in the case that the first AI / ML model is deployed to the network and the second AI / ML model is deployed to the terminal, the first AI / ML model and the second AI / ML model are used to complete business through interaction.

[0008] Therefore, in some potential AI / ML model business scenarios, the business requires at least two models deployed at different model consumers to interact with each other. In this case, the network service consumer entity can send a first message requesting the network service producer entity to provide the AI / ML model needed to complete the business. Correspondingly, the network service producer entity can send a second message to the network service consumer entity based on the first message. This second message includes information about the AI / ML model, which can refer to the model required to complete the business. For example, it could include a first AI / ML model for deployment on the network and a second AI / ML model for deployment on the terminal. Therefore, compared to the authorization scenarios for a single AI / ML model in existing technologies, this solution can authorize multiple functionally related AI / ML models to adapt to more complex business scenarios.

[0009] In one possible design, the first message includes first information, which is used to indicate the business. That is, the network service consumer entity can request models at the business level, deploying different models for different business needs, making model deployment more flexible.

[0010] Optionally, the first piece of information includes a service identifier. For example, the service identifier could be an analytics ID to indicate the service. The analytics ID can be an existing message structure. If a new service requires implementation through at least two models deployed separately to the network and terminals, its service requirements can be implemented using existing information elements, reducing implementation complexity and making it more standards-friendly. Alternatively, a new information element structure can be defined for the analytics ID, decoupling the new service from existing services and enabling greater flexibility.

[0011] In one possible design, the method described in the first aspect further includes: in response to the second message, sending information about the second AI / ML model to the terminal so that the second AI / ML model can be deployed to the terminal.

[0012] Optionally, sending the information of the second AI / ML model to the terminal includes: if the terminal is a terminal authorized to use the second AI / ML model, sending the information of the second AI / ML model to that terminal. This ensures that the information of the second AI / ML model is sent to authorized terminals, avoiding sending the information of the second AI / ML model to terminals without authorization, which could lead to information security risks and unnecessary overhead.

[0013] Optionally, the second message may also include a terminal identifier list, which indicates the identifiers of terminals that are permitted to use the second AI / ML model.

[0014] Optionally, the method further includes: obtaining the identifier of the terminal; confirming that the identifier of the terminal is in the terminal identifier list.

[0015] It can be seen that when the terminal's identifier is in the terminal identifier list corresponding to the second AI / ML model, it means that the terminal has the right to use the second AI / ML model. Thus, the information of the second AI / ML model is sent to the terminal to ensure information security.

[0016] Optionally, any terminal can be allowed to use the second AI / ML model. In other words, network service consuming entities do not need to compare whether the terminal's identifier matches the terminal identifier list to send the information of the second AI / ML model to any terminal with business needs, avoiding the overhead of judging whether to deploy the model and expanding the scope of business applicability.

[0017] Optionally, the terminal may be identified as a permanent device identifier, a type approval code, or a user permanent identifier to indicate the type or device of the terminal.

[0018] In one possible design, the first message also includes the terminal's identifier. That is, the network service consumer entity can send the identifier of the terminal whose model is to be acquired to the network service producer entity, so that the network service producer entity can confirm that the terminal is allowed to use the second AI / ML model based on the terminal's identifier, and then send the second message to the network service consumer entity. Conversely, for terminals that fail the verification and are confirmed to have no usage rights, the second message is not sent to the network service consumer entity to avoid redundant overhead.

[0019] In one possible design, the method described in the first aspect further includes: in response to the second message, sending information about the first AI / ML model to other entities in the network besides the network service consumer entity. That is, the network service consumer entity can act as a proxy to obtain the model for other entities in the network that may not be able to communicate directly with the network service producer entity. For example, the network service producer entity is a network data analytics function (NWDAF) network element, the network service consumer entity is an access and mobility management function (AMF) network element, and the other entity that needs to deploy the model is the radio access network (RAN). In a standard architecture, the RAN cannot directly communicate with the NWDAF network element through a service interface; therefore, it can obtain the model on behalf of the RAN through the AMF network element and send the information about the first AI / ML model to the RAN, thus achieving model deployment.

[0020] Optionally, sending the information of the first AI / ML model to other entities in the network besides the network service consuming entity includes: sending the information of the first AI / ML model to other entities if those entities are authorized to use the first AI / ML model. This ensures that the information of the first AI / ML model is sent to other entities with usage rights, avoiding sending the information of the first AI / ML model to other entities without usage rights, which could lead to information security risks and unnecessary costs.

[0021] Optionally, the second message also includes a list of entity identifiers, which indicates the identifiers of entities authorized to use the first AI / ML model. Sending the first AI / ML model to other entities whose identifiers belong to the list of entity identifiers ensures that the first AI / ML model is only sent to entities with model usage rights, thus ensuring information security.

[0022] Optionally, the network service consumption entity is the access and mobility management network element, and other entities are the access network equipment for terminal access.

[0023] In one possible design, the network service consumer entity is an entity in the network, and the method further includes: in response to the second message, deploying the first AI / ML model locally on the network service consumer entity, so that the first AI / ML model is deployed in the network.

[0024] Optionally, the network service consumer entity is the access network equipment accessed by the terminal.

[0025] Secondly, a communication method is provided, applied to a network service producing entity. The method includes: receiving a first message from a network service consuming entity, the first message being used to request the network service producing entity to provide an artificial intelligence (AI) / machine learning (ML) model for completing business between the network and a terminal; and sending a second message to the network service consuming entity, the second message including information about the AI / ML model, the AI / ML model including a first AI / ML model and a second AI / ML model, wherein, in the case that the first AI / ML model is deployed to the network and the second AI / ML model is deployed to the terminal, the first AI / ML model and the second AI / ML model are used to complete business through interaction.

[0026] It is understandable that the technical effects of the method described in the second aspect can also refer to the relevant introduction of the method described in the first aspect above, and will not be repeated here.

[0027] In one possible design, the first message further includes first information, which is used to indicate the business. The method described in the second aspect further includes: obtaining information about a first AI / ML model and information about a second AI / ML model based on the first information. That is, the network service production entity can determine, based on the first information, that the business requested by the network service message entity requires at least two functionally related AI / ML models to complete, and further determine the first AI / ML model and the second AI / ML model to achieve model deployment in complex business scenarios.

[0028] In one possible design, sending a second message to the network service consumer entity includes: sending a second message to the network service consumer entity based on its identifier corresponding to a service permission identifier. The service permission identifier indicates the entity authorized to request the AI / ML model corresponding to the service. The service permission identifier can be an interoperability ID, which can be an existing structure reused for new services requiring at least two AI / ML models to be deployed to the network and terminals respectively; alternatively, a new structure can be defined. The second message is sent only after confirming that the network service consumer entity is authorized to request the AI / ML model corresponding to the service, ensuring information security.

[0029] Optionally, the first message also includes the identifier of the network service consumer entity. That is, the network service consumer entity can send its own identifier to the network service producer entity to ensure that the information of the first AI / ML model and the information of the second AI / ML model are sent to the network service consumer that is allowed to request the AI / ML model, thereby avoiding information security risks and unnecessary expenses.

[0030] In one possible design, the second message further includes at least one of the following: a terminal identifier list or an entity identifier list, wherein the terminal identifier list indicates identifiers of terminals permitted to use the second AI / ML model, and the entity identifier list indicates identifiers of entities permitted to use the first AI / ML model. It is understood that the technical effects of the method described in this solution can also be referred to the relevant description of the method described in the first aspect above, and will not be repeated here.

[0031] In one possible design, the entity that can be used in the first AI / ML model is the access network device.

[0032] Thirdly, a communication device is provided, the communication device including a module for performing the method described in any one of the first to second aspects.

[0033] In one possible design, the communication device described in the third aspect may further include a transceiver. This transceiver may be a transceiver circuit or an interface circuit. The transceiver can be used for communication between the communication device described in the third aspect and other communication devices.

[0034] In one possible design, the communication device described in the third aspect may further include a memory. This memory may be integrated with the processor or disposed separately. The memory may be used to store instructions relating to the methods of any of the first to second aspects.

[0035] In the embodiments of this application, the communication device described in the third aspect may be a network device, or a chip (system) or other component or assembly disposed in the network device, or a device containing the network device.

[0036] It is understood that the technical effects of the device described in the third aspect can also be referred to the relevant descriptions of the methods in any of the first to second aspects above, and will not be repeated here.

[0037] Fourthly, a communication device is provided. The communication device includes a processor coupled to a memory, the processor being configured to execute instructions stored in the memory such that the communication device performs the method described in any one of the first to second aspects.

[0038] In one possible design, the communication device described in the fourth aspect may further include a transceiver. This transceiver may be a transceiver circuit or an interface circuit. The transceiver can be used for communication between the communication device described in the fourth aspect and other communication devices.

[0039] In the embodiments of this application, the communication device described in the fourth aspect may be a network device described in any one of the first to second aspects, or a chip (system) or other component or assembly disposed in the network device, or a device containing the network device.

[0040] Furthermore, the technical effects of the communication device described in the fourth aspect can be referred to the technical effects of the method described in any one of the first or second aspects, and will not be repeated here.

[0041] Fifthly, a communication device is provided, comprising: a processor and a memory; the memory being used to store instructions that, when executed by the processor, cause the communication device to perform the method as described in any one of the first to second aspects.

[0042] In one possible design, the communication device described in the fifth aspect may further include a transceiver. This transceiver may be a transceiver circuit or an interface circuit. The transceiver can be used by the communication device described in the third aspect to communicate with other communication devices.

[0043] In the embodiments of this application, the communication device described in the fifth aspect may be a network device described in any one of the first to second aspects, or a chip (system) or other component or assembly disposed in the network device, or a device containing the network device.

[0044] Furthermore, the technical effects of the communication device described in the fifth aspect can be referred to the technical effects of the method described in any one of the first or second aspects, and will not be repeated here.

[0045] A sixth aspect provides a chip comprising: a controller and an interface circuit, wherein the controller is configured to interact with other devices via the interface circuit to perform the method as described in any one of the first to second aspects.

[0046] A seventh aspect provides a communication system. The communication system includes a first manager for performing the method described in the first aspect, and a second manager for performing the method described in the second aspect.

[0047] Eighthly, a computer-readable storage medium is provided, the computer-readable storage medium including storage of a computer program or instructions that, when executed, cause the method described in any one of the first to second aspects to be performed.

[0048] A ninth aspect provides a computer program product comprising a computer program or instructions that, when executed, cause the method described in any one of the first to second aspects to be performed. Attached Figure Description

[0049] Figure 1 is a schematic diagram of the architecture of a 5G system;

[0050] Figure 2 is a schematic diagram of the licensing process for AI / ML models;

[0051] Figure 3 is a schematic diagram of the architecture of the communication system provided in an embodiment of this application;

[0052] Figure 4 is a flowchart illustrating the communication method provided in an embodiment of this application;

[0053] Figure 5 is a schematic flowchart of the communication method provided in an embodiment of this application;

[0054] Figure 6 is a flowchart illustrating the communication method provided in an embodiment of this application.

[0055] Figure 7 is a schematic flowchart of the communication method provided in the embodiment of this application;

[0056] Figure 8 is a schematic diagram of the communication device provided in an embodiment of this application;

[0057] Figure 9 is a schematic diagram of the structure of the communication device provided in the embodiment of this application. Detailed Implementation

[0058] The technical solutions of this application embodiment can be applied to various communication systems, such as Wi-Fi systems, vehicle-to-everything (V2X) communication systems, device-to-device (D2D) communication systems, vehicle-to-everything (V2X) communication systems, fourth-generation (4G) mobile communication systems, such as long-term evolution (LTE) systems, worldwide interoperability for microwave access (WiMAX) communication systems, fifth-generation (5G) mobile communication systems, such as new radio (NR) systems, and future communication systems.

[0059] For ease of understanding, the technical terms used in this application will be introduced below.

[0060] 1. The fifth-generation (5G) mobile communication system (5G system, 5GS), abbreviated as 5G system:

[0061] Figure 1 is a schematic diagram of the 5GS architecture. As shown in Figure 1, 5GS includes: an access network (AN) and a core network (CN), and may also include: terminals.

[0062] The aforementioned terminal can be a terminal with transceiver capabilities, or a chip or chip system that can be installed on the terminal. This terminal can also be referred to as user equipment (UE), access terminal, subscriber unit, user station, mobile station (MS), mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication equipment, user agent, or user apparatus. The terminals in the embodiments of this application may be mobile phones, cellular phones, smartphones, tablets, wireless data cards, personal digital assistants (PDAs), wireless modems, handsets, laptop computers, machine-type communication (MTC) terminals, computers with wireless transceiver capabilities, virtual reality (VR) terminals, augmented reality (AR) terminals, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in remote medical care, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, vehicle-mounted terminals, roadside units (RSUs) with terminal functions, etc. The terminal of this application may also be an on-board module, on-board unit, on-board component, on-board chip or on-board unit that is built into a vehicle as one or more components or units.

[0063] The aforementioned AN is used to implement access-related functions, providing network access capabilities for authorized users and determining transmission links of different quality levels to transmit user data based on user level, service requirements, etc. The AN forwards control signals and user data between the terminal and the CN. The AN may include access network equipment, also known as radio access network (RAN) equipment.

[0064] The Network Center (CN) is primarily responsible for maintaining the subscription data of the mobile network and providing terminals with functions such as session management, mobility management, policy management, and security authentication. The CN mainly includes all or some of the following functions: User Plane Function (UPF), Authentication Server Function (AUSF), Access and Mobility Management Function (AMF), Session Management Function (SMF), Network Slice Selection Function (NSSF), Network Exposure Function (NEF), Network Repository Function (NRF), Policy Control Function (PCF), Unified Data Management (UDM), Unified Data Repository (UDR), Network Data Analytics Function (NWDAF), Analytics Data Repository Function (ADRF), and Application Function (AF).

[0065] As shown in Figure 1, the UE accesses the 5G network through the RAN device. The UE communicates with the AMF through the N1 interface (N1 for short); the RAN communicates with the AMF through the N2 interface (N2 for short); the RAN communicates with the UPF through the N3 interface (N3 for short); the SMF communicates with the UPF through the N4 interface (N4 for short); and the UPF accesses the data network (DN) through the N6 interface (N6 for short). Furthermore, the control plane functions shown in Figure 1, such as AUSF, AMF, SMF, NSSF, NEF, NRF, PCF, UDM, UDR, or AF, interact using service-oriented interfaces. For example, AUSF provides the service interface Nausf; AMF provides the service interface Namf; SMF provides the service interface Nsmf; NSSF provides the service interface Nnssf; NEF provides the service interface Nnef; NRF provides the service interface Nnrf; PCF provides the service interface Npcf; UDM provides the service interface Nudm; UDR provides the service interface Nudr; NWDAF provides the service interface Nnwdaf; ADRF provides the service interface Nadrf; and AF provides the service interface Naf.

[0066] RAN equipment can be a device that provides access to terminals. For example, RAN equipment may include: access network equipment in future mobile communication systems, such as base stations, or in future mobile communication systems, the network equipment may have other naming conventions, all of which are covered within the protection scope of the embodiments of this application, and this application does not limit them in any way. Alternatively, RAN equipment may also include 5G, such as gNB in ​​a new radio (NR) system, or one or a group of antenna panels (including multiple antenna panels) of a base station in 5G, or it may be a network node constituting a gNB, transmission and reception point (TRP) or transmission point (TP) or transmission measurement function (TMF), such as a building base band unit (BBU), or a centralized unit (CU) or distributed unit (DU), an RSU with base station function, or a wired access gateway, or the core network of 5G. Alternatively, RAN equipment may also include access points (APs) in wireless fidelity (WiFi) systems, wireless relay nodes, wireless backhaul nodes, various forms of macro base stations, micro base stations (also known as small stations), relay stations, access points, wearable devices, vehicle-mounted equipment, and so on.

[0067] The User-Defined Processing (UPP) is primarily responsible for user data processing (forwarding, receiving, billing, etc.). For example, a UPF can receive user data from a data network (DN) and forward it to the terminal through access network equipment. Alternatively, a UPF can receive user data from the terminal through access network equipment and forward it to the DN. A DN refers to the operator's network that provides data transmission services to users. Examples include Internet Protocol (IP) multimedia services (IMS) and the Internet. A DN can be an external network of the operator or a network controlled by the operator, used to provide services to terminals. In a Protocol Data Unit (PDU) session, the UPF directly connected to the DN via N6 is also called the Protocol Data Unit Session Anchor (PSA).

[0068] AUSF is primarily used to perform security authentication for terminals.

[0069] AMF is primarily used for mobility management in mobile networks. Examples include user location updates, user network registration, and user handover.

[0070] SMF is primarily used for session management in mobile networks. This includes session establishment, modification, and release. Specific functions include assigning Internet Protocol (IP) addresses to users and selecting a UPF (User-Defined Provider) to handle packet forwarding.

[0071] The PCF primarily supports providing a unified policy framework to control network behavior, delivering policy rules to control-layer network functions, and acquiring user subscription information related to policy decisions. The PCF can provide policies to the AMF and SMF, such as Quality of Service (QoS) policies and slice selection policies.

[0072] NSSF is primarily used to select network slices for terminals.

[0073] NEF is primarily used to support the opening of capabilities and events.

[0074] UDM is primarily used to store user data, such as contract data and authentication / authorization data.

[0075] UDR is primarily used to store structured data, including contract data, policy data, externally exposed structured data, and application-related data.

[0076] AF primarily supports interaction with CN to provide services, such as influencing data routing decisions, policy control functions, or providing third-party services to the network side.

[0077] NRF is primarily used to receive NF discovery requests from NF instances and provide information about the (discovered) NF instances to the requesting NF instance; it also maintains NF configuration files for available NF instances and their supported services.

[0078] NWDAF is primarily used to obtain statistical information, metrics, and events from 5G network functions; to perform machine learning modeling on streaming data; and to develop AI / ML models based on the obtained statistical data.

[0079] ADRF is primarily used for storing and retrieving collected data and analysis, and also for storing AI / ML models.

[0080] It is understood that the functions mentioned in the embodiments of this application can also refer to functional network elements or functional entities. For example, UPF can be described as a UPF network element, AMF can be described as an AMF network element, SMF can be described as an SMF network element, PCF can be described as a PCF network element, and so on, without limitation. In addition, NWDAF network elements can also be called NWDAF instances or NWDAF entities, or any other possible naming, which is not limited in the embodiments of this application.

[0081] Furthermore, the aforementioned network elements can be collectively referred to as network functions (NFs). NFs that provide network services can be called NF service producers (NFp), and NFs that consume network services can be called NF service consumers (NFc). An NF can be either an NFc or an NFp, depending on whether it is consuming or providing services. For example, when an AMF requests an NWDAF to provide it with an AI / ML model service, the AMF, as the service consumer, can be called an NFc, and the NWDAF, as the service producer, can be called an NFp.

[0082] 2. Artificial intelligence (AI) / machine learning (ML):

[0083] AI can endow machines with human-like intelligence, allowing them to simulate certain intelligent human behaviors using computer hardware and software. To achieve artificial intelligence, machine learning (ML) methods can be used to obtain AI / ML models. These models can be used for reasoning, meaning they can be used to obtain output data corresponding to given input data. This allows AI / ML models to have rich application scenarios in numerous industries, especially in the recently emerging terminal intelligent applications (applications backed by large models, such as chatGPT), applied to areas such as dialogue, writing, code generation, and image generation. The interactive experience depends primarily on the effectiveness of the third-party application's (APP) AI / ML model and the quality of the communication network.

[0084] Currently, AI / ML models can be applied to communication networks, such as 5GS network architecture, to assist network services (NFs) in providing network services such as data processing, user analysis, and beam selection. The deployment process of AI / ML models in communication networks is described in detail below, and the authorization process for AI / ML models is shown in Figure 2.

[0085] S201, NFp is registered with NRF.

[0086] NFp, such as NWDAF network elements containing model training logical function (MTLF), registers its NF profile in NRF as described in Section 5.2 of TS23.288

[0105] , including the analytics ID and its corresponding interoperability indicator. The interoperability indicator is a list of NWDAF providers (device vendors) that are allowed to obtain AI / ML models from the NFp.

[0087] S202, NFC is registered with NRF.

[0088] For NWDAF network elements that include analytics logical function (AnLF), NFc registers its NF profile in NRF, including the vendor ID. If NFc is an NWDAF that includes MTLF, it also registers the analytics ID and its corresponding interoperability ID.

[0089] S203, Encrypted AI / ML Model.

[0090] AI / ML models are required to be stored in encrypted format on NFp, NFc, and the platform used to store the AI / ML models (such as ADRF), unless NFp, NFc, and ADRF all belong to the same communication system and are within the same equipment vendor and operator's security domain. The encryption method for AI / ML models is vendor-specific, and this article does not cover key distribution.

[0091] S204, NFp sends a Nadrf_MLModelManagement_Storage Request message to ADRF.

[0092] If an NFp (such as an NWDAF containing an MTLF) decides to store the AI / ML model in the ADRF, the NFp triggers the model management storage request message as described in TS23.288

[0105] . Optionally, this message includes the MTLF ID, the AI / ML model ID, the model's address in the MTLF (UR11), and a list of allowed NFcs. If there is no list of allowed NFcs, it means that only the NFp storing the AI / ML model is allowed to retrieve it.

[0093] S205, ADRF sends a model management storage response message to NFp.

[0094] As described in TS23.288

[0105] , ADRF sends a model management storage response message to NFp. Optionally, the message includes the AI / ML model ID and the model's address in ADRF (UR12).

[0095] S206, NFC should be set to NFp.

[0096] NF uses the requested analytics ID to perform an NF discovery request (Nnrf_NFDiscovery_Request) operation to select the appropriate NFp.

[0097] In the case where an NFc (such as an NWDAF containing an MTLF) requests an AI / ML model on behalf of another AI / ML model consumer (such as an NWADAF containing an AnLF), prior to step S206, the AI / ML model consumer may request a token for analysis ID from the NRF according to the process described in step S207 to authorize the AI / ML model consumer to use the model retrieval service provided by the NFp, and send the model retrieval service request along with the analysis ID, the obtained token, the device vendor ID, and the AI / ML model consumer's clear channel assessment (CCA) to the NFp.

[0098] S207, NFc sends a token acquisition request (Nnrf_AccessToken_Get Request) message to the NRF network element.

[0099] The token acquisition request message contains the device vendor ID and analytics ID of the NFC. If the NFC is an NWDAF that includes MTLF, the request also includes an interoperability indicator corresponding to the analytics ID.

[0100] When an NFc (such as an NWDAF containing MTLF) requests an AI / ML model on behalf of another AI / ML model consumer (such as an NWADAF containing AnLF), the token acquisition request should also include the NF instance ID and device vendor ID of the AI / ML model consumer, and the NFc should also include the CCA of the AI / ML model consumer that it received in the service request from the AI / ML model consumer.

[0101] S208, the NRF network element sends a token acquisition response (Nnrf_AccessToken_Get Response) message to the NFC.

[0102] The NRF checks whether NFc is authorized to access the requested services of NFp. The NRF verifies whether the NFc's vendor ID is included in the interoperability indicator corresponding to the NFp's analytics ID. If NFc is an NWDAF containing MTLF, the NRF also verifies that the interoperability indicator of NFc's analytics ID is a subset of the interoperability indicator corresponding to the NFp's analytics ID. If authorization is successful, the NRF grants Token 1 to NFc, which contains the analytics ID specified in Clause 13.4.1.

[0103] When an NFc (such as an NWDAF containing an MTLF) requests an AI / ML model on behalf of another AI / ML model consumer (such as an NWADAF containing an AnLF), the NRF also verifies whether the AI / ML model consumer's vendor ID is included in the interoperability indicator corresponding to the NFp's analytics ID. If authorization is successful, a token containing the AI / ML model consumer's NF instance ID is granted. The NRF can also authenticate the AI / ML model consumer based on an SBA method described in Clause 13.3.1.2.

[0104] S209, NFc sends an AI / ML model preparation request (Nnwdaf_MLModelProvision Request) message to NFp.

[0105] The AI / ML model preparation request message sent by NFc to NFp includes the analytics ID, vendor ID, and token 1 to retrieve the AI / ML model corresponding to the analytics ID in NFp. When NFc (such as NWDAF containing MTLF) requests an AI / ML model on behalf of another AI / ML model consumer (such as NWADAF containing AnLF), the AI / ML model preparation request message should also include the AI / ML model consumer's NF instance ID, vendor ID, and CCA.

[0106] Note 1: NFp relies on NRF for authorization and checks the device vendor ID of NFC. NFp cannot verify the device vendor ID on its own.

[0107] S210, authorize NFc and store NFc ID.

[0108] NFp authenticates NFc and verifies the access token specified in Clause 13.4.1.1.2, ensuring that the analytics ID is included in the access token. If authentication is successful, NFp determines the AI / ML model for the requested analytics ID and stores the NFc's NF instance ID as part of the list of allowed NFcs for the AI / ML model.

[0109] When an NFc (such as an NWDAF containing MTLF) requests an AI / ML model on behalf of another AI / ML model consumer (such as an NWADAF containing AnLF), NFp also verifies the AI / ML model consumer in accordance with Terms 13.3.2 and 13.4.1, and ensures that the AI / ML model consumer is authorized based on the NF instance ID in the access token provided by the NFc. NFp also stores the NF instance ID of the AI / ML model consumer as part of the list of allowed NFcs for the AI / ML model.

[0110] S211, NFp sends a model management storage request (Nadrf_MLModelManagement_Storage Request) message to the ADRF network element.

[0111] If the identified AI / ML model is stored in the ADRF, and the NFc is not yet in the ADRF's list of allowed NFcs, the NFp sends a model management storage request message to the ADRF element. This message includes the NFc's NF instance ID and the AI / ML model ID, triggering an update to add the NFc to the list of allowed NFcs. The ADRF verifies whether the NFp sending the request is the NFp storing the model. Then, the ADRF stores the list of allowed NFc instance IDs for the AI / ML model referenced by the AI / ML model ID.

[0112] S212, the ADRF network element returns a model management storage response message to NFp.

[0113] Specifically, the response message includes the AI / ML model ID.

[0114] S213, NFp sends an AI / ML model preparation response (Nnwdaf_MLModelProvision Response) message to NFc.

[0115] The AI / ML model preparation response message includes the AI / ML model ID, the AI / ML model file address, and optionally, the ADRF set ID.

[0116] Optionally, if the AI / ML model prepare response message in step S213 provides an ADRF set ID, then the process further includes:

[0117] S214, NFc sends a token acquisition request (Nnrf_AccessToken_Get Request) message to the NRF network element.

[0118] NFc requests an access token from ADRF from NRF to obtain authorization and retrieve the AI / ML model stored in ADRF as described in Clause 13.4.1.

[0119] S215, the NRF network element sends a token acquisition response (Nnrf_AccessToken_Get Response) message to the NFC.

[0120] The NRF verifies whether the NFc is authorized to access the services provided by the ADRF. If the verification is successful, the NRF grants a token (token2) based on the information provided in the NF profile in the ADRF.

[0121] S216, NFc sends a model management retrieval request (Nadrf_MLModelManagement_Retrieva Request) message to the ADRF network element.

[0122] As described in Clause 10.3.4TS23.288

[0105] , the message includes the analytics ID and a token (token2).

[0123] S217, the ADRF network element sends a model management retrieval response (Nadrf_MLModelManagement_Retrieva Response) message to the NFc.

[0124] ADRF authenticates the NFc and verifies the token (token2) as specified in Clause 13.4.1.1.12. ADRF also verifies that the NFc's NF ID is included in the list of allowed NFcs for the AI / ML model, and / or is the same as the NF ID of the NFp storing the model. If the verification is successful, ADRF sends a model management retrieval response message to the NFc, containing the address of the AI / ML model stored in ADRF.

[0125] S218, NFc acquires and decrypts AI / ML models.

[0126] NFc retrieves the AI / ML model from NFp or ADRF based on the AI / ML model file address and decrypts the AI / ML model according to the equipment vendor's implementation plan.

[0127] As shown in Figure 1, the authorization process for the aforementioned AI / ML model is primarily used to respond to requests from a model service consumer entity and deploy the AI / ML model to that entity. However, the inventors have discovered that as future business demands increase, a single business may require the collaboration of multiple AI / ML models, and these models may need to be deployed to different entities / network elements involved in the business. For example, if a business involves data transmission and processing between a terminal and access network equipment, and the terminal uses an AI / ML model to compress the data and sends the compressed data to the access network equipment, then the access network equipment needs to deploy another related AI / ML model to decompress the compressed data to obtain the original data. Therefore, how to achieve multi-entity deployment of multiple AI / ML models is a current research challenge.

[0128] To address the aforementioned technical problems, the embodiments of this application propose the following technical solutions.

[0129] The technical solutions in this application will now be described with reference to the accompanying drawings.

[0130] In the embodiments of this application, "instruction" can include direct and indirect instructions, as well as explicit and implicit instructions. The information indicated by a certain piece of information is called the information to be instructed. In the specific implementation process, there are many ways to instruct the information to be instructed, such as, but not limited to, directly instructing the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly instruct the information to be instructed by instructing other information, where there is a relationship between the other information and the information to be instructed. It can also instruct only a part of the information to be instructed, while the other parts are known or pre-agreed upon. For example, the instruction of specific information can be achieved by using a pre-agreed (e.g., protocol-defined) arrangement of various pieces of information, thereby reducing instruction overhead to some extent. At the same time, common parts of various pieces of information can be identified and uniformly indicated to reduce the instruction overhead caused by individually indicating the same information.

[0131] Furthermore, the specific indication method can also be any existing indication method, such as, but not limited to, the above-mentioned indication methods and their various combinations. Specific details of various indication methods can be found in existing technologies, and will not be repeated here. As described above, for example, when multiple pieces of information of the same type need to be indicated, the indication methods for different pieces of information may differ. In the specific implementation process, the required indication method can be selected according to specific needs. This application embodiment does not limit the selected indication method; therefore, the indication methods involved in this application embodiment should be understood to cover various methods that enable the party to be indicated to obtain the information to be indicated.

[0132] It should be understood that the information to be indicated can be sent as a whole or divided into multiple sub-information messages sent separately, and the sending period and / or timing of these sub-information messages can be the same or different. The specific sending method is not limited in this application embodiment. The sending period and / or timing of these sub-information messages can be predefined, for example, according to a protocol, or configured by the sending device by sending configuration information to the receiving device.

[0133] In this application, "sending information" can be understood as one device sending information to another device, or it can also be understood as one logical module within a device sending information to another logical module. For example, "network device sending information" can be understood as a network device sending information to another device (such as a terminal or other network device), or it can be understood as a network service producer in a network device sending information to a network service consumer in a network device.

[0134] In this application, "receiving information" can be understood as one device receiving information from another device, or it can also be understood as a logical module within a device receiving information from another logical module. For example, "network device receiving information" can be understood as a network device receiving information from another device (such as a terminal or other network device), or it can be understood as a network service producer in a network device receiving information from a network service consumer in a network device.

[0135] In this application, phrases such as "sending information to... (e.g., a terminal)" or related illustrations in the accompanying drawings can be understood as indicating that the destination of the information is a terminal. This can include sending information directly or indirectly to a terminal. Similarly, phrases such as "receiving information from... (e.g., a terminal)," "receiving information from... (e.g., a terminal)," or "receiving information sent by (e.g., a terminal)," or related illustrations in the accompanying drawings, can be understood as indicating that the source of the information is a terminal. This can include receiving information directly or indirectly from a terminal. Information may undergo necessary processing between the source and destination, such as format changes, but the destination can understand the valid information from the source. Similar expressions in this application can be interpreted similarly and will not be elaborated further here.

[0136] "Predefined" or "pre-configured" can be achieved by pre-saving corresponding codes, tables, or other means that can be used to indicate relevant information in the device. This application does not limit the specific implementation method. "Saving" can refer to saving in one or more memories. These memories can be separate installations or integrated into the encoder, decoder, processor, or communication device. Alternatively, some memories can be separately installed, while others are integrated into the decoder, processor, or communication device. The type of memory can be any form of storage medium, and this application does not limit this.

[0137] The “protocol” mentioned in the embodiments of this application may refer to a protocol family in the field of communication, a standard protocol with a similar protocol family frame structure, or a related protocol applied to future communication systems. The embodiments of this application do not specifically limit this.

[0138] In the embodiments of this application, descriptions such as "when," "under the circumstances," "if," and "if" all refer to the device making corresponding processing under certain objective circumstances, and are not limited to a specific time. They do not require the device to make a judgment action during implementation, nor do they imply any other limitations.

[0139] In the description of the embodiments of this application, unless otherwise stated, " / " indicates that the objects before and after are in an "or" relationship. For example, A / B can represent A or B. "And / or" in the embodiments of this application is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone, where A and B can be singular or plural. Furthermore, in the description of the embodiments of this application, unless otherwise stated, "multiple" refers to two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple. Additionally, to facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first," "second," etc., do not limit the quantity or order of execution, and that "first," "second," etc., are not necessarily different. Furthermore, in the embodiments of this application, words such as "exemplary" or "for example" are used to indicate that something is being used as an example, illustration, or description. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner for ease of understanding.

[0140] The network architecture and business scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0141] To facilitate understanding of the embodiments of this application, a communication system will be used as an example to describe in detail the communication system applicable to the embodiments of this application.

[0142] As shown in Figure 3, which is a schematic diagram of the architecture of a communication system, the communication system may include network service consuming entities and network service producing entities, and optionally, it may also include access network equipment (such as RAN) and / or terminals (such as UE).

[0143] The network service consumer entity can be an NFC (Network Functions Controlled Controller), specifically a RAN (Radio Router) or core network element (such as an AMF). The term "network service consumer entity" is an exemplary designation; other names are possible in future communication systems, without limitation. For a detailed explanation of the network service consumer entity, please refer to the above introduction to NFC, which will not be repeated here.

[0144] Network service production entities (NFPs) are used to provide AI / ML models and / or information about AI / ML models to implement services. For example, a NFP can be an NFp, specifically a core network element (such as NWDAF). The term "Network Service Production Entity" is an exemplary designation; other names are possible in future communication systems, without limitation. A detailed explanation of NFPs can be found in the above introduction and will not be repeated here.

[0145] In one specific architecture of the communication system, as shown in Figure 3(a), the network service consumer entity is AMF and the network service producer entity is NWDAF. AMF and NWDAF interact using a service-oriented interface. AMF can also communicate with RAN and / or UE. In some possible implementations, AMF requests AI / ML models from NWDAF on behalf of RAN and / or UE so that AI / ML models can be deployed on RAN and / or UE.

[0146] In one specific architecture of this communication system, as shown in Figure 3(b), the network service consumer entity is the RAN, and the network service producer entity is the RAN-NWDAF. RAN-NWDAF refers to an NWDAF that can communicate directly with the RAN through the defined service interface N2. RAN-NWDAF is an exemplary name; other names are possible in future communication systems, without specific limitations. Furthermore, the RAN-NWDAF and AMF can also interact through the SBI interface to provide services to other RANs in the network through the AMF.

[0147] In one specific architecture of this communication system, as shown in Figure 3(c), the network service consumer entity is RAN, and the network service producer entity is RAN-NWDAF. RAN-NWDAF can exclusively serve RAN, and the two communicate directly through a defined service interface.

[0148] In this communication system, a network service consumer entity can send a service request to a network service producer entity. This service requires interaction between at least two models deployed on different entities, enabling the network service producer entity to provide at least two AI / ML models to complete the service. In response to the request, the network service producer entity provides a first AI / ML model and a second AI / ML model to the network service consumer entity. If the first AI / ML model is deployed on the network and the second AI / ML model is deployed on the terminal, the first and second AI / ML models interact to complete the service. Compared to existing AI / ML model authorization processes, this system enables the authorization of multiple functionally related AI / ML models, adapting to more complex business scenarios.

[0149] The communication method and apparatus of this application embodiments will be further described below with reference to the accompanying drawings. It is understood that this application uses network service consuming entities and network service producing entities as examples to illustrate the execution of the interaction, but this application does not limit the execution entities of the interaction. The interaction process between various network elements / devices in the above-described communication system will be specifically described below through method embodiments. The communication method provided in this application embodiments can be applied to the above-described communication system and specifically applied to various scenarios mentioned in the above-described communication system, which will be described in detail below.

[0150] Figure 4 is a flowchart illustrating the communication method provided in an embodiment of this application. This communication method is applicable to the aforementioned communication system and mainly involves the interaction between network service consuming entities and network service producing entities.

[0151] As shown in Figure 4, the specific process of this method is as follows:

[0152] S401, the network service consumer entity sends a first message to the network service producer entity, requesting the network service producer entity to provide an AI / ML model for completing the business between the network and the terminal.

[0153] The first message can be used to request the provision of an AI / ML model to complete the business between the network and the terminal. The first message can be any possible signaling / message used in the interaction between the network service consumer entity and the network service producer entity. It can be a reused existing signal element or a newly defined signaling, such as named a model request message, or any other possible name, without any restrictions.

[0154] S402, the network service consumer entity receives a second message from the network service producer entity, the second message including information about the AI / ML model.

[0155] The second message can be used to provide information about the AI / ML model requested by the network service consumer entity. It can be a model notification message, or any other message that can be named as such, without limitation. In one example, the aforementioned AI / ML model includes a first AI / ML model and a second AI / ML model. The second message includes information about both the first and second AI / ML models. If the first AI / ML model is deployed to the network and the second AI / ML model is deployed to the terminal, the first and second AI / ML models interact to complete the aforementioned service.

[0156] The S401 will be described in detail below:

[0157] The first message may contain initial information. This initial information can be used to indicate a service, such as including an identifier for that service. This service may be one that requires an AI / ML model to complete, such as one that requires interaction between at least two AI / ML models deployed to the network and the terminal respectively. In one example, this service is a data compression and decompression service. The terminal needs an AI / ML model to compress the data for this service before sending it to the network, and the network needs an associated AI / ML model to decompress the received data.

[0158] Furthermore, the identifier for a service can be an analytics ID. When a service requires an AI / ML model to complete its task, the analytics ID can be understood as being related to the AI / ML model, or it can be understood as being related to the network service provider that delivers the AI / ML model. The analytics ID can be an existing information element or a newly defined identifier; there are no specific restrictions.

[0159] It is understandable that, since the business requires interaction between at least two AI / ML models deployed on the network and the terminal, the first message, by carrying the first information, can implicitly indicate that it is used to request the model to complete the business. Alternatively, the first message can also implicitly request the AI / ML model used to complete the above business through message name / type, or request the above model in other ways, without any specific restrictions.

[0160] Optionally, the first message may also include a vendor ID, which can be used to indicate the manufacturer of the access network equipment or core network element, ensuring the uniqueness and identifiability of the equipment or network element.

[0161] Optionally, the first message may also include a terminal identifier, which indicates the manufacturer type or specific device of the terminal. For example, the terminal identifier can be any one of a permanent equipment identifier (PEI), a subscription permanent identifier (SUPI), a subscription concealed identifier (SUCI), an international mobile equipment identity (IMEI), or a type allocation code. Among these, PEI is the terminal's factory identifier, SUPI / SUCI is the identifier assigned by the operator's network to the user using the terminal, IMEI is the terminal's device identification code, and TAC is a part of IMEI, which is a code that distinguishes the terminal manufacturer and model. Therefore, any one of PEI, SUPI / SUCI, or IMEI can be used to distinguish different terminals, and TAC can be used to distinguish terminals from different manufacturers / models.

[0162] The network service consumer entity can be referred to in the relevant description of the communication system shown in Figure 3. Optionally, the network service consumer entity can be an access network device or a core network element. The following further explains S401 based on the two cases where the network service consumer entity is an access network device or a core network element.

[0163] Scenario 1: The network service consumer is an access network device, such as a RAN.

[0164] A network service consumer entity can obtain the terminal's identifier from the terminal or the network. For example, the network service consumer entity receives a third message sent by the terminal, which contains the terminal's identifier. Alternatively, the network service consumer entity can send a request to the network (such as an AMF) to request the terminal's identifier from the AMF. Of course, this application embodiment does not limit how the AMF obtains the terminal's identifier in advance.

[0165] Optionally, the third message can also be used to request services between the network and the terminal, or to request AI / ML models to complete services between the network and the terminal. For example, the third message can request the aforementioned services or models through implicit or explicit indications in the message name / type, such as being named an AI capability support indication or a model acquisition request. Thus, upon receiving the third message, the network service consumer entity can determine the analysis ID based on the service or model indicated by the third message. It should be understood that the terminal can act as an AI / ML model consumer for a service, requesting model acquisition; that is, any AI / ML model consumer involved in the aforementioned service can request the AI / ML model used to complete the service through the network service consumer entity.

[0166] The network service consumer entity can encapsulate one or more of the following into a first message: the analysis ID, its own device vendor ID, and the terminal identifier, and send it to the network service producer entity. Communication between the network service consumer entity and the network service producer entity can be implemented based on any of the communication systems shown in Figure 3(a), Figure 3(b), or Figure 3(c), and will not be elaborated further.

[0167] Scenario 2: The network service consumer is a core network element, such as an access and mobility management function element (e.g., AMF).

[0168] A network service consumer entity can obtain one or more of the following information from the RAN: the RAN vendor ID, the terminal identifier, and the analytics ID. For example, the network service consumer entity receives a fourth message sent by the RAN, which contains one or more of the aforementioned information. Optionally, the fourth message may also include an identifier of the interface between the RAN and the UE (e.g., RAN_UE_NGAP_ID), used to uniquely identify the UE through the interface within the 5G base station, thereby identifying the terminal.

[0169] It should be understood that when the network service consumer entity obtains the analysis ID from the RAN, the AI / ML model implicitly requested by the analysis ID is actually a model required by the RAN to complete its business, and the network service consumer entity is requesting the AI / ML model on behalf of the RAN. The method by which the RAN obtains the aforementioned information can be referred to the relevant explanation in scenario 1 above where the network service consumer entity is an access network device, and will not be repeated here. Of course, the network service consumer entity can also obtain the terminal's identifier from itself, without restriction.

[0170] The network service consumer entity can encapsulate any one or more of the following into a first message: the analysis ID, its own equipment vendor ID, the RAN equipment vendor ID, and the terminal identifier, and send it to the network service producer entity through the service interface.

[0171] The S402 will be described in detail below:

[0172] Upon receiving the first message from the network service consumer entity, the network service producer can determine the AI / ML model required to complete the service based on the service indicated in the first message. For example, the network service producer entity may be associated with one or more analytics IDs, which can be understood as the network service producer entity providing the AI / ML model for the service corresponding to each of the one or more analytics IDs.

[0173] The network service production entity can identify the AI / ML model and the entity to which it needs to be deployed based on the service indicated by the first message and / or the information elements in the first message. For example, if the service indicated by the first message is data compression and decompression, the network service production entity can identify, based on the first message, the information of a first AI / ML model for decompressing data and the information of a second AI / ML model for compressing data, wherein the first AI / ML model is used for deployment to the RAN and the second AI / ML model is used for deployment to the terminal.

[0174] The information for the first AI / ML model may include the model file of the first AI / ML model and / or model information used to retrieve the first AI / ML model file (e.g., model identifier, model URL, etc.). The information for the second AI / ML model may include the model file of the second AI / ML model or the model address used to retrieve the second AI / ML model. For the specific process of retrieving the AI / ML model file based on the AI / ML model address, please refer to the aforementioned description of the model licensing process in the prior art, which will not be repeated here.

[0175] Optionally, the second message may further include at least one of the following: a terminal identifier list or an entity identifier list. The entity identifier list indicates identifiers of entities permitted to use the first AI / ML model. For example, the entity identifier list may be a list of entity vendor IDs to send the information of the first AI / ML model to entities whose vendor IDs belong to the entity identifier list, ensuring that the information of the first AI / ML model is sent to entities permitted to use the first AI / ML model. The terminal identifier list indicates identifiers of terminals permitted to use the second AI / ML model. For example, the terminal identifier list may be any one of an permitted TAC list, a permitted PEI list, a permitted SUPI / SUCI list, or a permitted IMEI list to send the information of the second AI / ML model to terminals whose terminal identifiers belong to the terminal identifier list, ensuring that the information of the second AI / ML model is sent to terminals permitted to use the second AI / ML model. If the second message does not contain a list of entity identifiers, or if the field of the entity identifier list is ALL, the first AI / ML model can be deployed to any entity in the network; if the second message does not contain a list of terminal identifiers, or if the field of the terminal identifier list is ALL, the second AI / ML model can be deployed to any terminal.

[0176] In some possible implementations, the network service producing entity can verify the network service consuming entity to determine whether the consuming entity has the authority to request the aforementioned service. If the verification is successful, a second message is sent to the consuming entity. In one example, the network service producing entity can send a second message to the consuming entity based on the consuming entity's identifier corresponding to a service authorization identifier. The service authorization identifier indicates the entity authorized to request the AI / ML model used to complete the aforementioned service. For example, the service authorization identifier can be an interoperability ID, which is a list of identifiers of network service consuming entities authorized to obtain the AI / ML model. Exemplarily, the identifier of the consuming entity can be a network element ID, a device vendor ID, or other IDs capable of identifying entity attributes, identity, or type. The service permission identifier corresponds to the service identifier. The mapping between service permission identifiers (such as interoperability IDs) and service identifiers (such as analytics IDs) can be pre-configured in the network service production entity. When the network service production entity receives the first message, it can obtain the corresponding service permission identifier based on the service identifier in the first message and verify whether the identifier of the network service consumption entity belongs to the service permission identifier. For example, verifying that the identifier of the network service consumption entity is in the interoperability ID indicates that the network service consumption entity can request the AI / ML model corresponding to the analytics ID. Depending on the type of network service consumption entity, the identifier of the network service consumption entity can be the vendor ID of the access network equipment (such as RAN) or the vendor ID of the core network element (such as AMF).

[0177] In conjunction with S401-S402 above, after the network service consumer entity receives the information from the first AI / ML model and the information from the second AI / ML model, the method may further include:

[0178] Step S1: The network service consumer entity deploys the first AI / ML model to itself, or sends the information of the first AI / ML model to one or more other entities in the network besides the network service consumer entity.

[0179] Step S2: The network service consumer entity sends the information of the second AI / ML model to the terminal, or to one or more other entities in the network other than the network service consumer entity.

[0180] Optionally, steps S1 and S2 can be two separate steps, or they can be combined into a single message. For example, when the network service consumer entity is a core network element (such as AMF), the AMF can first send the information of the first AI / ML model to one or more other entities in the network besides the AMF, and then send the information of the second AI / ML model to the terminal; the AMF can also encapsulate the information of the first AI / ML model and the information of the second AI / ML model into a single message and send it to one or more other entities in the network besides the network service consumer entity. When steps S1 and S2 are two separate steps, the sequence numbers of steps S1 and S2 are not used to restrict the execution order. Step S2 can be executed first, followed by step S1. For example, the AMF can also first send the information of the second AI / ML model to the terminal, and then send the information of the first AI / ML model to one or more other entities in the network besides the AMF, without restriction.

[0181] The following explanation will focus on two scenarios: network service consumption entities being access network equipment and core network elements, and will be further elaborated in conjunction with S402, steps S1 and S2.

[0182] Scenario 3: The network service consumer is an access network device, such as a RAN.

[0183] The network service producing entity can send a second message to the network service consuming entity based on the service authorization identifier corresponding to the identifier of the network service consuming entity (i.e., the RAN equipment vendor ID). Optionally, the second message includes information about the first AI / ML model and information about the second AI / ML model. Optionally, the second message also includes terminal identification information.

[0184] The network service consumer deploys the first AI / ML model locally based on the information from the first AI / ML model.

[0185] The network service consumer entity sends the information of the second AI / ML model to the terminal based on the terminal's identifier in the terminal identifier list, so that the terminal can deploy the second AI / ML model.

[0186] Scenario 4: The network service consumer is a core network element, such as an access and mobility management function element (e.g., AMF).

[0187] The network service producing entity can simultaneously verify that the equipment vendor ID of the network service consuming entity corresponds to the service permission identifier, and that the equipment vendor ID of the AI / ML model consumer (such as RAN) in the network corresponds to the service permission identifier, before sending the second message.

[0188] In this scenario, the network service producer can encapsulate the information from the first AI / ML model and the second AI / ML model into a second message and send it to the network service consumer. Optionally, the second message may also include a list of terminal identifiers. The network service consumer can then directly send the information from the first AI / ML model to the RAN whose equipment vendor ID has been verified, thereby enabling the first AI / ML model to be deployed to the RAN.

[0189] The network service producing entity may also choose not to verify the network service consuming entity, but only to verify the actual AI / ML model consumer (such as RAN) in the network. For example, the network service producing entity may send a second message to the network service consuming entity only after verifying that the RAN's equipment vendor ID corresponds to the service permission identifier. Optionally, the first and second AI / ML models in the second message are encrypted models, which the network service consuming entity cannot decrypt. Therefore, the network service consuming entity does not need to have the permission to request the model, and information security can still be guaranteed. In this case, the second message may optionally include information about the first AI / ML model, information about the second AI / ML model, and a list of entity identifiers. Optionally, the second message may also include the list of entity identifiers. The network service consuming entity can send the information of the first AI / ML model to the RAN based on the RAN's identifier corresponding to the list of entity identifiers.

[0190] The network service consumer entity sends the information of the second AI / ML model to the terminal based on the terminal's identifier in the terminal identifier list, so that the terminal can deploy the second AI / ML model.

[0191] Network service consuming entities can also send the information of the first AI / ML model, the information of the second AI / ML model, and the terminal identifier list to the RAN. The RAN will deploy the first AI / ML model to itself and send the information of the second AI / ML model to the terminal. For specific methods, please refer to the relevant description in Case 3 above, which will not be repeated here.

[0192] In conjunction with S401-S402 above, before the network service consumer entity sends the first message to the network service producer entity in case 4, the method may further include:

[0193] Step S3: The network service consumer entity sends an authorization token request message to the core network element (such as NRF) and receives an authorization token response message from the core network element.

[0194] The Authorization Token Request message is used to request an authorization token, which authorizes or allows a network service consumer entity to obtain a model from a network service producer entity. The Authorization Token Request message includes the RAN's equipment vendor ID. Upon receiving the Authorization Token Response message, the network service consumer entity sends a first message to the network service producer entity.

[0195] Optionally, the NRF can send an authorization token response message to the network service consumer entity, based on the fact that both the network service consumer entity's vendor ID and the RAN's vendor ID are in the network service producer entity's service permission identifier. In this case, after receiving the first message, the network service producer entity can send a second message without further verifying the identifiers of the network service consumer entity (e.g., AMF) or the AI / ML model consumer (e.g., RAN).

[0196] Optionally, the authorization token response message includes the RAN's equipment vendor ID. The NRF can send the authorization token response message to the network service consumer entity based on the RAN's equipment vendor ID in the service authorization identifier of the network service producer entity. In this case, after receiving the first message, the network service producer entity sends the second information based on the RAN's equipment vendor ID in the service authorization identifier.

[0197] The overall flow of the communication method provided in the embodiments of this application has been described above with reference to Figure 4. The flow of the communication method provided in the embodiments of this application in specific scenarios will be described below with reference to Figures 5-7.

[0198] Figure 5 is a schematic flowchart of the communication method provided in this application embodiment. This communication method is applicable to the aforementioned communication system and mainly involves the interaction between the RAN (e.g., a network service consumer entity), the RAN-NWDAF (e.g., a network production consumer entity), the UE (e.g., a terminal), and the AMF (e.g., a network). For example, the RAN can send a first message to the RAN-NWDAF, requesting the RAN-NWDAF to provide an AI / ML model for completing services between the network and the terminal, and receive a second message from the RAN-NWDAF. The second message includes at least information about the first AI / ML model and information about the second AI / ML model. The first AI / ML model is used for deployment to the RAN, and the second AI / ML model is used for deployment to the UE.

[0199] Specifically, as shown in Figure 5, the flow of this communication method is as follows:

[0200] S501, RAN sends a RAN information registration message to RAN-NWDAF.

[0201] RAN information registration messages are used to register or configure RAN device information on the NWDAF. Specifically, the message may include the RAN ID, RAN vendor ID, serving cell, and tracking area identity (TAI). The specific registration process is shown in Figure 2 and will not be elaborated further.

[0202] This step is understandable as it is optional. The RAN ID can also be pre-configured in the NWDAF.

[0203] S502, AMF sends a UE identification message to RAN.

[0204] The UE identification message is used to enable the RAN to obtain the UE's identification in advance. For details, please refer to the relevant introduction of S401 above, which will not be repeated here.

[0205] It is understood that this step is optional. The RAN may also request the UE identifier from the UE, or obtain the UE identifier through other means.

[0206] S503, the UE sends an AI / ML model acquisition request message #1 to the RAN.

[0207] Optionally, the AI / ML model acquisition request message #1 may include the UE's identifier. For details, please refer to the relevant introduction of S401 and Case 1 above, which will not be repeated here.

[0208] S504, RAN sends AI / ML model retrieval request message #2 to RAN-NWDAF.

[0209] AI / ML Model Acquisition Request Message #2 is used to request an AI / ML model for completing services between the network and the terminal, and includes an analysis ID. Optionally, AI / ML Model Acquisition Request Message #2 may also include the RAN's equipment vendor ID and / or UE identifier(s). For details, please refer to the relevant description in S401 above, which will not be repeated here.

[0210] S505, RAN-NWDAF determines that the RAN has the authority to access the model based on the RAN's equipment vendor ID being located in the interoperability ID corresponding to the analysis ID, and determines the AI / ML model information based on the analysis ID.

[0211] Specifically, RAN-NWDAF determines that RAN has the authority to obtain information about the first AI / ML model and / or the second AI / ML model, and determines the information about the first AI / ML model, the information about the second AI / ML model, and the list of terminal identifiers corresponding to the second AI / ML model.

[0212] S506, RAN-NWDAF sends AI / ML model notification messages to RAN.

[0213] The AI / ML model notification message includes information about the first AI / ML model, information about the second AI / ML model, and a list of terminal identifiers corresponding to the second AI / ML model.

[0214] It is understandable that the specific implementation of S505-S506 can be referred to the relevant introduction of case 3 in step S402 above, and will not be repeated here.

[0215] S507, the UE sends an AI / ML model acquisition request message #1 to the RAN.

[0216] Other UEs may send AI / ML model acquisition request messages #1 to the RAN. If the RAN determines that the UE needs to use a local AI / ML model (such as a model pre-configured in the RAN or a model obtained from NWDAF), then the subsequent steps will continue. For specific implementation details, please refer to S503, which will not be elaborated further.

[0217] It is understood that step S507 is an optional step.

[0218] S508, RAN determines whether it is necessary to obtain the UE identifier.

[0219] The RAN determines whether it needs to obtain the UE identifier based on the information of the second AI / ML model and / or the corresponding terminal list identifier. The RAN may obtain the UE identifier in the following ways: the RAN obtains the UE identifier stored locally [obtained in step S502], or it may execute steps S509-S510 to obtain the UE identifier from the AMF, or it may execute steps S501-S512 to obtain the UE identifier from the UE.

[0220] S509, RAN sends UE identifier acquisition request message #1 to AMF.

[0221] S510, AMF sends UE identifier acquisition response message #1 to RAN.

[0222] S511, the RAN sends a UE identifier acquisition request message #2 to the UE.

[0223] S512, the UE sends a UE identifier acquisition response message #2 to the RAN.

[0224] It is understood that steps S509-S510 and steps S511-S512 are optional steps.

[0225] Optionally, if the second message received by the RAN does not contain a terminal identifier list, or if the field of the terminal identifier list is ALL, the RAN does not need to obtain the UE.

[0226] When the RAN needs to obtain the UE, it can obtain the terminal identifier stored locally through step S502, or through steps S509-S510, or through steps S511-S512. For details, please refer to the relevant introduction of case 1 in S401 above, which will not be repeated here.

[0227] S513, RAN verifies whether the UE identifier is in the terminal identifier list.

[0228] S514, the RAN sends a second AI / ML model delivery message to the UE.

[0229] The UE identifier being in the terminal identifier list indicates that the UE is allowed to use the second AI / ML model. For details on steps S513-S514, please refer to the relevant description of case 3 in S402 above, which will not be repeated here.

[0230] Figure 6 is a flowchart illustrating the communication method provided in this embodiment. This communication method is applicable to the aforementioned communication system and mainly involves interactions between the AMF (e.g., a network service consumer entity), the NWDAF (e.g., a network producer consumer entity), the RAN (e.g., a network), the UE (e.g., a terminal), and the NRF (e.g., a core network element). For example, the AMF can send a first message to the NWDAF, requesting the network service producer entity to provide an AI / ML model for completing services between the network and the terminal, and receive a second message from the RAN-NWDAF. The second message includes at least information about the first AI / ML model and information about the second AI / ML model. The first AI / ML model is used for deployment to the RAN, and the second AI / ML model is used for deployment to the UE.

[0231] Specifically, as shown in Figure 6, the communication method flow is as follows:

[0232] S601: AMF sends the UE identifier to RAN.

[0233] The UE identification message is used to enable the RAN to obtain the UE's identification in advance. For details, please refer to the relevant introduction of S401 above, which will not be repeated here.

[0234] It is understandable that this step is optional.

[0235] S602: The UE sends an AI / ML model acquisition request message #1 to the RAN.

[0236] Optionally, the AI / ML model acquisition request message #1 may include the UE's identifier. For details, please refer to the relevant introduction of S401 and Case 1 above, which will not be repeated here.

[0237] S603: RAN sends AI / ML model retrieval request message #2 to AMF.

[0238] AI / ML model retrieval request message #2 may include an analysis ID, allowing the AMF to request an AI / ML model on behalf of the RAN. For details, please refer to the relevant description in S401, Case 2, which will not be repeated here.

[0239] S604: The AMF sends a token acquisition request message to the NRF.

[0240] S605: The NRF sends an authorization token based on the interoperability ID of the RAN's equipment vendor ID located in the NFp.

[0241] S606: NRF sends a token to AMF to obtain a response message.

[0242] The token acquisition request message includes the RAN device vendor ID or the RAN instance ID, which is used to uniquely distinguish different RAN devices.

[0243] Optionally, the token acquisition response message includes the RAN equipment vendor ID.

[0244] Optionally, in step S605, the NRF can send an authorization token based on the fact that the NFc is an AMF and the RAN's equipment vendor ID is located in the interoperability ID of the NFp, or it can send an authorization token based on the fact that both the NFc's equipment vendor ID and the RAN's equipment vendor ID are located in the interoperability ID of the NFp.

[0245] The specific implementation methods for steps S604-S606 can be found in the description of step S3 above, and will not be repeated here.

[0246] S607: AMF sends AI / ML model retrieval request message #3 to NWDAF.

[0247] AI / ML Model Acquisition Request Message #3 is used to request an AI / ML model for completing services between the network and the terminal, and includes an analysis ID. Optionally, AI / ML Model Acquisition Request Message #3 may also include the RAN's equipment vendor ID and / or the UE identifier. For details, please refer to the relevant description in S401 above, which will not be repeated here.

[0248] S608: NWDAF determines the AI / ML model information based on the AMF's equipment vendor ID and the RAN's equipment vendor ID, which are located in the interoperability ID corresponding to the analysis ID.

[0249] Specifically, NWDAF determines that AMF and RAN have the authority to obtain the model based on the fact that the AMF's equipment vendor ID and RAN's equipment vendor ID are located in the interoperability ID corresponding to the analysis ID. It also determines the information of the first AI / ML model, the information of the second AI / ML model, and the terminal identifier list corresponding to the second AI / ML model based on the analysis ID.

[0250] Alternatively, NWDAF can also determine whether the RAN has permission to access the model based on the fact that the RAN's equipment vendor ID is located in the interoperability ID corresponding to the analysis ID.

[0251] Optionally, NWDAF can determine the information of the first AI / ML model, the information of the second AI / ML model, the list of entity identifiers corresponding to the first AI / ML model, and the list of terminal identifiers corresponding to the second AI / ML model based on the analysis ID.

[0252] For details, please refer to the relevant introduction of S402 case 4 above, which will not be repeated here.

[0253] S609: NWDAF sends AI / ML model notification message #1 to AMF.

[0254] AI / ML model notification message #1 includes information about the first AI / ML model, information about the second AI / ML model, and a list of terminal identifiers. For details, please refer to the relevant introduction of situation 4 in S402 above, which will not be repeated here.

[0255] S610: AMF sends AI / ML model notification message #2 to RAN.

[0256] AI / ML model notification message #2 includes information about the first AI / ML model, information about the second AI / ML model, and a list of terminal identifiers. For details, please refer to the relevant introduction in S402 above for situation 4, which will not be repeated here.

[0257] S611: RAN verifies whether the UE identifier is in the terminal identifier list.

[0258] S612: The RAN sends a second AI / ML model delivery message to the UE.

[0259] The specific implementation methods of steps S611-S612 can be referred to the relevant introduction of case 3 in S402 above, and will not be repeated here.

[0260] Figure 7 is a flowchart illustrating the communication method provided in this embodiment. This communication method is applicable to the aforementioned communication system and mainly involves interactions between the AMF (e.g., a network service consumer entity), the NWDAF (e.g., a network producer consumer entity), the RAN (e.g., a network), the UE (e.g., a terminal), and the NRF (e.g., a core network element). For example, the AMF can send a first message to the NWDAF, requesting the network service producer entity to provide an AI / ML model for completing services between the network and the terminal, and receive a second message from the RAN-NWDAF. The second message includes at least information about the first AI / ML model and information about the second AI / ML model. The first AI / ML model is used for deployment to the RAN, and the second AI / ML model is used for deployment to the UE.

[0261] Specifically, as shown in Figure 7, the communication method flow is as follows:

[0262] S701: AMF sends the UE identifier to RAN.

[0263] The UE identification message is used to enable the RAN to obtain the UE's identification in advance. For details, please refer to the relevant introduction of S401 above, which will not be repeated here.

[0264] S702: The UE sends an AI / ML model acquisition request message #1 to the RAN.

[0265] Optionally, the AI / ML model acquisition request message #1 may include the UE's identifier. For details, please refer to the relevant introduction of S401 and Case 1 above, which will not be repeated here.

[0266] S703: RAN sends AI / ML model retrieval request message #2 to AMF.

[0267] AI / ML model acquisition request message #2 may include the analysis ID and UE identifier, enabling the AMF to request the AI / ML model on behalf of the RAN, and subsequently send a second AI / ML model delivery message to the UE based on the UE identifier. For details, please refer to the relevant description in S401, Case 2, which will not be repeated here.

[0268] S704: The AMF sends a token acquisition request message to the NRF.

[0269] S705: The NRF sends an authorization token based on the interoperability ID in the NFp containing the RAN's equipment vendor ID.

[0270] S706: NRF sends a token to AMF to obtain a response message.

[0271] The token acquisition request message includes the RAN device vendor ID or the RAN instance ID, which is used to uniquely distinguish different RAN devices.

[0272] Optionally, in step S705, the NRF can send an authorization token based on the fact that the NFc is an AMF and the RAN's vendor ID is located in the interoperability ID of the NFp, or it can send an authorization token based on the fact that both the NFc's vendor ID and the RAN's vendor ID are located in the interoperability ID of the NFp.

[0273] The specific implementation methods for steps S704-S706 can be found in the description of step S3 above, and will not be repeated here.

[0274] S707: AMF sends AI / ML model retrieval request message #3 to NWDAF.

[0275] AI / ML Model Acquisition Request Message #3 is used to request an AI / ML model for completing services between the network and the terminal, and includes an analysis ID. Optionally, AI / ML Model Acquisition Request Message #3 may also include the RAN's equipment vendor ID and / or the UE identifier. For details, please refer to the relevant description in S401 above, which will not be repeated here.

[0276] S708: NWDAF determines the information of the first AI / ML model, the information of the second AI / ML model, the list of entity identifiers corresponding to the first AI / ML model, and the list of terminal identifiers corresponding to the second AI / ML model based on the analysis ID.

[0277] For details, please refer to the relevant introduction of S402 case 4 above, which will not be repeated here.

[0278] S709: NWDAF sends AI / ML model notification message #1 to AMF.

[0279] The AI / ML model notification message #1 includes information about the first AI / ML model, information about the second AI / ML model, a list of entity identifiers, and a list of terminal identifiers. For details, please refer to the relevant introduction of case 4 in S402 above, which will not be repeated here.

[0280] S710, AMF verifies whether the RAN equipment vendor ID is in the entity identifier list.

[0281] S711: AMF sends AI / ML model notification message #2 to RAN.

[0282] AI / ML model notification message #2 includes information about the first AI / ML model.

[0283] For details on steps S710-S711, please refer to the relevant description of case 4 in S402 above, which will not be repeated here.

[0284] S712, AMF verifies whether the UE identifier is in the terminal identifier list.

[0285] S713: AMF sends a second AI / ML model delivery message to UE.

[0286] The message sent by the second AI / ML model includes information about the second AI / ML model.

[0287] The specific implementation methods of steps S712-S713 can be found in the relevant description of case 3 in S402 above, and will not be repeated here.

[0288] The communication method provided by the embodiments of this application has been described in detail above with reference to Figures 4-7. The communication apparatus used to perform the communication method provided by the embodiments of this application is described in detail below with reference to Figures 8-9.

[0289] Figure 8 is a schematic diagram of the structure of a communication device provided in an embodiment of this application. As exemplarily shown in Figure 8, the communication device 800 includes a transceiver module 801 and a processing module 802. For ease of explanation, Figure 8 only shows the main components of the communication device.

[0290] The communication device 800 can be applied to the communication methods shown in Figures 4-7 to achieve the corresponding functions. For example, the transceiver module 801 can be used to implement the transceiver function in the communication methods shown in Figures 4-7, and the processing module 802 can be used to implement other functions in the communication methods shown in Figures 4-7 besides the transceiver function.

[0291] Optionally, the transceiver module 801 may include a transmitting module (not shown in FIG8) and a receiving module (not shown in FIG8). The transmitting module is used to implement the transmitting function of the communication device 800, and the receiving module is used to implement the receiving function of the communication device 800.

[0292] Optionally, the communication device 800 may further include a storage module (not shown in FIG8) that stores programs or instructions. When the processing module 802 executes the program or instructions, the communication device 800 can perform the functions in the methods shown in FIG4-FIG7.

[0293] It is understood that the communication device 800 may be a network device, or a chip (system) or other component or assembly that can be set in the network device, or a device that includes the network device. This application does not limit this.

[0294] Furthermore, the technical effects of the communication device 800 can be referenced from the technical effects of the communication method described above, and will not be repeated here.

[0295] Figure 9 is a second schematic diagram of the structure of the communication device provided in an embodiment of this application. Exemplarily, the communication device can be a terminal, or a chip (system) or other component or assembly that can be disposed in the terminal. As shown in Figure 9, the communication device 900 may include a processor 901. Optionally, the communication device 900 may further include a memory 902 and / or a transceiver 903. The processor 901 is coupled to the memory 902 and the transceiver 903, for example, they can be connected via a communication bus.

[0296] The following section, with reference to Figure 9, provides a detailed description of each component of the communication device 900:

[0297] The processor 901 is the control center of the communication device 900. It can be a single processor or a collective term for multiple processing elements. For example, the processor 901 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement the embodiments of this application, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0298] Optionally, the processor 901 can perform various functions of the communication device 900 by running or executing software programs stored in the memory 902 and calling data stored in the memory 902, such as performing the communication methods shown in Figures 4-7 above.

[0299] In a specific implementation, as one example, processor 901 may include one or more CPUs, such as CPU0 and CPU1 shown in FIG9.

[0300] In a specific implementation, as one embodiment, the communication device 900 may also include multiple processors, such as processors 901 and 904 shown in FIG. 9. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0301] The memory 902 is used to store the software program that executes the solution of this application, and is controlled by the processor 901 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0302] Optionally, the memory 902 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 902 may be integrated with the processor 901 or may exist independently and be coupled to the processor 901 through the interface circuit of the communication device 900 (not shown in FIG. 9). This application embodiment does not specifically limit this.

[0303] Transceiver 903 is used for communication with other communication devices. For example, if communication device 900 is a terminal, transceiver 903 can be used to communicate with a network device or with another terminal device. As another example, if communication device 900 is a network device, transceiver 903 can be used to communicate with a terminal or with another network device.

[0304] Optionally, transceiver 903 may include a receiver and a transmitter (not shown separately in Figure 9). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0305] Optionally, the transceiver 903 can be integrated with the processor 901 or exist independently and be coupled to the processor 901 through the interface circuit of the communication device 900 (not shown in FIG9). This application embodiment does not specifically limit this.

[0306] It is understood that the structure of the communication device 900 shown in Figure 9 does not constitute a limitation on the communication device. The actual communication device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0307] Furthermore, the technical effects of the communication device 900 can be referred to the technical effects of the method described in the above method embodiments, and will not be repeated here.

[0308] It should be understood that the processor in the embodiments of this application can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0309] It should also be understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0310] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0311] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0312] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0313] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0314] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0315] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0316] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0317] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0318] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A communication method characterized by comprising: Applied to network service consuming entities, the method includes: Send a first message to the network service production entity, the first message being used to request the network service production entity to provide an artificial intelligence (AI) / machine learning (ML) model for completing the business between the network and the terminal; The system receives a second message from the network service production entity. The second message includes information about the AI / ML model, which includes a first AI / ML model and a second AI / ML model. When the first AI / ML model is deployed to the network and the second AI / ML model is deployed to the terminal, the first AI / ML model and the second AI / ML model are used to complete the service through interaction.

2. The method of claim 1, wherein, The first message includes first information, which is used to indicate the service.

3. The method of claim 2, wherein, The first information includes the identifier of the service.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: In response to the second message, information about the second AI / ML model is sent to the terminal.

5. The method of claim 4, wherein, Sending the information of the second AI / ML model to the terminal includes: If the terminal is a terminal that allows the use of the second AI / ML model, the information of the second AI / ML model is sent to the terminal.

6. The method according to claim 4 or 5, characterized in that, The second message also includes a terminal identifier list, which indicates the identifiers of terminals that are permitted to use the second AI / ML model.

7. The method of claim 6, wherein, The method further includes: Obtain the identifier of the terminal; The identifier of the terminal is confirmed to be in the terminal identifier list.

8. The method of claim 5, wherein, The terminal that can use the second AI / ML model is any terminal.

9. The method according to any one of claims 1 to 8, characterized in that, The first message also includes the identifier of the terminal.

10. The method according to claim 6 or 9, characterized in that, The terminal is identified by any one of the following: a permanent device identifier, a model approval code, or a user permanent identifier.

11. The method according to any one of claims 1 to 3, characterized in that, The method further includes: In response to the second message, information about the first AI / ML model is sent to one or more other entities in the network besides the network service consumer entity.

12. The method of claim 11, wherein, Sending the information of the first AI / ML model to one or more other entities in the network besides the network service consumer entity includes: If the other entity is an entity that is permitted to use the first AI / ML model, the information of the first AI / ML model is sent to the other entity.

13. The method according to claim 11 or 12, characterized in that, The second message also includes an entity identifier list, which indicates the identifiers of entities that are permitted to use the first AI / ML model.

14. The method according to any one of claims 10 to 13, characterized in that, The network service consumption entity is the access and mobility management network element, and the other entities are the access network devices accessed by the terminal.

15. The method according to any one of claims 1-3, characterized in that, The network service consumer entity is an entity in the network, and the method further includes: In response to the second message, the first AI / ML model is deployed locally to the network service consumer entity.

16. The method of claim 15, wherein, The network service consumer entity is the access network device accessed by the terminal.

17. A method of communication, comprising: Applied to network service production entities, the method includes: Receive a first message from a network service consumer entity, the first message being used to request the network service producer entity to provide an artificial intelligence (AI) / machine learning (ML) model for completing the business between the network and the terminal; A second message is sent to the network service consumer entity. The second message includes information about the AI / ML model, which includes a first AI / ML model and a second AI / ML model. When the first AI / ML model is deployed to the network and the second AI / ML model is deployed to the terminal, the first AI / ML model and the second AI / ML model are used to complete the service through interaction.

18. The method of claim 17, wherein, The first message further includes first information, which is used to indicate the service, and the method further includes: Based on the first information, obtain information about the first AI / ML model and information about the second AI / ML model.

19. The method of claim 17, wherein, Sending the second message to the network service consumer entity includes: Based on the business permission identifier corresponding to the identifier of the network service consumer entity, a second message is sent to the network service consumer entity, wherein the business permission identifier is used to indicate the entity that is allowed to request the AI / ML model corresponding to the business.

20. The method of claim 19, wherein, The first message also includes the identifier of the network service consuming entity.

21. The method of claim 17, wherein, The second message further includes at least one of the following: a terminal identifier list or an entity identifier list, wherein the terminal identifier list indicates identifiers of terminals that are allowed to use the second AI / ML model, and the entity identifier list indicates identifiers of entities that are allowed to use the first AI / ML model.

22. The method according to any one of claims 17-21, characterized by, The entity that can use the first AI / ML model is the access network device.

23. A communications device, characterized by The apparatus includes a module for performing the method as described in any one of claims 1-22.

24. A communications device, characterized by The communication device includes a processor and a memory; the memory is used to store computer instructions, which, when executed by the processor, cause the communication device to perform the method as described in any one of claims 1-22.

25. A computer readable storage medium, characterized in that, The computer-readable storage medium includes a computer program or instructions that, when executed, cause the method as described in any one of claims 1-22 to be performed.

26. A computer program product, characterised in that, It includes a computer program or instructions that, when run, cause the method as described in any one of claims 1-22 to be performed.

Images