Communication method and communication device

Abstract

Provided are a communication method and a communication device, the method comprising: a first network element determining a first entity and a first identity, and sending a first request message to the first entity, the first request message being used for requesting the first entity to execute a first federated learning task with the first identity, the first identity is used for indicating the identity of the first entity in executing the first federated learning task. According to the method, the first identity is sent to the first entity, so that the first entity can effectively execute the first federated learning task with the first identity, and the execution efficiency of the first federated learning task is improved.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of communication technologies, and more particularly, to a communication method and a communication device. Background Art

[0002] Federated Learning (FL) enables multiple participants to interact model parameters through a security mechanism without exchanging training data, thereby achieving a collaborative training effect. Among them, federated learning includes horizontal federated learning and vertical federated learning. Among them, horizontal federated learning is also known as feature-aligned federated learning, that is, the data features of the participants are aligned, and horizontal federated learning can increase the total amount of training samples. Vertical federated learning is to perform federated learning on different data features of the common samples of multiple participants, that is, the training data of each participant is vertically partitioned. Vertical federated learning is also known as sample-aligned federated learning, that is, the training samples of the participants are aligned, and vertical federated learning can increase the dimensionality of training data features. That is, federated learning can effectively help multiple institutions to perform data usage and learning modeling while meeting the requirements of user privacy protection, data security, and government regulations.

[0003] However, in the current mobile network system, such as the 5th-generation core (5GC) network, the creation and execution of vertical federated learning tasks are not supported. Therefore, how to effectively execute vertical federated learning tasks is a problem that needs to be considered currently. Summary of the Invention

[0004] The present application provides a communication method and a communication device, which can effectively execute vertical federated learning tasks.

[0005] In a first aspect, a communication method is provided. This method can be executed by a first network element (for example, a task initiator), or can also be executed by a chip, circuit, or logic module of the first network element. The present application does not make any limitation in this regard. For the sake of description, the following will take the execution by the first network element as an example for illustration.

[0006] The method includes: the first network element determines a first entity and a first identity, where the first identity is used to indicate the identity of the first entity in performing a first federated learning task; the first network element sends a first request message to the first entity, and the first request message is used to request the first entity to perform the first federated learning task with the first identity.

[0007] Among them, the first network element supports initiating a first federated learning task. Or rather, the first network element has the ability to initiate a first federated learning task. For example, the first network element can be a coordinator, NWDAF, AF, etc.

[0008] Optionally, the first request message includes a first identity.

[0009] That is to say, the first entity is an entity that has the ability to execute the first federated learning task with the first identity.

[0010] Optionally, the first federated learning task can be replaced with other names such as: first federated learning, or first federated learning function, or first federated learning process, or first federated learning activity, etc. Here, the entity that supports executing the first federated learning task can be understood as: an entity (or network element or device or node, etc.) that has the ability to execute the first federated learning task. Or rather, the entity supports executing the first federated learning task (or activity, or process, etc.).

[0011] Exemplarily, the first federated learning task can include one or more of the following: network performance analysis, artificial intelligence (AI) model training, face recognition, or cross-institutional medical data analysis and disease prediction, etc. Optionally, the federated learning task in the embodiments of the present application can refer to a vertical federated learning task, and the specific interpretation can refer to the relevant description above.

[0012] According to the above solution, after determining the first entity and the first identity, the first network element makes the first entity effectively execute the first federated learning task with the first identity by sending the first identity to the first entity, thereby improving the execution efficiency of the first federated learning task.

[0013] In some implementation manners, the first identity includes a main participant or a subordinate participant. Among them, the main participant supports providing labels in the first federated learning task, and the labels correspond to the first federated learning task; or, the main participant supports providing labels or, labels and data in the first federated learning task; the subordinate participant supports providing data in the first federated learning task.

[0014] That is to say, the first entity supports providing labels and / or data in the execution of the first federated learning task.

[0015] Among them, the labels are used for the first federated learning task. For example, they are used for the training and / or model evaluation of the model corresponding to the first federated task; the data is used for the first federated learning task. For example, it is used for the analysis, model training and / or inference corresponding to the first federated task. The data can refer to the samples used by the first entity in the execution of the first federated learning task. For example, the traffic data of the terminal on the AF, or the service data of the terminal on the AF, etc.

[0016] In some implementations, the method further includes: a first network element obtaining at least one candidate entity and identity information of the at least one candidate entity, where the identity information of the at least one candidate entity is used to indicate the identities supported by the at least one candidate entity in a first federated learning task; the first network element determining a first entity and a first identity, including: the first network element determining the first entity and the first identity from the at least one candidate entity and the identity information of the at least one candidate entity.

[0017] Based on the above solution, the first network element can select or determine the first entity and the first identity from the obtained at least one candidate entity and the identity information of the at least one candidate entity, that is, find the first entity and the first identity suitable for executing the first federated learning task, so that the first entity can effectively execute the first federated learning task with the first identity subsequently, improving the execution efficiency of federated learning.

[0018] In some implementations, the first network element obtaining at least one candidate entity and identity information of the at least one candidate entity includes: the first network element receiving at least one second entity and identity information of the at least one second entity from a second network element, where the second network element supports discovering entities for executing the first federated learning task, or in other words, the second entity supports providing information about entities capable of executing the first federated learning task, and among them, the at least one second entity supports executing the first federated learning task, and the at least one second entity includes at least one candidate entity.

[0019] In some implementations, the first network element receiving at least one second entity and identity information of the at least one second entity from the second network element includes: the first network element sending a third request message to the second network element, where the third request message is used to obtain entities that support executing the first federated learning task; the first network element receiving a third response message from the second network element, where the third response message includes at least one candidate entity and identity information of the at least one candidate entity.

[0020] Among them, an entity that supports executing the first federated learning task can be understood as: the entity has the ability to execute the first federated learning task.

[0021] Optionally, the third request message includes one or more of the following: multiple types, a first analysis identifier, a first group identifier, information about a vertical federation, or a first interoperability identifier, etc., where the type refers to the type of the entity, such as the AF type and / or the NF type.

[0022] Based on the above solution, the first network element can obtain at least one candidate entity and the identity information of at least one candidate entity from a second network element (e.g., an entity discovery function network element). Among them, the at least one candidate entity can be an AF instance and / or an NF instance determined by the second network element according to multiple types, or can be an AF instance and / or an NF instance determined by the second network element according to at least one of a first set of identifiers, a first analysis identifier, or a first interoperability identifier.

[0023] In some implementation manners, before the first network element determines the first entity and the first identity, the method further includes: the first network element sends a second request message to at least one candidate entity, where the second request message is used to request the at least one candidate entity to prepare to execute a first federated learning task; the first network element receives a second response message from the at least one candidate entity, where the second response message is used to indicate that the at least one candidate entity agrees to support executing the first federated learning task with the first identity.

[0024] Exemplarily, if the at least one candidate entity includes the first entity, it means that the first network element sends a second request message to the first entity, and the second request message is used to request the first entity to prepare to execute the first federated learning task; the first network element receives a second response message from the first entity, and the second response message is used to indicate that the first entity agrees or supports executing the first federated learning task with the first identity.

[0025] In some implementation manners, the second request message includes at least one identity supported by the first entity in the first federated learning task. The at least one identity supported by the first entity in the first federated learning task includes the first identity, and the supported at least one identity includes a main participant and / or a subordinate participant.

[0026] Optionally, the second request message includes the first identity, where the first identity refers to the identity that the first network element requests the first entity to support, or in other words, requests the first entity to execute the subsequent first federated learning task with the first identity. The identity requested for the first entity to support can be a main participant and / or a subordinate participant.

[0027] In some implementation manners, the second response message includes the first identity.

[0028] Based on the above solution, in the case that the first network element determines that the first entity agrees to execute the first federated learning task with the first identity, the first network element finally determines the participants and their identity information for executing the first federated learning task, which is convenient for improving the execution efficiency of the first federated learning task.

[0029] In some implementation manners, the second request message further includes information about a third network element, and the third network element supports coordinating the first entity to execute the first federated learning task.

[0030] In some implementations, the method further includes: a first network element obtaining information of a third network element and identity information of the third network element, where the identity information of the third network element is used to indicate that the third network element supports coordinating a first entity to execute a first federated learning task, or in other words, the identity information of the third network element is used to indicate that the third network element has the ability to coordinate the first entity to execute the first federated learning task.

[0031] In some implementations, the first network element obtaining the information of the third network element and the identity information of the third network element includes: the first network element receiving the information of the third network element and the identity information of the third network element from a second network element.

[0032] Optionally, the first network element obtaining the information of the third network element and the identity information of the third network element may be signaling configured or pre-configured.

[0033] In some implementations, the method further includes: the first network element sending a fifth request message to the third network element, where the fifth request message is used to request the third network element to coordinate the first entity to execute the first federated learning task; the first network element receiving a fifth response message from the third network element, where the fifth response message is used to indicate that the third network element agrees to coordinate the first entity to execute the first federated learning task.

[0034] Based on the above solution, through information interaction between the first network element and the third network element, it is determined that the third network element coordinates the first entity to execute the first federated learning task in the role of a coordinator, which is convenient for improving the execution efficiency of the subsequent first federated learning task.

[0035] In some implementations, before the first network element sends a first request message to the first entity, the method further includes: the first network element sending a sixth request message to the third network element, where the sixth request message is used to request to create or coordinate the first federated learning task, and the sixth request message includes the first entity and a first identity.

[0036] Based on the above solution, the first network element sends the first entity participating in the execution of the first federated learning task and the first identity to the third network element, that is, clarifies the identity of each first entity in the execution of the first federated learning task, such as providing labels and / or data, which is convenient for the third network element to coordinate or organize the first entities to participate in the execution of the first federated learning task with the first identity, effectively improving the execution efficiency of federated learning.

[0037] In some implementations, before the first network element determines the first entity and the first identity, the method further includes: the first network element obtaining a first analysis identifier corresponding to the first federated learning task; the first network element determining to trigger the first federated learning task according to the first analysis identifier.

[0038] Based on the above solution, the first network element can determine to initiate the first federated learning task according to the first analysis identifier, and then determine the first entity that executes the first federated learning task and the identity of the first entity when executing the first federated learning task.

[0039] Exemplarily, the way for the first network element to obtain the first analysis identifier can be: the first network element can receive a subscription request message from a consumer (for example, an NWDAF including AnLF or an NWDAF including MTLF), where the subscription request message includes the first analysis identifier, and the subscription request message is used to subscribe to the federated learning model provisioning or training (such as subscription request for MLmodel provisioning / training). Among them, the subscription request message can be a model subscription request message or an analysis subscription request message.

[0040] Exemplarily, the first analysis identifier can be predefined or preconfigured. Among them, predefined can include predefined in advance, such as protocol definition, and preconfiguration can be implemented by pre-saving corresponding codes, tables, strings or other means for indicating the first analysis identifier in the first network element. The present application does not limit its specific implementation manner.

[0041] It should be understood that the first analysis identifier can be used to indicate a certain specific function or service, and the specific function or service is related to the model, that is, the model can be used to execute the specific function or service, or in other words, the model supports the specific function or service corresponding to the first analysis identifier. Among them, the specific function or service can be network performance analysis, etc. The first analysis identifier corresponds to the first federated learning task, and it can be understood that: the first analysis identifier is used to indicate the first federated learning task.

[0042] Optionally, the first network element can also obtain other information and determine to trigger the first federated learning task according to the correspondence between the other information and the first federated learning task. Exemplarily, the other information can include one or more of the following: the first group identifier, the first analysis identifier, the first interoperability identifier, the task identifier corresponding to the first federated learning task, or information about the vertical federation alliance. For specific interpretations, reference can be made to the relevant descriptions below.

[0043] In a second aspect, a communication method is provided. This method can be executed by a second network element (for example, an entity discovery function network element), or can also be executed by a chip, circuit or logic module of the second network element. The present application does not limit this. For the sake of description, the following takes the execution by the second network element as an example for illustration.

[0044] The method includes: a second network element receives a third request message for obtaining an entity that supports executing a first federated learning task, and the second network element supports discovering an entity that executes the first federated learning task; the second network element determines at least one candidate entity and identity information of the at least one candidate entity, and the identity information of the at least one candidate entity is used to indicate the identity supported by the at least one candidate entity in the first federated learning task; the second network element sends a third response message, and the third response message includes the at least one candidate entity and the identity information of the at least one candidate entity.

[0045] In some implementations, the third request message includes multiple types, and each type in the multiple types corresponds to at least one candidate entity. Here, the type refers to the type of the entity, such as the AF type and / or the NF type.

[0046] In some implementations, the method further includes: the second network element sends information of the third network element and identity information of the third network element to the first network element, and the identity information of the third network element is used to indicate that the third network element supports coordinating a first entity to execute the first federated learning task.

[0047] In some implementations, the method further includes: the second network element receives a registration request message from at least one second entity, and the registration request message includes identity information of the at least one second entity, where the second entity supports executing the first federated learning task, and it should be understood that the at least one second entity includes at least one candidate entity, and the at least one candidate entity includes a first entity.

[0048] In some implementations, the second network element receiving the third request message includes: the second network element receives the third request message from the first network element, and the first network element supports initiating the first federated learning task; or, the second network element receives the third request message from the third network element, and the third network element supports coordinating a first entity to execute the first federated learning task, and the at least one candidate entity includes a first entity.

[0049] The beneficial effects of the second aspect and some implementations can be correspondingly referred to the description related to the first aspect, and will not be elaborated here.

[0050] In a third aspect, a communication method is provided. This method can be executed by a third network element (for example, a task coordinator), or can also be executed by a chip, a circuit or a logic module of the third network element. This application does not make any limitations in this regard. For the sake of description, the following will take the execution by the third network element as an example for illustration.

[0051] The method includes: a third network element obtains a first entity and a first identity, where the first identity is used to indicate the identity of the first entity in performing a first federated learning task, and the third network element supports coordinating the first entity to perform the first federated learning task; the third network element sends a seventh request message to the first entity, where the seventh request message is used to request the execution of the first federated learning task, and the seventh request message includes the first identity.

[0052] In some implementation manners, the third network element obtains the first entity and the first identity, including: the third network element receives the first entity and the first identity from a first network element, and the first network element supports initiating the first federated learning task.

[0053] In some implementation manners, the third network element receives the first entity and the first identity from the first network element, including: the third network element receives a sixth request message from the first network element, where the sixth request message is used to request the creation or coordination of the first federated learning task, and the sixth request message includes the first entity and the first identity.

[0054] In some implementation manners, before the third network element receives the first entity and the first identity information from the first network element, the method further includes: the third network element sends at least one candidate entity and the identity information of at least one candidate entity to the first network element, where the identity information of at least one candidate entity is used to indicate the identity supported by at least one candidate entity in the first federated learning task, and at least one candidate entity includes the first entity.

[0055] In some implementation manners, before the third network element sends at least one candidate entity and the identity information of at least one candidate entity to the first network element, the method further includes: the third network element receives a fourth request message from the first network element, where the fourth request message is used to request to prepare for the execution of the first federated learning task; the third network element obtains at least one candidate entity and the identity information of at least one candidate entity; the third network element sends an eighth request message to at least one candidate entity, where the eighth request message is used to request at least one candidate entity to execute the first federated learning task, and the eighth request message includes the identity information of at least one candidate entity; the third network element receives an eighth response message from at least one candidate entity, where the eighth response message is used to indicate that at least one candidate entity agrees to support executing the first federated learning task with the identity indicated by the identity information of the candidate entity.

[0056] In some implementation manners, the eighth request message further includes information about the third network element, and the eighth response message is further used to indicate agreement for the third network element to negotiate the first federated learning task.

[0057] In some implementation manners, the third network element obtains at least one candidate entity and the identity information of at least one candidate entity, including: the third network element receives at least one candidate entity and the identity information of at least one candidate entity from a second network element, and the second network element supports discovering entities for performing the first federated learning task.

[0058] In some implementations, a third network element receives at least one candidate entity and identity information of at least one candidate entity from a second network element, including: the third network element sends a third request message to the second network element, and the third request message is used to obtain entities that support the execution of a first federated learning task; the third network element receives a third response message from the second network element, and the third response message includes at least one candidate entity and identity information of at least one candidate entity.

[0059] In some implementations, before the third network element receives a first entity and a first identity from a first network element, the method further includes: the third network element receives a fifth request message from the first network element, and the fifth request message is used to request the third network element to coordinate the first entity to execute a first federated learning task; the third network element sends a fifth response message to the first network element, and the fifth response message is used to indicate that the third network element agrees to coordinate the first entity to execute the first federated learning task.

[0060] The beneficial effects of the above-mentioned third aspect and some implementations can be correspondingly referred to the descriptions related to the first aspect, and will not be elaborated here.

[0061] In a fourth aspect, a communication device is provided. The communication device can be used for a first network element and may include modules or units corresponding one by one to the methods / operations / steps / actions described in the first aspect. The module or unit can be a hardware circuit, software, or a combination of a hardware circuit and software.

[0062] In some implementations, the device includes: a processing unit, configured to determine a first entity and a first identity, where the first identity is used to indicate the identity of the first entity in executing a first federated learning task; a transceiver unit, configured to send a first request message to the first entity, and the first request message is used to request the first entity to execute the first federated learning task with the first identity.

[0063] The transceiver unit can perform the reception and transmission processing in the foregoing first aspect and its possible implementations, and the processing unit can perform other processing in the foregoing first aspect and its possible implementations except for reception and transmission.

[0064] In a fifth aspect, a communication device is provided. The communication device can be used for a second network element and may include modules or units corresponding one by one to the methods / operations / steps / actions described in the second aspect. The module or unit can be a hardware circuit, software, or a combination of a hardware circuit and software.

[0065] In some implementations, the apparatus includes: a transceiver unit configured to receive a third request message for obtaining an entity that supports executing a first federated learning task, where a second network element supports discovering an entity that executes the first federated learning task; a processing unit configured to determine at least one candidate entity and identity information of the at least one candidate entity, where the identity information of the at least one candidate entity is used to indicate identities supported by the at least one candidate entity in the first federated learning task; and the transceiver unit is further configured to send a third response message, where the third response message includes the at least one candidate entity and the identity information of the at least one candidate entity.

[0066] The transceiver unit may perform the receiving and sending processes in the foregoing second aspect and its possible implementations. Optionally, the apparatus further includes a processing unit, and the processing unit may perform other processes other than receiving and sending in the foregoing second aspect and its possible implementations.

[0067] In a sixth aspect, a communication apparatus is provided. The communication apparatus may be used for a third network element and may include modules or units corresponding one by one to the methods / operations / steps / actions described in the third aspect. The modules or units may be hardware circuits, software, or a combination of hardware circuits and software.

[0068] In some implementations, the apparatus includes: a processing unit configured to obtain a first entity and a first identity, where the first identity is used to indicate the identity of the first entity in executing a first federated learning task, and a third network element supports coordinating the first entity to execute the first federated learning task; and a transceiver unit configured to send a seventh request message to the first entity, where the seventh request message is used to request executing the first federated learning task, and the seventh request message includes the first identity.

[0069] The transceiver unit may perform the receiving and sending processes in the foregoing third aspect and its possible implementations. Optionally, the apparatus further includes a processing unit, and the processing unit may perform other processes other than receiving and sending in the foregoing third aspect and its possible implementations.

[0070] In a seventh aspect, a communication apparatus is provided, including at least one processor configured to cause the communication apparatus to execute the method in any one of the first aspect to the third aspect, or any possible implementation manner of these aspects, by executing computer programs or instructions, and / or by means of logic circuits.

[0071] In some implementations, the at least one processor is coupled to at least one memory, and the at least one memory stores the foregoing computer programs or instructions. Optionally, the communication apparatus further includes the foregoing at least one memory. Optionally, the at least one processor and the at least one memory are integrated together.

[0072] In an eighth aspect, a chip is provided, which includes a processor and a communication interface. The communication interface is configured to receive information and / or data to be processed and send the information and / or data to be processed to the processor. The processor is configured to process the information and / or data to be processed, so that a communication device installed with the chip executes the method according to any one of the first aspect to the third aspect, or any possible implementation manner of these aspects.

[0073] In a ninth aspect, a computer-readable storage medium is provided, in which computer instructions are stored. When the computer instructions run on a computer, the method according to any one of the first aspect to the third aspect, or any possible implementation manner of these aspects is implemented.

[0074] In a tenth aspect, a computer program product is provided, which includes computer program code. When the computer program code runs on a computer, the method according to any one of the first aspect to the third aspect, or any possible implementation manner of these aspects is implemented.

[0075] In an eleventh aspect, a communication system is provided, which includes a communication device according to any one or more of the fourth aspect to the sixth aspect.

[0076] Among them, for the technical effects of the technical solutions in the fourth aspect to the eleventh aspect, reference can be made to the descriptions of the corresponding technical effects in the first aspect to the third aspect, and details are not repeated here. Description of the Drawings

[0077] Figure 1 is a schematic diagram of a network architecture applicable to an embodiment of the present application;

[0078] Figure 2 is a schematic flowchart of a method for obtaining an entity method for performing federated learning;

[0079] Figure 3 is a schematic flowchart of a federated learning execution method;

[0080] Figure 4 is a schematic flowchart of a communication method provided by an embodiment of the present application;

[0081] Figure 5 is a schematic flowchart of another communication method provided by an embodiment of the present application;

[0082] Figure 6 is a schematic flowchart of yet another communication method provided by an embodiment of the present application;

[0083] Figure 7 is a schematic flowchart of yet another communication method provided by an embodiment of the present application;

[0084] Figure 8It is a schematic flowchart of another communication method provided by an embodiment of the present application;

[0085] Figure 9 It is a schematic diagram of a communication device provided by an embodiment of the present application;

[0086] Figure 10 It is a schematic diagram of another communication device provided by an embodiment of the present application. Detailed implementation manners

[0087] Next, the technical solutions in the present application will be described with reference to the accompanying drawings.

[0088] The technical solutions provided by the present application can be applied to various communication systems, such as: New Radio (NR) systems, Long Term Evolution (LTE) systems, LTE Frequency Division Duplex (FDD) systems, LTE Time Division Duplex (TDD) systems, etc. The technical solutions provided by the present application can also be applied to device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, machine-to-machine (M2M) communication, machine type communication (MTC), and Internet of Things (IoT) communication systems or other communication systems.

[0089] In a communication system, the part operated by an operator can be called a Public Land Mobile Network (PLMN), or can also be called an operator network, etc. A PLMN is a network established and operated by a government or its approved operator for the purpose of providing public land mobile communication services, mainly a public network in which a Mobile Network Operator (MNO) provides mobile broadband access services for users. The PLMN described in the embodiments of the present application can specifically be a network that meets the requirements of the 3rd Generation Partnership Project (3GPP) standard, abbreviated as a 3GPP network. A 3GPP network usually includes but is not limited to a 5th-generation (5G) mobile communication network, a 4th-generation (4G) mobile communication network, and other future communication systems, such as a 6th-generation (6G) mobile communication network, etc.

[0090] For ease of description, in the embodiments of the present application, the PLMN or 5G network will be taken as an example for illustration.

[0091] Figure 1 It is a schematic diagram of a network architecture 100. Taking the 5G network architecture based on the service-based architecture (SBA) in the non-roaming scenario defined in the 3GPP standardization process as an example. As Figure 1 shown, the network architecture may include a terminal device part, a data network (DN) part, and a PLMN part of the operator network. Among them, the PLMN part of the operator network may include, but is not limited to, a (radio) access network ((R)AN) 120 and a core network (CN) part.

[0092] The functions of the network elements of each part will be briefly described below.

[0093] The terminal device part may include a terminal device 110, which is a device that provides voice and / or data connectivity to users. The terminal device 110 may also be referred to as a user equipment UE. The terminal device 110 in this application is a device with wireless transceiver functions and can communicate with one or more core network (CN) devices via an access network device (or also referred to as an access device) in the (radio) access network (R)AN 120. The terminal device 110 may also be referred to as an access terminal, a terminal, a user unit, a user station, a mobile station, a mobile device, a remote station, a remote terminal, a mobile device, a user terminal, a user agent, or a user device, etc. The terminal device 110 may be deployed on land, including indoor or outdoor, handheld or vehicle-mounted; it may also be deployed on water (such as a ship, etc.); it may also be deployed in the air (such as an airplane, a balloon, a satellite, etc.). The terminal device 110 may be a cellular phone, a cordless phone, a session initiation protocol (SIP) phone, a smart phone, a mobile phone, a wireless local loop (WLL) station, a personal digital assistant (PDA), etc. Alternatively, the terminal device 110 may also be a handheld device with wireless communication functions, a computing device, or other devices connected to a wireless modem, a vehicle-mounted device, a wearable device, a drone device, or a terminal in the Internet of Things, the Internet of Vehicles, any form of terminal in a 5G network and future networks, a relay user equipment, or a terminal in a future evolved 6G network, etc. Among them, the relay user equipment may be, for example, a 5G residential gateway (RG). For example, the terminal device 110 may be a virtual reality (VR) terminal, an augmented reality (AR) terminal, a wireless terminal in industrial control, a wireless terminal in unmanned driving, a wireless terminal in remote medical treatment, a wireless terminal in a smart grid, a wireless terminal in transportation safety, a wireless terminal in a smart city, a wireless terminal in a smart home, etc. The terminal device here refers to a 3GPP terminal. The embodiments of this application do not limit the type or category of the terminal device. For ease of description, the following embodiments of this application will use UE to refer to the terminal device as an example for illustration.

[0094] The (R)AN 120 may include one or more access network elements or access network devices. The interface between the access network device and the terminal device may be the Uu interface (or also known as the air interface, that is, the messages exchanged between the access network device and the terminal device may be called air interface messages). Of course, in future communications, the interface name may remain unchanged or may be replaced by other names, and this application is not limited thereto. The (R)AN 120 is a device that provides wireless communication functions for the terminal device 110, and can be a node or device that connects the terminal device to the wireless network, and can also be called a network device. The above RAN may be a 3GPP-related cellular system, such as a 5G mobile communication system, or an evolved system for the future (such as a 6G mobile communication system). The RAN may also be an open radio access network (open RAN, O-RAN or ORAN), a cloud radio access network (cloud radio access network, CRAN), or a wireless fidelity (WiFi) system. The (R)AN 120 can be regarded as a sub-network of the operator network and is an implementation system between the service node and the terminal device 110 in the operator network. For example, the terminal device 110 can be connected to the service node of the operator network through the (R)AN 120 to obtain the services provided by the service node. The (R)AN120 includes but is not limited to: the next generation node base station (gNB) in the 5G system, the evolved node B (eNB) in the long term evolution (LTE), the radio network controller (RNC), the node B (NB), the base station controller (BSC), the base transceiver station (BTS), the home base station (for example, home evolved nodeB, or home node B, HNB), the base band unit (BBU), the transmitting and receiving point (TRP), the transmitting point (TP), the small base station device, the mobile switching center, or the network device in the future network, etc.The access network device can also be a module or unit that completes the functions of a base station. For example, it includes a central unit (CU) and a distributed unit (DU). In a possible network structure, the CU can be used to support communications under protocols such as radio resource control (RRC), packet data convergence protocol (PDCP), and service data adaptation protocol (SDAP). The DU can be used to support communications under the radio link control (RLC) layer protocol, the medium access control (MAC) layer protocol, and the physical layer protocol. The specific technologies and device forms adopted by the access network device in the embodiments of this application are not limited. In systems using different radio access technologies, the names of the devices with the functions of the access network device may be different. For the convenience of description, in all embodiments of this application, the device that provides wireless communication functions for the terminal device 110 is collectively referred to as the access network device or simply as RAN. It should be understood that the specific types of the access network device are not limited herein.

[0095] In different systems, the CU (including CU-CP or CU-UP), or DU or RU may also have different names, but those skilled in the art can understand their meanings. For example, in the O-RAN system, the CU can also be called O-CU (open CU), the DU can also be called O-DU, the CU-CP can also be called O-CU-CP, the CU-UP can also be called O-CU-UP, and the RU can also be called O-RU. For the convenience of description, CU, CU-CP, CU-UP, DU, and RU are used as examples in this application. Any unit among CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented through software modules, hardware modules, or a combination of software modules and hardware modules.

[0096] The CN part may include, but is not limited to, the following network functions (NFs): User Plane Function (UPF) 130, Network Exposure Function (NEF) 131, Network Function Repository Function (NRF) 132, Policy Control Function (PCF) 133, Unified Data Management (UDM) 134, Unified Data Repository (UDR) 135, Network Data Analytics Function (NWDAF) 136, Application Function (AF) 141, Authentication Server Function (AUSF) 137, Access and Mobility Management Function (AMF) 138, and Session Management Function (SMF) 139.

[0097] The data network DN 140, which may also be referred to as a Packet Data Network (PDN), is typically a network located outside the operator's network, such as a third-party network.

[0098] The NF functions included in the CN are further briefly described below.

[0099] 1. The UPF 130 is a gateway provided by the operator and serves as the gateway for communication between the operator's network and the DN 140. The network functions of the UPF 130 include user plane-related functions such as packet routing and transmission, packet detection, service usage reporting, Quality of Service (QoS) handling, lawful interception, uplink packet detection, and downlink packet storage. For example, it is responsible for forwarding and receiving user data in the terminal device 110. It can receive user data from the DN 140 and transmit it to the terminal device 110 through the access network device 120; the UPF 130 can also receive user data from the terminal device 110 through the access network device 120 and forward it to the DN 140. The transmission resources and scheduling functions provided for the terminal device 110 in the UPF 130 are managed and controlled by the SMF 139.

[0100] 2. The NEF 131 is a control plane function provided by the operator, mainly enabling third parties to use the services provided by the network, supporting the network to open its capabilities, events, and data analysis, equipping the PLMN with security information from external applications, and converting the information exchanged inside and outside the PLMN, etc.

[0101] 3. The NRF 132 is a control plane function provided by the operator, which can be used to maintain the real-time information of network functions and services in the network.

[0102] 4. The PCF 133 is a control plane function provided by the operator, mainly supporting the provision of a unified policy framework to control network behavior, providing policy rules to the control layer network functions, and at the same time being responsible for obtaining the subscriber subscription information related to policy decisions. Exemplarily, the PCF 133 can be divided into two PCFs with different functions, namely UE-PCF and AMF-PCF.

[0103] 5. The UDM 134 is a control plane function provided by the operator, responsible for storing information such as the subscriber permanent identifier (SUPI) of the subscribed users in the operator's network, the generic public subscription identifier (GPSI) of the subscribed users, and the credential, etc.

[0104] 6. The UDR 135 is a control plane function provided by the operator, providing the UDM with the function of storing and obtaining subscription data, providing the PCF with the function of storing and obtaining policy data, and storing and obtaining the NF group ID information of the users, etc.

[0105] 7. NWDAF 136 is a control plane function provided by the operator, with functions such as data collection, model training, data analysis, and model inference. The NWDAF network element containing the analytics logical function (AnLF) can be used to infer and export analysis information and expose analysis services, where the analysis can refer to statistical information and / or predictions generated or provided according to the requests of the analysis consumer. The NWDAF network element containing the model training logical function (MTLF) can be used to train a machine learning (ML) model or an artificial intelligence (AI) model and expose new training services, such as providing the trained AI model or ML model to the AnLF. In this application, for the process of obtaining model-related data, the AnLF can be used as the data producer network element, denoted as NFp; the MTLF can be used as the data consumer network element, denoted as NFc. Currently, the AnLF can request a model from the MTLF through the model subscription (MLModelProvision_Subscribe) service or a message, and the model can be obtained by the MTLF training according to the model-related data (such as samples). In addition, the AnLF can use a data set tag to specify a data set, and the data set can contain model-related data. Therefore, the data set tag can be used to mark a data set, that is, mark the model-related data. The data set tag can also be understood as the index of the data set or the index of the data. The MTLF can obtain the data stored by the data producer network element in the data storage by the data storage network element as the consumer network element of the model-related data. Among them, the data producer network element can be the AnLF, and the data storage network element can be the analytics data repository functional (ADRF) and other network elements for storing data. In this application, the relevant data can be understood as the data used for the vertical federated learning task, such as input data, training data, inference data, or sample data, etc.

[0106] 8. AF 141 is a control plane function provided by the operator, which mainly provides corresponding services by interacting with other NFs in the PLMN, such as providing roaming UE visited network selection information, guiding the routing of data streams, accessing the NEF 131, etc.

[0107] 9. AUSF 137 is a control plane function provided by the operator, which is usually used for primary authentication, that is, the authentication between the terminal device 110 (subscribed user) and the operator network.

[0108] 10. The AMF 138 is a control plane network function provided by the operator network, responsible for access control and mobility management of the terminal device 110 accessing the operator network, such as functions including mobile status management, allocation of user temporary identity identifiers, authentication and authorization of users, etc.

[0109] 11. The SMF 139 is a control plane network function provided by the operator network, responsible for managing the protocol data unit (PDU) sessions of the terminal device 110 (including establishment, modification, and release of sessions), for selection and reselection of user plane function network elements, allocation of internet protocol (IP) addresses of the terminal device, quality of service (QoS) control, etc. Among them, the PDU session is a channel for transmitting PDUs, and the terminal device and the DN 140 transmit PDUs to each other through the PDU session. The establishment, maintenance, deletion, etc. of the PDU session are the responsibilities of the SMF network function 139. The SMF network function 139 includes session management (such as session establishment, modification, and release, including tunnel maintenance between the user plane function UPF 130 and the (R)AN 120), selection and control of the UPF network function 130, selection of service and session continuity (SSC) mode, roaming, and other session-related functions.

[0110] It can be understood that the above network elements or functions can be either physical entities in hardware devices, software instances running on dedicated hardware, or virtualized functions instantiated on a shared platform (such as a cloud platform). Simply put, an NF can be implemented by hardware or by software.

[0111] Figure 1 Among Nnef, Nnrf, Npcf, Nudm, Nudr, Nnwdaf, Naf, Nausf, Namf, Nsmf, N1, N2, N3, N4, and N6 are interface sequence numbers. Exemplarily, the meanings of the above interface sequence numbers can be referred to the meanings defined in the 3GPP standard protocol, and this application does not limit the meanings of the above interface sequence numbers. It should be noted that Figure 1 The interface names between the various network functions in are merely examples. In specific implementations, the interface names of this system architecture may also be other names, and this application does not limit this. In addition, the names of the messages (or signaling) transmitted between the above network elements are also merely examples and do not impose any limitations on the functions of the messages themselves.

[0112] It should be noted that in Figure 1In the architecture shown, the interface between the (R)AN and the CN can also be referred to as the NG interface (not shown in the figure), and the (R)AN and the CN are connected through the NG interface. The NG interface can include an NG-C interface and an NG-U interface. Among them, the NG-C interface is a control plane interface, and the two connected parties are the (R)AN and the AMF, which is used to transmit control plane data; the NG-U interface is a user plane interface, and the two connected parties are the (R)AN and the UPF, which is used to transmit user plane data.

[0113] It should be understood that the above network architecture 100 is only a network architecture described from the perspective of a service-based architecture. In this service-based architecture, the PLMN can, according to specific scenario requirements, orderly combine some or all network functions as needed to realize the customization of network capabilities and services, so as to deploy a dedicated network for different services, that is, to realize 5G network slicing. The network slicing technology enables operators to respond more flexibly and quickly to customer needs and supports the flexible allocation of network resources.

[0114] For the convenience of description, in the embodiments of the present application, network functions (such as NEF 131…SMF139) are collectively / abbreviated as NF, that is, the NF described later in the embodiments of the present application can be replaced by any network function. In addition, in the embodiments of the present application, the session management function SMF 139 is abbreviated as SMF, and the terminal device 110 is called UE, that is, the SMF described later in the embodiments of the present application can be replaced by the session management function, and the UE can be replaced by the terminal device. Figure 1 Only some network functions are schematically described, and the NF described later is not limited to Figure 1 the network functions shown in.

[0115] It should be understood that Figure 1 the AMF, SMF, UPF, NEF, AUSF, NRF, PCF, and UDM shown in can be understood as network elements in the core network for implementing different functions. For example, they can be combined into network slices as needed. These core network elements can be independent devices or can be integrated into the same device to implement different functions. The present application does not limit the specific form of the above network elements.

[0116] It should also be understood that the above naming is only defined for the convenience of distinguishing different functions and should not constitute any limitation to the present application. The present application does not exclude the possibility of using other names in 5G networks and future other networks. For example, in 6G networks, some or all of the above network elements may continue to use the terms in 5G, or other names may also be used.

[0117] For the convenience of understanding, the following first introduces the relevant terms, concepts, or technologies that may be involved in the embodiments of the present application:

[0118] 1. Federated Learning;

[0119] Federated Learning (FL) is a distributed machine learning method in which multiple participants interact model parameters through a security mechanism without interacting or sharing raw training data, thereby achieving the effect of collaborative training. In other words, federated learning is an encrypted distributed machine learning technology. Federated learning can make full use of the data and computing capabilities of participants, enabling multiple parties to collaboratively build a general and robust machine learning model without the need to share data. Therefore, federated learning can effectively help multiple institutions use data and build learning models while meeting the requirements of user privacy protection, data security, and government regulations. It can be said that federated learning is designed to share knowledge and parameters without exchanging any of its own data.

[0120] Federated learning includes horizontal federated learning and vertical federated learning. In an environment where data supervision is becoming increasingly strict, federated learning can solve key problems such as data ownership, data privacy, data access rights, and access to heterogeneous data. A horizontal row of a data matrix represents a training sample, and a vertical column represents a data feature. Horizontal federated learning is to conduct federated learning by combining multiple rows of samples with the same features of multiple participants, that is, the training data of each participant is horizontally divided. Horizontal federated learning is also called feature-aligned federated learning, that is, the data features of the participants are aligned, and horizontal federated learning can increase the total amount of training samples. As a machine learning technology, VFL can be used to solve model training and inference in the case where each member is unwilling to share raw data, and is applicable to the situation where the training sample identifiers (IDs) of the participants overlap more, while the data features of the participants overlap less. VFL conducts federated learning by combining different data features of the common samples of multiple participants, that is, the training data of each participant is vertically divided, so it is called vertical federated learning. Vertical federated learning is to conduct federated learning by combining different data features of the common samples of multiple participants, that is, the training data of each participant is vertically divided. Vertical federated learning is also called sample-aligned federated learning, that is, the training samples of the participants are aligned, and vertical federated learning can increase the dimension of training data features.

[0121] Federated learning has been applied in the industrial community. For example, Google has applied it to its own GBoard project. Another example is the FATE federated learning framework proposed by WeBank. These are all typical examples of the practical application of the federated learning method with the above-mentioned parameter server architecture. They all require a very high degree of consistency among different devices participating in the learning, with consistent model structures and aligned data.

[0122] 2. Analysis ID;

[0123] The analysis ID can be used to indicate an analysis service, which is abbreviated as service. This service is related to the model, that is, the model can be used to execute this service. Or it can be described as the analysis ID is related to the model, that is, the model is used to execute the service corresponding to the analysis ID.

[0124] Or it can also be understood that MTLF is related to the analysis ID, that is, the model provided by MTLF supports the execution of the service corresponding to this analysis ID. Exemplarily, MTLF can be related to one or more analysis IDs. It can be understood that this MTLF can provide models for the services corresponding to each of the one or more analysis IDs. For example, MTLF1 is related to analysis ID 1 and analysis ID 2, that is, MTLF1 corresponds to analysis ID 1 and analysis ID 2, then MTLF1 can provide models for the services corresponding to analysis ID 1 and analysis ID 2.

[0125] 3. Interoperability indicator;

[0126] Exemplarily, the interoperability indicator can correspond to MTLF, or to the analysis ID, or to the analysis ID corresponding to MLTF. Or it can be described as the interoperability indicator is related to MTLF, or the interoperability indicator is related to the analysis ID. Among them, the interoperability indicator can also be called an interoperability indication, or a machine learning (ML) model interoperability indicator, or an ML Model interoperability indicator.

[0127] The interoperability indicator includes a list of manufacturers, or is described as a list of NWDAF providers (or suppliers). The manufacturers in this list of manufacturers are allowed to retrieve or use the models provided by MTLF. The interoperability indicator also indicates that MTLF supports the NWDAF of the manufacturers in this list of manufacturers to request the models provided by MTLF. The interoperability indicator also indicates that the manufacturers in this list of manufacturers are allowed to obtain models from MTLF. The interoperability indicator also indicates that this MTLF allows the manufacturers in this list of manufacturers to obtain models from this MTLF.

[0128] A list of interoperability identifiers for MTLF providers. For example, the interoperability identifier represents a manufacturer identifier, or the interoperability identifier is associated with the manufacturer identifier. The interoperability identifier can be associated with an analysis identifier. For example, if they are in a one-to-one correspondence, it means that MTLF allows the corresponding manufacturer or the MTLF included in the manufacturer to obtain the model corresponding to the analysis identifier, and / or it means that MTLF is allowed to interoperate with AnLF for the model corresponding to the analysis identifier. Optionally, an MTLF can have one or more interoperability identifiers. If there are multiple interoperability identifiers, the multiple interoperability identifiers respectively correspond to different analysis identifiers. For example, MTLF NF ID 1 corresponds to analysis identifier 1 and analysis identifier 2. Among them, the MTLF to which MTLF NF ID 1 belongs has interoperability identifier 1 and interoperability identifier 2. Interoperability identifier 1 corresponds to analysis identifier 1, and interoperability identifier 2 corresponds to analysis identifier 2. Optionally, if the MTLF to which MTLF NF ID 1 belongs and the MTLF to which MTLF NF ID 2 belongs support interoperability, the MTLF to which MTLF NF ID 2 belongs can have interoperability identifier 1 and interoperability identifier 2, where interoperability identifier 1 corresponds to analysis identifier 1, and / or interoperability identifier 2 corresponds to analysis identifier 2, that is, MTLFs of the same manufacturer can have the same interoperability identifier for the same analysis identifier. In addition, if the MTLF to which MTLF NF ID 1 belongs and the MTLF to which MTLF NF ID 2 belongs belong to different manufacturers, the MTLF to which MTLF NF ID 2 belongs can have interoperability identifier 3 and interoperability identifier 4, where interoperability identifier 3 corresponds to analysis identifier 1, and / or interoperability identifier 4 corresponds to analysis identifier 2, that is, MTLFs of the same manufacturer can have different interoperability identifiers for the same analysis identifier.

[0129] Exemplarily, analysis identifier 1 is related to model 1, that is, model 1 is used to execute the service corresponding to analysis identifier 1. Analysis identifier 1 is related to interoperability identifier 1, that is, model 1 is related to interoperability identifier 1. Assume that interoperability identifier 1 includes the identifiers of manufacturer 1 and manufacturer 2, that is, model 1 can be provided for manufacturer 1 and manufacturer 2 to use, or it can be understood that if the manufacturer of NWDAF is manufacturer 1 or manufacturer 2, then this NWDAF can use model 1.

[0130] Exemplarily, an MTLF can have one or more interoperability identifiers. If there are multiple interoperability identifiers, the multiple interoperability identifiers can correspond to different analysis identifiers. For example, MTLF1 corresponds to analysis identifier 1 and analysis identifier 2. Among them, interoperability identifier 1 corresponds to analysis identifier 1, and interoperability identifier 2 corresponds to analysis identifier 2.

[0131] 4. Model producer;

[0132] The model producer is the entity that produces the model or the entity that has the right to provide model information to other entities according to the network configuration. Consumers can obtain the model based on the model information. The model information includes, but is not limited to, the uniform resource locator (URL) of the model file, the model itself, etc.

[0133] 5. Manufacturer identifier;

[0134] The vendor ID of a vendor can also be represented as network element information, which identifies the vendor of the network element or the manufacturer. For example, Vendor ID1 identifies the vendor of the NWDAF network element.

[0135] The manufacturer identifier can be used to identify a device manufacturer. One manufacturer identifier can correspond to one or more NWDAF device identifiers. For example, both NWDAF NF ID 1 and NWDAF NF ID 2 can correspond to manufacturer identifier 1, that is, the MTLF to which MTLF NF ID 1 belongs and the MTLF to which MTLF NF ID 2 belongs are from the same manufacturer, and the manufacturer identifier of this manufacturer is manufacturer identifier 1.

[0136] The terms involved in the present application are briefly described above and will not be elaborated in the following embodiments. In addition, the above description of the terms is only for easy understanding and does not limit the protection scope of the embodiments of the present application.

[0137] Figure 2 It is a schematic flowchart of a method for obtaining a federated learning entity. As Figure 2 shown, it includes the following multiple steps. For the parts not elaborated in detail, reference can be made to the existing protocol. Among them, steps S201 - S205 are the registration process of NWDAF, steps S206 - S208 are the discovery process of the client NWDAF, and steps S209 - S212 are the selection process for the server NWDAF to determine the client NWDAF that executes federated learning.

[0138] S201. The server NWDAF sends a registration request message #1 to the NRF. Correspondingly, the NRF receives the registration request message #1 from the server NWDAF, where the registration request message #1 is used to request registration to the network.

[0139] Exemplarily, the registration request message #1 may be an Nnrf_NFManagement_NFRegister_request message, and the registration request message #1 includes configuration parameters (such as Server NWDAF profile), where the configuration parameters include one or more of the following: NWDAF NF type, analysis ID(s), address information of the server NWDAF, service area, FL capability type information (such as FL server), or the time interval for which the server NWDAF supports FL.

[0140] S202, the client NWDAF sends a registration request message #2 to the NRF. Correspondingly, the NRF receives the registration request message #2 from the client NWDAF, where the registration request message #2 is used to request registration to the network.

[0141] Exemplarily, the registration request message #2 may be an Nnrf_NFManagement_NFRegister_request message, and the registration request message #2 includes configuration parameters (such as Client NWDAF profile), where the configuration parameters include one or more of the following: NWDAF NF type, analysis ID(s), address information of the client NWDAF, service area, FL capability type information (such as FL client), or the time interval for which the client NWDAF supports FL.

[0142] It should be understood that both the server NWDAF and the client NWDAF are NWDAFs that include MTLF and can participate in federated learning training.

[0143] Optionally, the number of client NWDAFs in this application is not limited. For example, there may be one or more, such as NWDAF 1, …, NWDAF N, where N is an integer greater than or equal to 1.

[0144] S203, the NRF stores the configuration parameters of the server NWDAF and the client NWDAF.

[0145] Exemplarily, the NRF stores Server NWDAF profile and Client NWDAF profile.

[0146] S204, the NRF sends a registration response message #2 to the client NWDAF. Correspondingly, the client NWDAF receives the registration response message #2 from the NRF.

[0147] S205, the NRF sends a registration response message #1 to the server NWDAF. Correspondingly, the server NWDAF receives the registration response message #1 from the NRF.

[0148] In S206, the server NWDAF sends a discovery request message to the NRF. Correspondingly, the NRF receives the discovery request message from the server NWDAF.

[0149] Exemplarily, the server NWDAF discovers available services in the network by sending a discovery request message to the NRF. For example, the server NWDAF invokes Nnrf_NFDiscovery_Request from a properly configured NRF within the same PLMN and carries the NF type of the desired NF instance in Nnrf_NFDiscovery_Request. Optionally, one or more of the following may also be carried: the desired service name, the NF type of the server NWDAF, and the desired target NF location.

[0150] S207, NRF authorization.

[0151] Exemplarily, the NRF authorizes the Nnrf_NFDiscovery_Request in step S206. For example, the NRF decides whether to allow the server NWDAF to discover the desired NF instance according to the profile of the desired NF / NF service and the type of the server NWDAF. If the desired NF instance or NF service instance is deployed in a certain network slice, the NRF authorizes the discovery request according to the discovery configuration of the network slice. For example, the desired NF instance can only be discovered by NFs in the same network slice.

[0152] In S208, the NRF sends a discovery response message to the server NWDAF. Correspondingly, the server NWDAF receives the discovery response message from the NRF.

[0153] Among them, one or more client NWDAFs may be carried in the discovery response message, such as client NWDAF1,..., client NWDAF N.

[0154] Exemplarily, if the NRF authorizes, the NRF determines a set of matching NF instances according to Nnrf_NFDiscovery_Request and the NRF internal policy, and issues the NF profile of the NF instance, such as the client NWDAF. Exemplarily, the discovery response message may be an Nnrf_NFDiscovery_Response message, and the NF profile of each NF instance can be sent to the server NWDAF through the Nnrf_NFDiscovery_Response message.

[0155] Optionally, if the server NWDAF carried the desired target NF location in step S206, the NRF shall not restrict the set of discovered NF instances or NF service instances to the target NF location. For example, if no NF instance or NF service instance can be found for the preferred target NF location, the NRF may provide NF instances or NF service instances whose locations are not the preferred target NF locations.

[0156] S209, the server NWDAF sends a federated learning preparation request message to one or more client NWDAFs. Correspondingly, one or more client NWDAFs receive the federated learning preparation request message from the server NWDAF.

[0157] Exemplarily, the federated learning preparation request message may be a Federated Learning preparation request message. For example, the server NWDAF uses the Nnwdaf_MLModel training_subscription or Nnwdaf_MLModel training information_request service with an ML preparation flag to send a federated learning preparation request to the FL client NWDAF, to check whether the client NWDAF can meet the ML model training requirements (such as analysis ID, ML model interoperability information), available data requirements (a list of event IDs of local data for training, the available data requirements may also include dataset statistical attributes, the time window of data samples, and the minimum number of data samples), or availability time requirements (the time span required for the FL process), etc.

[0158] S210, one or more client NWDAFs determine whether to join the federated learning.

[0159] Exemplarily, the client NWDAF(s) checks whether it can meet the ML model training requirements, and / or, if model information is provided in the federated learning preparation request message in step S209, the client NWDAF(s) also needs to check whether it can successfully download the model and decide whether to join the federated learning process according to the implementation. The example criteria used by the client NWDAF(s) may be based on its availability, computing and communication capabilities, and ML model interoperability information.

[0160] S211, one or more client NWDAFs send a federated learning preparation response message to the server NWDAF. Correspondingly, the server NWDAF receives the federated learning preparation response message from one or more client NWDAFs.

[0161] Exemplarily, the federated learning preparation response message can be a Federated Learning preparation response message. For example, the client NWDAF(s) invokes the Nnwdaf_MLModel training_subscription response service operation or the Nnwdaf_MLModel training information_request response service operation to indicate whether it will join the federated learning process. If it cannot join the federated learning process, the client NWDAF(s) can carry a cause value in the federated learning preparation response message. For example, the client NWDAF(s) currently cannot support federated learning, or the client NWDAF(s) is currently overloaded, etc.

[0162] S212. The server NWDAF selects one or more client NWDAFs that support the execution of federated learning.

[0163] Exemplarily, based on the federated learning preparation response messages sent by one or more client NWDAFs received in step S211, the server NWDAF selects or finally determines one or more client NWDAFs that support the execution of federated learning. The specific process for one or more client NWDAFs to execute federated learning can refer to the solution Figure 3 shown below.

[0164] Figure 3 is a schematic flowchart of a federated learning execution method. As Figure 3 shown, it includes the following multiple steps. For parts not described in detail, reference can be made to the existing protocol.

[0165] S301. The consumer sends a subscription request message #1 to the server NWDAF. Correspondingly, the server NWDAF receives the subscription request message #1 from the consumer. The subscription request message #1 is used to subscribe to ML model provisioning or training.

[0166] Exemplarily, the consumer (for example, the NWDAF containing AnLF, or the NWDAF containing MTLF) uses the Nnwdaf_MLModelProvision service to send a subscription request message #1 to the server NWDAF, such as subscription request for ML model provisioning / training, to retrieve the ML model.

[0167] Among them, the subscription request message #1 includes one or more of the following: analysis ID, ML model metrics (e.g., ML model accuracy), accuracy reporting interval, predetermined status (ML model accuracy threshold or time when an ML model is needed). It should be understood that the ML model accuracy threshold can be used to indicate the target ML model accuracy of the training process. When the ML model accuracy threshold is reached during the training process, the server NWDAF can stop the training process. If the consumer provides the time when an ML model is needed, the server NWDAF can consider this information to determine the maximum response time of its client NWDAF.

[0168] S302, the server NWDAF determines the client NWDAF(s).

[0169] Among them, the specific implementation method can refer to the relevant description of the above method 200. For the sake of brevity, it will not be elaborated here. For example, the client NWDAF(s) determined by the server NWDAF includes: client NWDAF1,..., client NWDAF N.

[0170] S303, the server NWDAF sends the subscription request message #2 to the client NWDAF(s). Correspondingly, the client NWDAF(s) receives the subscription request message #2 from the server NWDAF.

[0171] Exemplarily, the subscription request message #2 can be Nnwdaf_MLModelTraining_Subscribe. For example, the server NWDAF sends Nnwdaf_MLModelTraining_Subscribe or Nnwdaf_MLModelTrainingInfo_Request to the client NWDAF(s) to perform local model training.

[0172] Among them, the subscription request message #2 can carry one or more of the following: Initial Federated Learning parameters provisioning, ML model metrics, initial ML model, or maximum response time, where the maximum response time refers to the maximum response time for the client NWDAF to report temporary local ML model information to the server NWDAF.

[0173] S304, optionally, the client NWDAF(s) collect data.

[0174] Exemplarily, if the client NWDAF does not yet have available local data, each client NWDAF can use the current mechanism in TS23.288 to collect its local data from the NF (data provider).

[0175] S305, the client NWDAF(s) sends a subscription response message #2 to the server NWDAF. Correspondingly, the server NWDAF receives the subscription response message #2 from the client NWDAF(s).

[0176] Exemplarily, the subscription response message #2 can be Nnwdaf_MLModelTraining_Notify, which is used to report local model training information. For example, during the federated learning training process, each client NWDAF trains the ML model provided by the server NWDAF based on its own data and reports the temporary local ML model information to the server NWDAF in Nnwdaf_MLModelTraing_Notify or Nnwdaf_MLModelTraingInfo_Response.

[0177] Optionally, Nnwdaf_MLModelTraining_Notify or Nnwdaf_MLModelTrainingInfo_Response can also include the local ML model metrics and training input data information (such as the area covered by the dataset, the sampling ratio, the maximum / minimum values of the data in each dimension, etc.) calculated by the client NWDAF(s).

[0178] Optionally, the ML model is sent from the client NWDAF(s) to the server NWDAF during the federated learning training process, which is the information required for the server NWDAF to build an aggregated model based on the locally trained ML model. If the client NWDAF cannot complete the training of the temporary local ML model within the maximum response time provided by the server NWDAF, the client NWDAF(s) can send a delay event notification, including a delay event indication, an optional reason code (e.g., the local ML model training failed, or more time is required for the local ML model training), and the expected time to complete the training available to the client NWDAF before the maximum response time has passed, etc.

[0179] S306, optionally, the server NWDAF sends a subscription request message #3 to the client NWDAF(s). Correspondingly, the client NWDAF(s) receives the subscription request message #3 from the server NWDAF.

[0180] Exemplarily, the subscription response message #3 can be Nnwdaf_MLModelTraining_Notify, carrying an extended response time and / or the current iteration round ID. For example, if the server NWDAF receives a notification / response that the client NWDAF(s) cannot complete training within the maximum response time, the server NWDAF can send an extended maximum response time to the client NWDAF in Nnwdaf_MLModelTraining_Subscribe or Nnwdaf_MLModel TrainingInfo_Request. Before this request, the client NWDAF needs to report the temporary local ML model information to the server NWDAF. Otherwise, the server NWDAF may instruct the client NWDAF to skip the reporting for this iteration. The client NWDAF includes the current iteration round ID in the request message to indicate that this request is for modifying the training parameters of the current iteration round.

[0181] Optionally, the server NWDAF can notify the client NWDAF to stop ML model training by sending a termination request and report the current local ML model update.

[0182] S307. The client NWDAF performs model aggregation.

[0183] Exemplarily, the client NWDAF aggregates all the local ML model information retrieved in step S305 to update the global ML model. Optionally, the server NWDAF can also calculate the global ML model metrics, such as based on the local ML model metrics, or by applying the global model on the validation dataset (if available). When the client NWDAF provides updated local ML model information each time, the server NWDAF can update the global ML model, or the server NWDAF can decide to wait for the local ML model information from all client NWDAFs before updating the global ML model.

[0184] If the server NWDAF provides the maximum response time for the client NWDAF to provide temporary local ML model information in step S303, or provides an extended maximum response time in step S306, the server NWDAF decides to wait for the client NWDAF that has not provided its temporary local ML model within the (extended) maximum response time, or only aggregates the retrieved local ML model information instances to update the global ML model. The server NWDAF makes the decision based on the notification / response from the client NWDAF, or based on the local configuration if no notification is received.

[0185] S308. Optionally, the client NWDAF sends an update message to the consumer. Correspondingly, the consumer receives the update message from the client NWDAF. The update message is used to indicate the current ML training status to the consumer.

[0186] Exemplarily, in response to step S301, the client NWDAF sends an Nnwdaf_MLModelProvision_Notify message to the consumer to dynamically update the global ML model metrics (e.g., reaching the ML model accuracy threshold or training time expiration) to the consumer periodically (e.g., a certain number of training epochs or every 10 minutes), or when certain predefined states are reached.

[0187] S309. Optionally, the consumer sends a subscription request message #4 to the server NWDAF. Correspondingly, the server NWDAF receives the subscription request message #4 from the consumer. The subscription request message #4 is used to modify the subscription for update or termination.

[0188] Exemplarily, the consumer determines whether the current model can meet the requirements. For example, the global ML model metrics are satisfactory to the consumer, and decides to stop or continue the training process. The user can re - invoke the Nnwdaf_MLModelProvision_Subscribe service operation used in step S301 to stop or continue the training process.

[0189] S310. Optionally, the server NWDAF updates or terminates the federated learning training process.

[0190] Exemplarily, the server NWDAF updates or terminates the current federated learning training process according to the subscription request message #4 sent by the consumer in step S309. Optionally, if the server NWDAF receives a request to stop the joint training process in step S309, the following steps S311 and S312 are skipped.

[0191] S311. Optionally, the server NWDAF sends the aggregated model information to the client NWDAF(s). Correspondingly, the client NWDAF(s) receives the aggregated model information from the server NWDAF.

[0192] In other words, if the federated learning training process continues, the server NWDAF will identify the client NWDAF and send an Nnwdaf_MLModelTraingInfo_Request including the aggregated ML model information to the selected client NWDAF(s) for the next round of joint training.

[0193] S312. Optionally, the client NWDAF(s) updates the local model according to the aggregated model information.

[0194] Exemplarily, each client NWDAF updates its local ML model according to the aggregated ML model information distributed by the server NWDAF in step S311.

[0195] It should be noted that by repeating the above steps S304 - S312 until the training termination condition is reached (for example, the maximum number of iterations, or the result of the loss function is lower than the threshold).

[0196] When the joint training process is completed, the server NWDAF requests the client NWDAF(s) to terminate the federated learning training process. In one implementation, the server NWDAF calls the Nnwdaf_MLModelTraing_Unsubscribe service with the cause code that the federated learning process has been completed and can optionally use the final aggregated ML model information. Then, the client NWDAF(s) terminates the local model training. If the final aggregated ML model information is received from the server NWDAF, the client NWDAF(s) can store the aggregated ML model information for further use.

[0197] Optionally, after the federated learning training process is completed, the client NWDAF can send a Nnwdaf_MLModelProvision_Notify containing the global optimal ML model information to the consumer.

[0198] Currently, 5GC only supports horizontal federated learning, and there is no process mechanism for how to support vertical federated learning. In vertical federated learning, there are multiple types of participants, such as UE, NWDAF, or AF. However, the existing network element discovery mechanism does not support discovering multiple types of NF types in the same request message. Moreover, in vertical federated tasks, various participants may have different label information, and the models trained using different label information may be different, resulting in the ineffective execution of the federated learning task. Therefore, how to effectively execute vertical federated learning tasks is a problem that needs to be considered.

[0199] In view of this, the present application provides a communication method and a communication device, which can effectively improve the execution efficiency of the first federated learning task by sending a first identity to the first entity, enabling the first entity to execute the first federated learning task with the first identity.

[0200] The following will describe in detail the communication method provided by the embodiments of the present application with reference to the accompanying drawings. The embodiments provided by the present application can be applied to a communication scenario where a sending device and a receiving device communicate, for example, it can be applied to the Figure 1 communication system shown above.

[0201] Figure 4It is a schematic flowchart of the communication method 400 provided by an embodiment of the present application. As Figure 4 shown, the method flow can be executed by a first network element (e.g., a task initiator), a second network element (e.g., an entity discovery function network element), or a third network element (e.g., a task coordinator), or can also be executed by a chip or circuit of the first network element, the second network element, or the third network element, or can also be implemented by a logic module or software capable of implementing all or part of the communication device functions. The present application does not limit this. The following will be described with the first network element, the second network element, or the third network element as the execution subject. The method includes the following multiple steps. For parts not described in detail, reference can be made to the above existing protocols.

[0202] S410, the second network element receives a third request message.

[0203] Among them, the third request message is used to request to obtain an entity that supports executing a first federated learning task.

[0204] In the present application, federated learning can be referred to as federated machine learning, joint learning, or coalition learning.

[0205] Optionally, the first federated learning task can be replaced with: other names such as first federated learning, or first federated learning function, or first federated learning process, or first federated learning activity, etc. The present application does not limit this. In addition, obtaining an entity that supports executing a first federated learning task can be understood as: obtaining an entity (or network element or device or node, etc.) with the ability to execute a first federated learning task, or rather, the entity supports executing a first federated learning task (or activity, or process, etc.).

[0206] Exemplarily, the first federated learning task can include one or more of the following: network performance analysis, artificial intelligence (AI) model training, face recognition, or cross-institutional medical data analysis and disease prediction, etc. Optionally, the federated learning task in the embodiments of the present application can refer to a vertical federated learning task, and the specific interpretation can be referred to the above relevant description.

[0207] In the present application, the second network element supports discovering an entity that executes a first federated learning task, or rather, the second entity supports providing information about an entity with the ability to execute a first federated learning task. For example, the second network element can be an NRF or an NEF or other network entities. Optionally, for discovering entities of the NF (e.g., NWDAF) type, the second network element can be an NRF or other network entities; for discovering entities of the AF type, the second network element can be an NEF or other network entities.

[0208] It should be understood that an entity that supports the execution of the first federated learning task can be understood as: the entity supports the first federated learning task, or in other words, the entity has the ability to execute the first federated learning task. Exemplarily, the entity may include one or more of the following: NWDAF, AF, a coordinator (i.e., an example of the third network element), or a network element or entity related to the first federated learning task. Optionally, in the present application, the entity that supports the execution of the first federated learning task may also be replaced with a network element or device that supports the execution of the first federated learning task, etc.

[0209] Optionally, in one implementation, the second network element may receive a third request message from the first network element or the third network element. For details, refer to steps S401 or S402 below.

[0210] S401, the first network element sends a third request message to the second network element. Correspondingly, the second network element receives the third request message from the first network element.

[0211] In the embodiments of the present application, the first network element supports initiating the first federated learning task, or in other words, the first network element has the ability to initiate the first federated learning task. For example, the first network element may be a coordinator, NWDAF, or AF. Assume that the first network element is AF1 and the second network element is NRF. Then AF1 may send a third request message to NRF to request to obtain an entity that supports network performance analysis.

[0212] S402, the third network element sends a third request message to the second network element. Correspondingly, the second network element receives the third request message from the third network element.

[0213] In the embodiments of the present application, the third network element supports coordinating the first entity to execute the first federated learning task, or in other words, the third network element has the ability to support coordinating the first entity to execute the first federated learning task. Optionally, the third network element may belong to the same network as the first network element or the second network element. For example, the third network element may be an NF, AF, or coordinator. Alternatively, the third network element may be deployed in the same network as the first network element or the second network element, such as an external network. The third network element may be a third-party entity. The present application does not make any limitations in this regard. Assume that the third network element is a task coordinator and the second network element is NRF. Then the task coordinator may send a third request message to NRF to request to obtain an entity that supports network performance analysis.

[0214] Optionally, the third request message includes one or more of the following:

[0215] (1) Multiple types;

[0216] Among them, each of the multiple types corresponds to at least one candidate entity. Here, the type can be understood as the type of the entity. For example, it can be multiple network function (NF) types and / or multiple application function (AF) types. Exemplarily, the at least one candidate entity may include one or more of the following: AF, NF, or a third network element (such as a task coordinator). The at least one candidate entity refers to a candidate entity with the ability to execute the first federated learning task. In the embodiments of the present application, the at least one candidate entity includes a first entity.

[0217] That is to say, by carrying multiple types in the third request message, the first network element or the third network element can improve the efficiency of the second network element in discovering at least one candidate entity. That is to say, the second network element can provide at least one candidate entity corresponding to the multiple types. That is to say, the types of candidate entities provided by the second network element are diverse and not limited to one NF type or one AF type.

[0218] (2) The first group of identifiers;

[0219] Among them, the first group of identifiers is used to indicate the first federated learning group. The entities in the first federated learning group support the execution of the first federated learning task, that is, the first federated learning group corresponds to the first federated learning task. For example, the first federated learning group includes at least one candidate entity, which also means that the first entity is included in the first federated learning group. Optionally, the entities and their identity information within the first federated learning group are pre-configured.

[0220] For example, the first group of identifiers is VFL ID1, and the information of the entities in the first federated learning group includes one or more of the following: AF identifier AF ID(s), AF instance(s) (such as AF1, AF2), AF type(s), name(s) of the application service (Application Name), name(s) of the manufacturer's device Vendor ID(s), NF identifier NF ID(s), NF instance(s) (such as NWDAF ID), or network identifier PLMN ID(s), where AF1 identifies a certain type of AF or a specific AF.

[0221] In other words, by carrying VFL ID1 in the third message, the second network element can correspondingly determine the first federated learning group, and thus determine the member information in the first federated learning group (including one or more of the member identifier, member type, member's identity information, member's ability information, and federated learning task supported by the member). Among them, the member information in the first federated learning group can be predefined or pre-configured.

[0222] Exemplarily, the first federated learning group may also be referred to as the first federated learning alliance, where the alliance member information (including one or more of the alliance member identifier, alliance member type, identity information of the alliance member, capability information of the alliance member, and federated learning tasks supported by the alliance member).

[0223] In a possible case, it further includes information about entities in the first federated learning group.

[0224] (3) The first analysis identifier;

[0225] Wherein, the first analysis identifier is used to indicate the first federated learning task.

[0226] Exemplarily, the first analysis identifier can be used to indicate a certain specific function or service, which is related to the model, that is, the model can be used to execute the specific function or service, or in other words, the model supports the specific function or service corresponding to the first analysis identifier. Among them, the specific function or service can be face recognition or network performance analysis, etc. Taking the analysis ID as an example, the first federated learning group can correspond to one or more analysis IDs. The analysis ID can, for example, indicate analysis services such as terminal anomaly detection and terminal session analysis (such as quality of service QoS analysis).

[0227] (4) The first interoperability indicator;

[0228] Wherein, the first interoperability indicator can correspond to the first analysis identifier and is used to indicate the first federated learning task.

[0229] (5) The task identifier corresponding to the first federated learning task, and this task identifier is used to identify the first federated learning task.

[0230] Optionally, this task identifier can be assigned by the first network element, or it can also be assigned by the third network element and then sent to the first network element. This application does not make any limitations in this regard.

[0231] (6) Information about the vertical federated alliance;

[0232] Exemplarily, the information of the vertical federation can be used to identify or distinguish the vertical federation (e.g., the first federated learning group). For example, the information of the vertical federation may include one or more of the identifier of the vertical federation (e.g., the first group identifier), the identifier of at least one analysis service corresponding to the vertical federation (e.g., the first analysis identifier), the identifier of at least one AF corresponding to the vertical federation, the information of at least one network corresponding to the vertical federation, or the information of at least one equipment manufacturer corresponding to the vertical federation. Among them, the identifier of the vertical federation can be used to indicate or identify a vertical federation, and can be called the alliance ID (i.e., the first group ID, such as VFL ID1).

[0233] It can be understood that the vertical federation (e.g., the first federated learning group) can be a professional group composed of two or more members (or called participants), aiming to participate in the vertical federation of common activities or achieve common results by sharing each other's resources. Among them, any member can be a natural person, a company, an organization, a network element entity (such as an application function or a service), or a network (such as a PLMN or a CN), etc., or can be a combination of any of the above members. All or part of the members within the alliance jointly participate in the vertical federation task (e.g., the first federated learning task). The vertical federation task refers to the task corresponding to vertical federated learning, and the vertical federation task may include training and / or inference. The vertical federation task can also be called the vertical federation activity or the vertical federated learning task, or simply called the learning task or the federation task, etc. In various embodiments of the present application, multiple AFs included in the vertical federation can respectively maintain their own data. For example, each AF maintains the terminal data of different terminals. The object of vertical federated learning can be a terminal.

[0234] For a vertical federated task (e.g., the first federated learning task), the data of the terminals respectively included in multiple AFs within the vertical federated alliance can be used as samples for the learning task. Among them, vertical federated learning generally requires vertical federated members to use the data of the same object for training. The determination of the samples for the vertical federated task can be understood as the vertical federated members determining which object's data to jointly use for the vertical federated task. Exemplarily, when the vertical federated task is the PDU session analysis of the terminal, and the vertical federated members include the CN, AF1, and AF2, then the CN, AF1, and AF2 all need to use the data of the same terminal for the vertical federated task. Among them, AF1 and AF2 can be connected to the CN. For example, if the vertical federated members include NF network elements (such as NWDAF) in the PLMN network, AF1, and AF2, then the NWDAF, AF1, and AF2 all need to use the data of the same terminal for the vertical federated task. Among them, AF1 and AF2 can be AFs in the PLMN network. Also, for example, the AF can also be an AF outside the network, or a trusted AF of the PLMN, or an untrusted AF. That is to say, the relationship between the AF and the network is not restricted. It can be understood that the relationship between the members within the vertical federated alliance is also not restricted. In this application, the AF can be an application provider, for example, it can be an application server (AS) or an application-related network element entity, etc.

[0235] S420, the second network element determines at least one candidate entity and the identity information of at least one candidate entity.

[0236] Among them, the identity information of at least one candidate entity is used to indicate the identities supported by at least one candidate entity in the first federated learning task. The at least one candidate entity refers to a candidate entity with the ability to execute the first federated learning task. In the embodiments of this application, the at least one candidate entity includes the first entity.

[0237] Exemplarily, the identities supported by the candidate entity in the first federated learning task include any one or more of the following:

[0238] (1) Support providing labels in the first federated learning task, where the labels correspond to the first federated learning task (which can be called the main participant). Among them, the labels are used for the first federated learning task, for example, for the training and / or model evaluation of the model corresponding to the first federated task;

[0239] (2) Support providing labels and data in the first federated learning task (which can be referred to as the primary participant), where the data is used for the first federated learning task, for example, for training and / or inference of the model corresponding to the first federated task; Optionally, the data granularity supported by the vertical federation can be terminal granularity, alliance identification granularity, analysis identification granularity, alliance member granularity, etc. That is, it represents the granularity of the data used in the first federated learning task. For example, terminal granularity indicates that the samples used in the first federated learning task are data of terminals within a certain range. For example, for terminal granularity, the data granularity supported by the vertical federation can be used to indicate that the samples used in the first federated learning task are the traffic data of the terminal on AF or the business data of the terminal on AF, etc.

[0240] (3) Support providing data in the first federated learning task (which can be referred to as the secondary participant);

[0241] (4) Support coordinating participants to complete the first federated learning task (which can be referred to as the coordinator).

[0242] In the embodiments of the present application, the identity information can be understood as role information, or rather, ability information. The supported identities can be understood as the supported roles, where the roles can be understood as the logical roles in the first federated learning task, such as the primary participant, the secondary participant, or the coordinator, etc. The supported identities can also be understood as the capabilities possessed, or the capabilities corresponding to the identities, including: the ability to provide federated training labels in the first federated learning task, and / or the ability to provide federated training data in the first federated learning task.

[0243] Next, the specific implementation manner for the second network element to determine at least one candidate entity and the identity information of at least one candidate entity will be described.

[0244] In one implementation manner, the third request message itself can indicate the first federated learning task. For example, the third request message can be such that the second network element can determine the first federated learning task and / or in the first federated learning group according to the received third request message, and use some or all of the entities in the first federated learning group as candidate entities, thereby determining at least one candidate entity and the identity information of at least one candidate entity.

[0245] In another implementation manner, the second network element determines the corresponding first federated learning group according to at least one of the first analysis identifier, the first group identifier, the information of the vertical federation, or the first interoperability identifier carried in the third request message, and then determines at least one candidate entity and the identity information of at least one candidate entity from the first federated learning group.

[0246] In yet another implementation, the second network element selects entities corresponding to the AF type and / or NF type from the first federated learning group according to multiple types carried in the third request message, and uses the entities of the AF type and / or NF type as candidate entities, so as to determine at least one candidate entity and the identity information of at least one candidate entity.

[0247] Optionally, at least one candidate entity can be understood as: at least one active entity in the first federated learning group, where the active entity can refer to: a member in the current first federated learning group that has or supports participating in the first federated learning task.

[0248] Optionally, before performing step S410, the second network element obtains at least one candidate entity in the first federated learning group and the identity information of at least one candidate entity.

[0249] In one example, the at least one candidate entity in the first federated learning group and the identity information of at least one candidate entity can be predefined, or signaled-configured or pre-configured. Among them, predefined can include predefined in advance, such as protocol definition, and pre-configuration can be implemented by pre-saving corresponding codes, tables, strings or other means for indicating candidate entities in the first federated learning group and their identity information in the second network element. The present application does not limit its specific implementation manner.

[0250] In another example, a second entity in the first federated learning group sends a registration request message to a second network element to request registration to the network. The registration request message carries the configuration parameter (profile) of the second entity. Correspondingly, the second network element can save the configuration parameter of the second entity. The configuration parameter may include at least one of the identifier of the second candidate entity, the type of the second candidate entity, the federated learning tasks supported by the second candidate entity (e.g., the first federated learning task), the identifier of the first group where the second candidate entity is located, the information of the vertical federated alliance to which the second candidate entity belongs, the first analysis identifier supported by the second candidate entity, the first interoperability identifier, or the identity supported by the second candidate entity in the first federated learning task. It should be understood that at least one second entity supports the execution of the first federated learning task, where at least one second entity includes at least one candidate entity, and at least one candidate entity includes the first entity. Therefore, the second network element can learn at least one candidate entity in the first federated learning group and its identity information in the execution of the first federated learning task, such as whether it supports providing labels in the first federated learning task, or whether it has the ability to coordinate the first entity to execute the first federated learning task, etc. For example, if the second entity includes AF1, AF2, NWDAF1, NWDAF2, and a third network element, then AF1, AF2, NWDAF1, NWDAF2, and the third network element respectively send registration request messages to the second network element and carry their respective configuration parameters. Among them, AF1 and AF2 support providing data in the first federated learning task, NWDAF1 and NWDAF2 support providing labels and data in the first federated learning task, and the third network element supports coordinating entities to execute the first federated learning task, etc. Further, the second network element can select some entities from the second entities as candidate entities. For example, at least one candidate entity includes AF1, AF2, NWDAF1, and the third network element.

[0251] S430, the second network element sends a third response message.

[0252] Among them, the third response message includes at least one candidate entity and the identity information of at least one candidate entity. Exemplarily, at least one candidate entity may include various types of entities, such as AF, NF, NWDAF, or a third network element (e.g., a task coordinator), etc.

[0253] Optionally, in one implementation, in response to the above step S401 or S402, the second network element may send a third response message to the first network element or the third network element. For details, see the following steps S403 or S404.

[0254] S403, the second network element sends a third response message to the first network element. Correspondingly, the first network element receives the third response message from the second network element.

[0255] Exemplarily, assume that the first network element is AF1 and the second network element is NRF. After receiving the third request message from AF1, NRF can, according to the NF type and AF type carried in the third request message, feedback NF instances and AF instances to AF1, such as AF2 and NF3.

[0256] Exemplarily, assume that the first network element is AF1 and the second network element is NEF. After receiving the third request message from AF1, NEF can, according to the first set of identifiers or information on vertical federated alliances carried in the third request message, determine the first federated training group or the first federated learning alliance, and feedback NF instances and / or AF instances to AF1, such as NF1, AF2, and NF2.

[0257] S404. The second network element sends a third response message to the third network element. Correspondingly, the third network element receives the third response message from the second network element.

[0258] Exemplarily, assume that the third network element is a task coordinator and the second network element is NRF. After receiving the third request message from the task coordinator, NRF can, according to the first set of identifiers or information on vertical federated alliances carried in the third request message, determine the first federated training group or the first federated learning alliance, and then can feedback NF instances and / or AF instances in the first federated training group to the task coordinator, such as one or more of AF2, NF1, or NF3.

[0259] Exemplarily, assume that the first network element is AF1 and the second network element is NEF. After receiving the third request message from AF1, NEF can, according to the NF type and AF type carried in the third request message, feedback NF instances and AF instances to AF1, such as AF2 and NF3.

[0260] Optionally, if the above steps S401 and S403 are not executed, or S403 is not executed, then after executing step S404, the third network element can send at least one candidate entity and the identity information of at least one candidate entity to the first network element. For details, see step S404a.

[0261] S404a. The third network element sends at least one candidate entity and the identity information of at least one candidate entity to the first network element. Correspondingly, the first network element receives at least one candidate entity and the identity information of at least one candidate entity from the third network element.

[0262] That is to say, the first network element can directly or indirectly obtain at least one candidate entity and the identity information of at least one candidate entity from the second network element.

[0263] S440. The first network element determines the first entity and the first identity.

[0264] Among them, the first identity is used to indicate the identity of the first entity in performing the first federated learning task. It can be understood that the first identity is used to indicate the identity, or role, or function assumed by the first entity in performing the first federated learning task.

[0265] Optionally, the present application does not limit the number of first entities, and it can be one or more.

[0266] In one implementation, based on step S403, the first network element obtains at least one candidate entity and the identity information of at least one candidate entity from the second network element, and then selects or determines the first entity and the first identity from at least one candidate entity and the identity information of at least one candidate entity.

[0267] In another implementation, the first network element obtains at least one candidate entity and the identity information of at least one candidate entity from the third network element, and then selects or determines the first entity and the first identity from at least one candidate entity and the identity information of at least one candidate entity. It can be an internal implementation behavior of the first network element, or the first network element selects or determines the first entity and the first identity from at least one candidate entity and the identity information of at least one candidate entity through predefined constraint rules. The present application does not make specific limitations on this.

[0268] Exemplarily, assume that there are three candidate entities AF1, AF2, and NF3. Among them, AF1 supports providing labels in the first federated learning task, AF2 supports providing labels and data in the first federated learning task, and NF3 supports providing data in the first federated learning task. For example, the first network element can use AF1 as the main participant and AF2 and NF3 as the slave participants. That is, the first entity includes AF1, AF2, and NF3. The first identity of AF1 is used to indicate that AF1 provides labels in the first federated learning task, the first identity information of AF2 is used to indicate that AF2 provides data in the first federated learning task, and the first identity information of NF3 is used to indicate that NF3 provides data in the first federated learning task. Again, for example, the first network element can use AF2 as the main participant and NF3 as the slave participant. That is, the first entity includes AF2 and NF3. The first identity of AF2 is used to indicate that AF2 provides labels and data in the first federated learning task, and the first identity information of NF3 is used to indicate that NF3 provides data in the first federated learning task. That is to say, the first network element can determine all entities in the candidate entities as the first entity, or can determine most entities in the candidate entities as the first entity. The present application does not make limitations on this. Optionally, it depends on factors such as the current load and / or current capability information of the candidate entities.

[0269] Optionally, when there are multiple entities among the candidate entities that can provide labels in the first federated learning task, the first network element may, according to content logic, or randomly select one of them as the primary participant. It should be understood that in the first federated learning task, there is only one entity that provides labels, that is, the entities participating in the execution of the first federated learning task can use the same labels for federated learning training.

[0270] Optionally, in one implementation, before the first network element determines the first entity and the first identity, the first network element obtains a first analysis identifier, and the first analysis identifier corresponds to the first federated learning task; the first network element determines to trigger the first federated learning task according to the first analysis identifier.

[0271] Exemplarily, the first network element obtaining the first analysis identifier may be: the first network element may receive a subscription request message from a consumer (for example, an NWDAF including AnLF, or an NWDAF including MTLF), and the subscription request message includes the first analysis identifier, and the subscription request message is used to subscribe to the federated learning model provisioning or training. Among them, the subscription request message may be a model subscription request message or an analysis subscription request message.

[0272] Exemplarily, the first analysis identifier may be predefined or preconfigured. Among them, predefined may include predefined in advance, such as protocol definition, and preconfiguration may be implemented by pre-saving corresponding codes, tables, strings, or other means for indicating the first analysis identifier in the first network element. The present application does not limit its specific implementation manner.

[0273] It should be understood that the first analysis identifier may be used to indicate a certain specific function or service, and the specific function or service is related to the model, that is, the model can be used to execute the specific function or service, or in other words, the model supports the specific function or service corresponding to the first analysis identifier. Among them, the specific function or service may be network performance analysis, etc. The first analysis identifier corresponds to the first federated learning task, which can be understood as: the first analysis identifier is used to indicate the first federated learning task.

[0274] Optionally, the first network element may also obtain other information, and determine to trigger the first federated learning task according to the correspondence between the other information and the first federated learning task. Exemplarily, the other information may include one or more of the following: the first group identifier, the first analysis identifier, the first interoperability identifier, the task identifier corresponding to the first federated learning task, or the information of the vertical federation alliance. The specific interpretation can refer to the relevant description below.

[0275] Optionally, before performing step S440, that is, before the first network element determines the first entity and the first identity, the first network element may confirm with at least one candidate entity whether it agrees or supports performing the first federated learning task with the assigned identity. It should be understood that the at least one candidate entity includes the first entity. For ease of description, the following uses the example of the first network element confirming with the first entity whether it agrees or supports performing the first federated learning task with the assigned first identity or the first identity carried in the second request message. For details, see the following steps S411 - S413 (not shown in the figure).

[0276] S411, the first network element sends a second request message to the first entity. Correspondingly, the first entity receives the second request message from the first network element. The second request message is used to request the first entity to prepare to perform the first federated learning task.

[0277] Optionally, the second request message includes one or more of the following. For specific entities, refer to the relevant descriptions above:

[0278] (1) At least one identity supported by the first entity in the first federated learning task, and the at least one supported identity includes the first identity;

[0279] It should be understood that the identity supported by the first entity in the first federated learning task can be one or more, for example, including one or more of the following: providing labels, providing data, or coordinating the first entity to perform the first federated learning task.

[0280] (2) Information of some or all entities performing the first federated learning task, for example, the identification and / or identity information of some or all entities;

[0281] (3) The first analysis identifier;

[0282] (4) The first group identifier;

[0283] (5) The first identity;

[0284] (6) Information of the third network element that supports coordinating the first entity to perform the first federated learning task, such as the coordinator ID;

[0285] (7) Network identifier, such as PLMN ID.

[0286] The network identifier is used to indicate the specific network where the first network element requests the first entity to prepare to perform the first federated learning task.

[0287] (8) The task identifier corresponding to the first federated learning task, and the task identifier is used to identify the first federated learning task;

[0288] (9) Information of the vertical federation.

[0289] It should be noted that in the embodiments of the present application, the identities supported by the first entity in the first federated learning task can be one or more. For example, assuming that the first entity supports multiple identities in the first federated learning task, that is, the first entity supports both providing labels and providing data in the first federated learning task. In contrast, the first identity represents the identity of the first entity in performing the first federated learning task, which can be that the first entity only provides labels in the first federated learning task, or, it can also be that the first entity only provides data in the first federated learning task, or, it can further be that the first entity provides labels and data in the first federated learning task. That is to say, the identities supported by the first entity in the first federated learning task can be the main participant and / or the subordinate participant, and the first identity can be the main participant or the subordinate participant, that is, the first identity belongs to the identities supported by the first entity in the first federated learning task.

[0290] Optionally, in one implementation, the first network element may send a fourth request message to the third network element. Correspondingly, the third network element receives the fourth request message from the first network element, where the fourth request message is used to request the first entity to prepare to execute the first federated learning task. That is to say, the first network element can directly or indirectly request the first entity to prepare to execute the first federated learning task. The parameters included in the fourth request message and their interpretations can refer to the relevant descriptions of the parameters included in the second request message, which will not be elaborated here.

[0291] S412, the first entity confirms whether it agrees to execute the first federated learning task.

[0292] Exemplarily, after verifying that the identity of the first network element is passed, the first entity can determine whether to agree to execute the first federated learning task according to its current load condition, and / or, whether it currently has the ability to execute the first federated learning task, such as agreeing or disagreeing.

[0293] Optionally, if the first identity is carried in the second request message, the first entity determines whether it supports executing the first federated learning task with the first identity, or rather, whether the first entity currently has the ability to execute the first federated learning task with the first identity, and then agrees or disagrees to execute the first federated learning task with the first identity.

[0294] Optionally, if the second request message carries the identities supported by the first entity in the first federated learning task, the first entity may determine whether to agree to execute the first federated learning task based on its current load condition and / or its current ability to execute the first federated learning task. Alternatively, the first entity may also determine whether to agree to execute the first federated learning task with the supported identities based on its current load condition and / or its current ability to execute the first federated learning task.

[0295] Optionally, if the first entity agrees to execute the first federated learning task, the following step S413 may be performed. It should be understood that the technical solution of this application is based on the premise that the first entity agrees to execute the first federated learning task.

[0296] Optionally, if the first entity does not agree to execute the first federated learning task, the first entity may send a failure reason value to the first network element to indicate the refusal to execute the first federated learning task. For example, the first entity refuses to execute the first federated learning task with the first identity. The failure reason value may be that the current load of the first entity is too high, or the first entity does not currently support the first identity, or the first entity does not currently have the ability to execute the first federated learning task with the first identity.

[0297] S413. The first entity sends a second response message to the first network element. Correspondingly, the first network element receives the second response message from the first entity. The second response message is used to indicate that the first entity agrees or supports executing the first federated learning task with the first identity.

[0298] Optionally, if the second request message carries the first identity, the first entity may carry the first identity in the second response message when agreeing to execute the first federated learning task, indicating that the first entity agrees or supports executing the first federated learning task with the first identity.

[0299] Optionally, if the second request message carries the identities supported by the first entity in the first federated learning task, the first entity may send the first identity to the first network element when agreeing to execute the first federated learning task, indicating that the first entity agrees or supports executing the first federated learning task with the first identity. It should be understood that the identities supported by the first entity in the first federated learning task include the first identity. Optionally, the second response message may carry the identities supported by the first entity in the first federated learning task, indicating that the first entity agrees to execute the first federated learning task with the identity or role assigned by the first network element, where the identity or role assigned by the first network element includes the first identity.

[0300] Optionally, if the second request message received by the first entity carries information of a third network element as a parameter (6), the second response message can also be used to indicate that the first entity agrees for the third network element to coordinate the first entity to execute the first federated learning task.

[0301] It should be noted that the above steps S411 - S413 are described by taking the first network element's confirmation with at least one candidate entity on whether to agree or support executing the first federated learning task with the assigned identity as an example. Optionally, before executing step S440, that is, before the first network element determines the first entity and the first identity, the third network element (e.g., the coordinator) can confirm with at least one candidate entity on whether to agree or support executing the first federated learning task with the assigned identity. For details, see the following steps S414 - S416 (not shown in the figure).

[0302] S414. The third network element sends an eighth request message to at least one candidate entity. Correspondingly, at least one candidate entity receives the eighth request message from the third network element. The eighth request message is used to request at least one candidate entity to execute the first federated learning task, and the eighth request message includes the identity information of at least one candidate entity.

[0303] Optionally, the eighth request message includes one or more of the following: For specific interpretations, refer to the relevant descriptions above:

[0304] (1) The identities supported by at least one candidate entity in the first federated learning task;

[0305] It should be understood that the identities supported by at least one candidate entity in the first federated learning task can be one or more, including one or more of the following: providing labels, providing data, or coordinating the first entity to execute the first federated learning task.

[0306] (2) Information of some or all entities executing the first federated learning task, e.g., the identification and / or identity information of some or all entities;

[0307] (3) The first analysis identifier;

[0308] (4) The first group identifier;

[0309] (5) The first identity;

[0310] (6) Information of the third network element, where the third network element supports coordinating the first entity to execute the first federated learning task, e.g., the coordinator ID;

[0311] (7) Network identifier, e.g., PLMN ID.

[0312] (8) The task identifier corresponding to the first federated learning task, and this task identifier is used to identify the first federated learning task;

[0313] (9) Information of the vertical federation alliance.

[0314] S415. At least one candidate entity confirms whether it agrees to execute the first federated learning task.

[0315] Exemplarily, after verifying that the identity of the third network element is passed, at least one candidate entity can determine whether to agree to execute the first federated learning task according to its current own load condition, and / or whether it currently has the ability to execute the first federated learning task, such as agreeing or disagreeing.

[0316] Optionally, if the first identity is carried in the eighth request message, at least one candidate entity determines whether it supports executing the first federated learning task with the first identity, or rather whether at least one candidate entity currently has the ability to execute the first federated learning task with the first identity, and then agrees or disagrees to execute the first federated learning task with the first identity.

[0317] Optionally, if at least one candidate entity's supported identity in the first federated learning task is carried in the eighth request message, at least one candidate entity can determine whether to agree to execute the first federated learning task according to its current own load condition, and / or whether it currently has the ability to execute the first federated learning task, or at least one candidate entity can also determine whether to agree to execute the first federated learning task with the supported identity according to its current own load condition, and / or whether it currently has the ability to execute the first federated learning task.

[0318] Optionally, if at least one candidate entity agrees to execute the first federated learning task, the following step S416 can be executed. It should be understood that the technical solution of this application is based on the premise that at least one candidate entity agrees to execute the first federated learning task.

[0319] Optionally, if at least one candidate entity disagrees to execute the first federated learning task, at least one candidate entity can send a failure reason value to the first network element, which is used to indicate the refusal to execute the first federated learning task. Among them, the failure reason value can be that the current load of at least one candidate entity is too large, or at least one candidate entity currently does not support the identity of at least one candidate entity, or at least one candidate entity currently does not have the ability to execute the first federated learning task with the identity of at least one candidate entity.

[0320] S416. At least one candidate entity sends an eighth response message to the third network element. Correspondingly, the third network element receives the eighth response message from at least one candidate entity. Among them, the eighth response message is used to indicate that at least one candidate entity agrees to execute the first federated learning task with the identity of the candidate entity.

[0321] Optionally, if the first identity is carried in the eighth request message, at least one candidate entity may carry the first identity in the eighth response message when agreeing to execute the first federated learning task, indicating that at least one candidate entity agrees to support the execution of the first federated learning task with the first identity.

[0322] Optionally, if the identity supported by at least one candidate entity in the first federated learning task is carried in the eighth request message, at least one candidate entity may send the identity supported in the first federated learning task to the third network element when agreeing to execute the first federated learning task, indicating that at least one candidate entity agrees to support the execution of the first federated learning task with the first identity. It should be understood that the at least one candidate entity includes a first entity, and the identity supported by the first entity in the first federated learning task includes the first identity. Optionally, the identity supported by at least one candidate entity in the first federated learning task may be carried in the eighth response message, indicating that at least one candidate entity agrees to execute the first federated learning task with the assigned identity or role.

[0323] Optionally, if the information of the third network element, which is parameter (6), is carried in the eighth request message received by at least one candidate entity, the eighth response message may also be used to indicate that at least one candidate entity agrees that the third network element coordinates at least one candidate entity to execute the first federated learning task.

[0324] It should be noted that the above steps S411 - S413 and steps S414 - S416 can be executed alternatively. That is to say, this application does not limit the execution entity for confirming whether at least one candidate entity agrees or supports executing the first federated learning task with the assigned identity. For example, it can be the first network element, or the third network element, or other network elements with triggering capabilities.

[0325] Based on the above implementation, when the first entity agrees to execute the first federated learning task with the first identity, the first network element can finally determine the first entity and the first identity for executing the first federated learning task, and then trigger the first entity to execute the first federated learning task. Specifically, refer to the following step S450. The specific execution process of the first federated learning task can refer to the relevant description of the above method 300 and will not be elaborated here.

[0326] S450, the first network element sends a first request message to the first entity. Correspondingly, the first entity receives the first request message from the first network element.

[0327] Among them, the first request message is used to request the first entity to execute the first federated learning task with the first identity. That is to say, the first network element requests the first entity to execute the first federated learning task with the first identity. Exemplarily, the first identity includes a main participant or a subordinate participant, and the specific interpretation can refer to the relevant description of the above step S420.

[0328] Optionally, the first network element may directly send a first request message to the first entity. Alternatively, the first network element may first send a sixth request message to a third network element (such as a coordinator), where the sixth request message is used to request the creation or coordination of a first federated learning task. The sixth request message includes the first entity and a first identity. Correspondingly, after receiving the sixth request message, the third network element organizes the first entity to execute the first federated learning task with the first identity. Furthermore, the third network element may send a first federated learning task initialization request message to the first entity, which is used to request the first entity to execute the first federated learning task with the first identity.

[0329] Optionally, the sixth request message may further include one or more of the following. For specific interpretations, reference may be made to the above relevant descriptions:

[0330] (1) A first analysis identifier;

[0331] (2) A first group identifier;

[0332] (3) Information about some or all of the entities that execute the first federated learning task, where the information about some or all of the entities includes identifiers and / or identity information;

[0333] (4) A task identifier corresponding to the first federated learning task, which is used to identify the first federated learning task;

[0334] (5) Information about a vertical federated alliance.

[0335] Optionally, the first request message may further include one or more of the following. For specific interpretations, reference may be made to the above relevant descriptions:

[0336] (1) A first analysis identifier;

[0337] (2) A first identity;

[0338] (3) A first group identifier;

[0339] (4) Information about some or all of the entities that execute the first federated learning task, where the information about some or all of the entities includes identifiers and / or identity information;

[0340] (5) Information about the third network element, such as the identifier and / or identity information of the third network element.

[0341] That is to say, the first network element may send the first identity to any first entity that executes the first federated learning task, or may also send the identifiers and / or identity information of other entities that execute the first federated learning task to the first entity.

[0342] (6) A task identifier corresponding to the first federated learning task, which is used to identify the first federated learning task;

[0343] (7) Information of the vertical federation alliance.

[0344] It should be noted that the above step S450 is described with the first network element (e.g., the task initiator) triggering a request to the first entity to execute the first federated learning task with the first identity. Optionally, in one implementation, the third network element (e.g., the coordinator) can also trigger a request to the first entity to execute the first federated learning task with the first identity. For details, please refer to the following steps S405 - S407.

[0345] S405, the third network element obtains the first entity and the first identity.

[0346] Among them, the first identity is used to indicate the identity of the first entity in executing the first federated learning task.

[0347] Optionally, in one implementation, the third network element can obtain the first entity and the first identity from the first network element. For details, please refer to the following step S406.

[0348] S406, the first network element sends the first entity and the first identity to the third network element. Correspondingly, the third network element receives the first entity and the first identity from the first network element.

[0349] Exemplarily, the first network element can carry the first entity and the first identity in the sixth request message in the above step S450 and send them to the third network element.

[0350] Optionally, in one implementation, the third network element can determine the first entity and the first identity by itself. For example, after the third network element obtains at least one candidate entity and the identity information of at least one candidate entity through step S404, and then confirms whether at least one candidate entity agrees to execute the first federated learning task by executing the above steps S414 - S416, finally determines the first entity and the first identity. Among them, the implementation method for the third network element to determine the first entity and the first identity can refer to the implementation method of the first network element in the above step S440, which is not specifically limited here.

[0351] Optionally, in one implementation, the third network element can obtain the first entity and the first identity from other network elements. The other network elements can be UDM, UDR, or a third - party server, etc. This application does not make specific limitations.

[0352] Exemplarily, after the third network element obtains the first entity and the first identity, it can then trigger the execution of the following step S407.

[0353] S407, the third network element sends a seventh request message to the first entity. Correspondingly, the first entity receives the seventh request message from the third network element.

[0354] Among them, the seventh request message is used to request the first entity to execute the first federated learning task, and the seventh request message includes a first identity.

[0355] Optionally, the first entity determines whether the sender of step S407, that is, the third network element, is consistent with the information of the third network element carried in the second request message in step S411. If they are the same, the first entity executes the first federated learning task with the first identity; if they are different, the first entity may refuse to execute the first federated learning task. In this implementation manner, by determining whether the information of the third network element allocated by the first network element is the same as the information of the third network element that triggers the execution of the first federated learning task, the first entity can avoid executing other federated learning tasks initiated by malicious third network elements, ensuring network security while reducing the processing load or signaling overhead of the first entity.

[0356] It should be noted that the above step S450 and step S407 can be executed alternatively, or rather, the present application does not limit the execution entity that triggers the request for the first entity to execute the first federated learning task. For example, it can be the first network element, or the third network element, or other network elements with triggering capabilities.

[0357] Optionally, before executing step S405 or S406, the first network element can confirm whether the third network element agrees or supports coordinating the first entity to execute the first federated learning task. For specific details, please refer to the following steps S417 - S418 (not shown in the figure).

[0358] S417, the first network element sends a fifth request message to the third network element. Correspondingly, the third network element receives the fifth request message from the first network element. Among them, the fifth request message is used to request the third network element to coordinate the first entity to execute the first federated learning task.

[0359] Optionally, the fifth request message may include one or more of the following: a first analysis identifier, a first group identifier, information of some or all of the entities that execute the first federated learning task. The information of the some or all of the entities includes an identifier and / or identity information, information of the third network element, or a task identifier corresponding to the first federated learning task. For specific interpretations, please refer to the above relevant descriptions.

[0360] S418, the third network element sends a fifth response message to the first network element. Correspondingly, the first network element receives the fifth response message from the third network element. Among them, the fifth response message is used to indicate that the third network element agrees or supports coordinating the first entity to execute the first federated learning task.

[0361] Optionally, if the third network element does not agree to coordinate the first entity to execute the first federated learning task, the third network element may send a failure reason value to the first network element, which is used to indicate the refusal to coordinate the first entity to execute the first federated learning task. The failure reason value may be that the current load of the third network element is too large, or the third network element does not currently support the identity of the coordinator, or the third network element does not currently have the ability to coordinate the first entity to execute the first federated learning task.

[0362] It should be understood that this implementation manner is based on the premise that the third network element agrees to coordinate the first entity to execute the first federated learning task.

[0363] According to the solution provided above, the first network element can finally determine the first entity and the first identity for executing the first federated learning task. By informing the first entity of the first identity in the execution of the first federated learning task, the first entity can execute the first federated learning task with the first identity, thereby effectively improving the execution efficiency of the first federated learning task.

[0364] Figure 5 It is a schematic flowchart of the communication method 500 provided by an embodiment of the present application. As Figure 5 shown, taking the first network element (such as the task initiator) as the NWDAF, the second network element (such as the entity discovery function network element) as the NRF, and the task participants AF1 and AF2 as the execution entities for interaction, this method can be regarded as a further refinement of the above method 400. It should be understood that Figure 5 the embodiment shown in Figure 4 can be coupled with the embodiment shown in Figure 4 and can be referred to each other. Therefore, the relevant descriptions in the above method 400 also apply to this implementation manner. There may be the same or similar technical means between the two. The content already described in the embodiment shown in

[0365] S501, the vertically federated participating entity initiates a registration process to the vertically federated entity discovery function network element. The specific registration process can refer to the relevant description of the above method 200 and will not be described here.

[0366] Optionally, in one implementation manner, the vertically federated participating entity includes the NWDAF, AF1, and AF2, and the federated entity discovery function network element may be the NRF or the NEF. Then, the NWDAF, AF1, and AF2 respectively send registration request messages to the NRF to request registration to the network. For ease of description, this implementation manner takes the NWDAF as the task initiator and AF1 and AF2 as the task participants as an example for illustration.

[0367] Among them, the registration request message may carry the configuration parameters (profile) of the vertical federation participating entity. The configuration parameters may include the identification information of the entity, the type of the entity, the first analysis ID, the first federated learning group ID joined (for example, the first federated learning group ID can be indicated by AF ID, PLMN ID, Vendor ID, or UE ID, etc.), or the supported identities, etc. For the specific interpretations of the parameters, reference can be made to the relevant descriptions of the above method 400. Exemplarily, the first federated learning task includes one or more of model training, face recognition, or cross-institutional medical data analysis.

[0368] It should be noted that the embodiment of the present application does not limit the granularity level of the parameters carried in the registration request message.

[0369] S502, the NWDAF sends an entity discovery request message (i.e., an example of the third request message) to the NRF. Correspondingly, the NRF receives the entity discovery request message from the NWDAF.

[0370] Among them, for the parameters included in the entity discovery request message and their specific interpretations, reference can be made to the relevant descriptions of the third request message in step S401 of the above method 400.

[0371] S503, the NRF determines at least one candidate entity and the identity information of at least one candidate entity according to the first analysis ID and the first federated learning group ID, such as AF1 and AF2. For the specific implementation method, reference can be made to the relevant descriptions of step S420 of the above method 400.

[0372] Optionally, the NRF can also determine at least one candidate entity and the identity information of at least one candidate entity according to other parameters carried in the entity discovery request message, such as the information of the vertical federation alliance and / or the first interoperability identifier.

[0373] S504, the NRF sends an entity discovery response message (i.e., an example of the third response message) to the NWDAF. Correspondingly, the NWDAF receives the entity discovery response message from the NRF.

[0374] Among them, for the parameters included in the entity discovery response message and their specific interpretations, reference can be made to the relevant descriptions of the third response message in step S403 of the above method 400.

[0375] Furthermore, the NWDAF can request at least one candidate entity (such as AF1 and AF2) to be prepared to execute the first federated learning task. For the specific details, see the following steps S505 - S510.

[0376] S505, the NWDAF sends a federated learning request task request message #1 (i.e., an example of the second request message) to AF1. Correspondingly, AF1 receives the federated learning request task request message #1 from the NWDAF.

[0377] Exemplarily, the NWDAF may send the federated learning request task request message #1 to AF1 through the NEF.

[0378] Among them, the parameters included in the federated learning request task request message #1 and their specific interpretations can refer to the relevant descriptions of the second request message in the above method 400.

[0379] S506, AF1 agrees to participate in the federated learning task according to the assigned identity.

[0380] S507, AF1 sends a federated learning request task response message #1 (i.e., an example of the second response message) to the NWDAF. Correspondingly, the NWDAF receives the federated learning request task response message #1 from AF1.

[0381] Exemplarily, AF1 may send the federated learning request task response message #1 to the NWDAF through the NEF.

[0382] Among them, the parameters included in the federated learning request task request response #1 and their specific interpretations can refer to the relevant descriptions of the second response message in the above method 400.

[0383] S508, the NWDAF sends a federated learning request task request message #2 (i.e., an example of the second request message) to AF2. Correspondingly, AF2 receives the federated learning request task request message #2 from the NWDAF.

[0384] Exemplarily, the NWDAF may send the federated learning request task request message #2 to AF2 through the NEF.

[0385] Among them, the parameters included in the federated learning request task request message #2 and their specific interpretations can refer to the relevant descriptions of the second request message in the above method 400.

[0386] S509, AF2 agrees to participate in the federated learning task according to the assigned identity.

[0387] S510, AF2 sends a federated learning request task response message #2 (i.e., an example of the second response message) to the NWDAF. Correspondingly, the NWDAF receives the federated learning request task response message #2 from AF2.

[0388] Exemplarily, AF2 may send the federated learning request task response message #2 to the NWDAF through the NEF.

[0389] Among them, the parameters and specific interpretations included in the federated learning request task request response #2 can refer to the relevant descriptions of the second response message of the above method 400.

[0390] It should be noted that the specific implementation manners of the above steps S505 - S510 can refer to the relevant descriptions of steps S411 - S413 of the above method 400. For the sake of brevity, they will not be elaborated here.

[0391] S511, the NWDAF determines the first entity and the first identity.

[0392] Exemplarily, the NWDAF can determine the first entity and the first identity according to the feedback of steps S507 and S510. For example, the first entity includes AF1 and AF2. Correspondingly, the first identity of AF1 can be the main participant, that is, AF1 provides labels in the execution of the first federated learning task, and the first identity of AF2 can be the secondary participant, that is, AF2 provides data in the execution of the first federated learning task.

[0393] Exemplarily, the NWDAF can determine the first entity and the first identity according to the feedback of steps S507 and S510. For example, the first entity includes AF1, AF2, and NWDAF. Correspondingly, the first identity of NWDAF can be the main participant, that is, NWDAF provides labels in the execution of the first federated learning task. Optionally, NWDAF1 can also provide data in the first learning task. The first identities of AF1 and AF2 can be the secondary participants, that is, AF1 and AF2 provide data in the execution of the first federated learning task.

[0394] Optionally, the NWDAF can participate in the execution of the first federated learning task or not participate in the execution of the federated learning task. This application does not make any limitations in this regard. That is, the first entity executing the first federated learning task can include AF1, AF2, and NWDAF, or the first entity executing the first federated learning task can include AF1 and AF2.

[0395] Further, after determining the first entity and the first identity, the NWDAF can trigger a request for the first entity to execute the first federated learning task, which specifically includes the following steps S512 - S514.

[0396] S512, the NWDAF sends an initialization request message #1 (i.e., an example of the first request message) to AF1, and the corresponding AF1 receives the initialization request message #1 from the NWDAF.

[0397] Among them, the parameters and specific interpretations included in the initialization request message #1 can refer to the relevant descriptions of the first request message of the above method 400.

[0398] S513, AF1 sends an initialization response message #1 to the NWDAF. Correspondingly, the NWDAF receives the initialization response message #1 from AF1.

[0399] Among them, the initialization response message #1 is used to indicate that AF1 confirms to perform the first federated learning task with the assigned identity next. For the specific federated learning process, reference can be made to the relevant description of the above method 300.

[0400] S514, the NWDAF sends an initialization request message #2 (i.e., an example of the first request message) to AF2. Correspondingly, AF2 receives the initialization request message #2 from the NWDAF.

[0401] Among them, the parameters included in the initialization request message #2 and their specific interpretations can be referred to the relevant description of the first request message of the above method 400.

[0402] S515, AF2 sends an initialization response message #2 to the NWDAF. Correspondingly, the NWDAF receives the initialization response message #2 from AF2.

[0403] Among them, the initialization response message #2 is used to indicate that AF2 confirms to perform the first federated learning task with the assigned identity next. For the specific federated learning process, reference can be made to the relevant description of the above method 300.

[0404] Based on the above solution, as the task initiator, the NWDAF determines the first entity and the first identity participating in the execution of the first federated learning task by obtaining at least one candidate entity and the identity information of at least one candidate entity, and triggering a confirmation to at least one candidate entity whether it agrees to perform the first federated learning task. This implementation method triggers the initialization request of the federated learning task by the task initiator. By informing the first entity of the first identity in the execution of the first federated learning task, the execution efficiency of the first federated learning task can be effectively improved.

[0405] It should be understood that the above Figure 5 is that the NWDAF, as the task initiator, initiates the first federated learning task. That is to say, the NWDAF determines the first entity and the first identity for executing the first federated learning task, and triggers a federated learning task initialization request to be sent to the first entity. Compared with Figure 5 , Figure 6 it is that AF1, as the task initiator, initiates the first federated learning task. At the same time, a third network element (such as a coordinator) is introduced. When the task initiator AF1 determines the first entity and the first identity for executing the first federated learning task, it triggers a request to the coordinator to organize or coordinate the first entity to complete the first federated learning task. That is, it is the coordinator that triggers a federated learning task initialization request to be sent to the first entity.

[0406] Figure 6 is a schematic flowchart of a communication method 600 provided by an embodiment of the present application. As Figure 6 shown, taking the first network element (e.g., task initiator) as AF1, the second network element (e.g., entity discovery function network element) as NRF, the third network element as the coordinator, and the task participants as NWDAF and AF2 as the execution entities for interaction. This method can be regarded as a further refinement of the above method 400. It should be understood that Figure 6 the embodiments shown in Figure 4 can be coupled with each other and can be referred to each other. Therefore, the relevant descriptions in the above method 400 also apply to this implementation manner. There may be the same or similar technical means between the two. The content already described in the embodiment shown in Figure 4 will not be elaborated again. This method includes the following multiple steps. For the parts not elaborated in detail, reference can be made to the above method 400 or existing protocols.

[0407] S601, NWDAF, AF1, and AF2 initiate a registration process to NRF. For the specific registration process, reference can be made to the relevant description of the above method 200. The parameters carried in the registration process can be referred to the relevant description of step S501 of the above method 500.

[0408] S602, AF1 sends an entity discovery request message (i.e., an example of the third request message) to NRF. Correspondingly, NRF receives the entity discovery request message from AF1.

[0409] Among them, the parameters included in the entity discovery request message and their specific interpretations can be referred to the relevant description of the third request message in step S402 of the above method 400.

[0410] S603, NRF determines at least one candidate entity and the identity information of at least one candidate entity, such as NWDAF and AF2, according to the first analysis ID and the first federated learning group ID. The specific implementation manner can be referred to the relevant description of step S420 of the above method 400.

[0411] Optionally, NRF can also determine at least one candidate entity and the identity information of at least one candidate entity according to other parameters carried in the entity discovery request message, such as information of vertical federated alliance and / or the first interoperability identifier.

[0412] S604, NRF sends an entity discovery response message (i.e., an example of the third response message) to AF1. Correspondingly, AF1 receives the entity discovery response message from NRF.

[0413] Among them, the parameters included in the entity discovery response message and their specific interpretations can be referred to the relevant description of the third response message in step S404 of the above method 400.

[0414] S605, AF1 determines at least one candidate entity and the identity information of at least one candidate entity.

[0415] Further, AF1 may request at least one candidate entity (such as NWDAF and AF2) to prepare for executing the first federated learning task. For details, refer to the following steps S606 - S613.

[0416] S606, AF1 sends a federated learning task request message #1 (i.e., an example of the second request message) to NWDAF. Correspondingly, NWDAF receives the federated learning task request message #1 from AF1.

[0417] Exemplarily, AF1 may send the federated learning task request message #1 to NWDAF through the NEF.

[0418] Among them, the parameters included in the federated learning request task request message #1 and their specific interpretations can refer to the relevant descriptions of the second request message in the above method 400.

[0419] S607, NWDAF agrees to participate in the federated learning task according to the assigned identity.

[0420] S608, NWDAF agrees to the designated negotiator, that is, NWDAF agrees that the negotiator organizes or negotiates NWDAF to execute the first federated learning task.

[0421] S609, NWDAF sends a federated learning request task response message #1 (i.e., an example of the second response message) to AF1. Correspondingly, AF1 receives the federated learning request task response message #1 from NWDAF.

[0422] Exemplarily, NWDAF may send the federated learning request task response message #1 to AF1 through the NEF.

[0423] Among them, the parameters included in the federated learning request task request response #1 and their specific interpretations can refer to the relevant descriptions of the second response message in the above method 400.

[0424] S610, AF1 sends a federated learning request task request message #2 (i.e., an example of the second request message) to AF2. Correspondingly, AF2 receives the federated learning request task request message #2 from AF1.

[0425] Exemplarily, AF1 may send the federated learning request task request message #2 to AF2 through the NEF.

[0426] Among them, the parameters included in the federated learning request task request message #2 and their specific interpretations can refer to the relevant descriptions of the second request message in the above method 400.

[0427] S611, AF2 agrees to participate in the federated learning task according to the assigned identity.

[0428] S612, AF2 agrees to the designated negotiator, that is, AF2 agrees that the negotiator organizes or negotiates for AF2 to execute the first federated learning task.

[0429] S613, AF2 sends a federated learning request task response message #2 (i.e., an example of the second response message) to AF1. Correspondingly, AF1 receives the federated learning request task response message #1 from AF2.

[0430] Exemplarily, AF2 can send the federated learning request task response message #1 to AF1 through the NEF.

[0431] Among them, the parameters and specific interpretations included in the federated learning request task request response #2 can refer to the relevant descriptions of the second response message in the above method 400.

[0432] It should be noted that the specific implementation manners of the above steps S606 - S609 or steps S610 - S613 can refer to the relevant descriptions of steps S411 - S413 in the above method 400. For the sake of brevity, they will not be elaborated here.

[0433] S614, AF1 determines the first entity and the first identity.

[0434] Exemplarily, AF1 can determine the first entity and the first identity according to the feedback in steps S609 and S613. For example, the first entity includes the NWDAF and AF2. Correspondingly, the first identity of the NWDAF can be the main participant, that is, the NWDAF provides labels and data in the execution of the first federated learning task, and the first identity of AF2 can be the secondary participant, that is, AF2 provides data in the execution of the first federated learning task.

[0435] Optionally, AF1 can participate in the execution of the first federated learning task or not participate in the execution of the federated learning task. This application does not make any limitations in this regard. That is, the first entity executing the first federated learning task can include AF1, AF2, and the NWDAF, or the first entity executing the first federated learning task can include the NWDAF and AF2.

[0436] Further, after determining the first entity and the first identity, AF1 can trigger the request for the negotiator to organize the first entity to execute the first federated learning task, specifically including the following steps S615 - S622.

[0437] S615, AF1 sends a federated learning task creation request message (i.e., an example of the sixth message) to the negotiator. Correspondingly, the negotiator receives the federated learning task creation request message from AF1.

[0438] Among them, the parameters included in the federated learning task creation request message and their specific interpretations can refer to the relevant descriptions of the sixth request message in the above method 400.

[0439] S616. The negotiator sends an initialization request message #1 to the NWDAF. Correspondingly, the NWDAF receives the initialization request message #1 from AF1.

[0440] Among them, the parameters included in the initialization request message #1 and their specific interpretations can refer to the relevant descriptions of the first request message in the above method 400.

[0441] S617. The NWDAF verifies the negotiator ID.

[0442] Exemplarily, the NWDAF compares whether the sender in step S616 is the same as the negotiator determined in step S608. If they are the same, the following step S618 is executed. If they are not the same, the NWDAF can reject the initialization request in step S616. In this implementation method, by judging the negotiator ID assigned by the task initiator AF1 and the negotiator ID triggering the federated learning task initialization request, the NWDAF can avoid executing the federated learning task initiated by a malicious negotiator, ensuring network security while reducing the processing load or signaling overhead of the NWDAF.

[0443] S618. The NWDAF sends an initialization response message #1 to AF1. Correspondingly, AF1 receives the initialization response message #1 from the NWDAF.

[0444] Among them, the initialization response message #1 is used to instruct AF1 to confirm that it will execute the first federated learning task with the assigned identity next. The specific federated learning process can refer to the relevant descriptions of the above method 300.

[0445] S619. The negotiator sends an initialization request message #2 to AF2. Correspondingly, AF2 receives the initialization request message #2 from the negotiator.

[0446] Among them, the parameters included in the initialization request message #2 and their specific interpretations can refer to the relevant descriptions of the first request message in the above method 400.

[0447] S620. AF2 verifies the negotiator ID.

[0448] S621. AF2 sends an initialization response message #2 to AF1. Correspondingly, AF1 receives the initialization response message #2 from AF2.

[0449] Among them, the initialization response message #2 is used to instruct AF2 to confirm that it will execute the first federated learning task with the assigned identity next. For the specific federated learning process, reference can be made to the relevant description of the above method 300.

[0450] Among them, for the specific implementation manners of the above steps S619 - S621, reference can be made to the relevant description of the above steps S616 - S618.

[0451] S622, the negotiator sends a federated learning task creation response message to AF1. Correspondingly, AF1 receives the federated learning task creation response message from the negotiator.

[0452] Among them, the federated learning task creation response message is used to indicate that the negotiator has organized or negotiated for NWDAF and AF2 to execute or complete the first federated learning task.

[0453] Based on the above solution, AF1, as the task initiator, determines the first entity and the first identity participating in the execution of the first federated learning task by obtaining at least one candidate entity and the identity information of at least one candidate entity, and triggering a confirmation to at least one candidate entity on whether it agrees to execute the first federated learning task. This implementation manner is for the negotiator to create or coordinate the first federated learning task, trigger an initialization request for the first federated learning task, and by informing the first entity of its first identity in the execution of the first federated learning task, the execution efficiency of the first federated learning task can be effectively improved.

[0454] It should be understood that the above Figure 6 is that AF1, as the task initiator, initiates the first federated learning task, AF1 confirms with at least one candidate entity whether it agrees to participate in the federated learning task with the assigned identity. When NWDAF determines the first entity and the first identity for executing the first federated learning task, it triggers a request to the third network element (such as the coordinator) to organize or coordinate the first entity to complete the first federated learning task, that is, the coordinator triggers a federated learning task initialization request to the first entity. Compared with Figure 6 , Figure 7 is that AF1, as the task initiator, initiates the first federated learning task, then introduces the third network element (such as the coordinator), the coordinator confirms with at least one candidate entity whether it agrees to participate in the federated learning task with the assigned identity, and when NWDAF determines the first entity and the first identity for executing the first federated learning task, the coordinator then triggers a federated learning task initialization request to the first entity.

[0455] Figure 7 is a schematic flowchart of the communication method 700 provided by an embodiment of the present application, as Figure 7As shown, the first network element is the initiator, the second network element is the NRF, and the third network element is the coordinator as the execution entity for interaction. This method can be regarded as a further refinement of the above method 400. It should be understood that Figure 7 the embodiments shown in Figure 4 and the embodiments shown in Figure 4 can be coupled to each other and can be referred to each other. Therefore, the relevant descriptions in the above method 400 also apply to this implementation manner. There may be the same or similar technical means between the two. The content already described in the embodiments shown in

[0456] will not be elaborated. This method includes the following multiple steps. For the parts not elaborated in detail, reference can be made to the above method 400 or existing protocols.

[0456] S701, NWDAF, AF1, and AF2 initiate a registration process to the NRF. For the specific registration process, reference can be made to the relevant description of the above method 200. The parameters carried in the registration process can be referred to the relevant description of step S601 of the above method 600.

[0457] S702, AF1 sends a federated learning task request message #1 to the coordinator. Correspondingly, the coordinator receives the federated learning task request message #1 from AF1.

[0458] Exemplarily, AF1 can send the federated learning task request message #1 to the coordinator through the NEF.

[0459] Among them, the parameters included in the federated learning request task request message #1 and their specific interpretations can be referred to the relevant description of the second request message of the above method 400.

[0460] Furthermore, the coordinator triggers the acquisition of at least one candidate entity for executing the federated learning task and the identity information of at least one candidate entity. For example, the coordinator can pre-configure or configure through signaling to obtain candidate entities for executing the federated learning task; or, the coordinator requests the NRF to obtain candidate entities for executing the federated learning task, which specifically includes the following steps S703 - S705.

[0461] S703, the coordinator sends an entity discovery request message (i.e., an example of the third request message) to the NRF. Correspondingly, the NRF receives the entity discovery request message from the coordinator.

[0462] Among them, the parameters included in the entity discovery request message and their specific interpretations can be referred to the relevant description of the third request message in step S402 of the above method 400.

[0463] S704, the NRF determines at least one candidate entity and the identity information of at least one candidate entity according to the first analysis ID and the first federated learning group ID, such as NWDAF and AF2. The specific implementation manner can be referred to the relevant description of step S420 of the above method 400.

[0464] Optionally, the NRF may also determine at least one candidate entity and the identity information of at least one candidate entity according to other parameters carried in the entity discovery request message, such as information on vertical federated alliances and / or the first interoperability identifier.

[0465] In S705, the NRF sends an entity discovery response message (i.e., an example of the third response message) to the coordinator. Correspondingly, the coordinator receives the entity discovery response message from the NRF.

[0466] Among them, the parameters included in the entity discovery response message and their specific interpretations can refer to the relevant descriptions of the third response message in step S404 of the above method 400.

[0467] Further, in the case where the coordinator determines at least one candidate entity (such as NWDAF and AF2) and the identity information of at least one candidate entity, it may confirm with at least one candidate entity whether it is willing or agrees to participate in the execution of the first federated learning task with the assigned identity, specifically including the following steps S706 - S713.

[0468] In S706, the coordinator sends a federated learning task request message #2 (i.e., an example of the eighth request message) to the NWDAF. Correspondingly, the NWDAF receives the federated learning task request message #2 from the coordinator.

[0469] Exemplarily, the coordinator may send the federated learning task request message #2 to the NWDAF through the NEF.

[0470] Among them, the parameters included in the federated learning request task request message #2 and their specific interpretations can refer to the relevant descriptions of the eighth request message of the above method 400.

[0471] In S707, the NWDAF agrees to participate in the federated learning task with the assigned identity.

[0472] In S708, the NWDAF agrees to the designated negotiator, that is, the NWDAF agrees that the negotiator organizes or negotiates the NWDAF to execute the first federated learning task.

[0473] In S709, the NWDAF sends a federated learning request task response message #2 (i.e., an example of the eighth response message) to the coordinator. Correspondingly, the NWDAF receives the federated learning request task response message #2 from the coordinator.

[0474] Exemplarily, the NWDAF may send the federated learning request task response message #2 to the coordinator through the NEF.

[0475] Among them, the parameters and specific interpretations included in the federated learning request task request response #2 can refer to the relevant descriptions of the eighth response message of the above method 400.

[0476] S710, the coordinator sends a federated learning request task request message #3 (i.e., an example of the eighth request message) to AF2. Correspondingly, AF2 receives the federated learning request task request message #3 from the coordinator.

[0477] Exemplarily, the coordinator can send the federated learning request task request message #3 to AF2 through the NEF.

[0478] Among them, the parameters and specific interpretations included in the federated learning request task request message #3 can refer to the relevant descriptions of the eighth request message of the above method 400.

[0479] S711, AF2 agrees to participate in the federated learning task according to the assigned identity.

[0480] S712, AF2 agrees to the designated negotiator, that is, AF2 agrees that the negotiator organizes or negotiates AF2 to execute the first federated learning task.

[0481] S713, AF2 sends a federated learning request task response message #3 (i.e., an example of the eighth response message) to AF1. Correspondingly, AF1 receives the federated learning request task response message #3 from AF2.

[0482] Exemplarily, AF2 can send the federated learning request task response message #3 to the coordinator through the NEF.

[0483] Among them, the parameters and specific interpretations included in the federated learning request task request response #3 can refer to the relevant descriptions of the eighth response message of the above method 400.

[0484] It should be noted that the specific implementation methods of the above steps S706 - S709 or steps S710 - S713 can refer to the relevant descriptions of steps S411 - S413 of the above method 400. For the sake of brevity, they will not be elaborated here.

[0485] S714, the negotiator determines at least one candidate entity and the identity information of at least one candidate entity.

[0486] Exemplarily, the negotiator can determine at least one candidate entity and the identity information of at least one candidate entity based on the feedback in steps S709 and S713. It should be noted that the at least one candidate entity and the identity information of at least one candidate entity determined in step S714 may be exactly the same as or different from the at least one candidate entity and the identity information of at least one candidate entity obtained in step S705. For example, the at least one candidate entity and the identity information of at least one candidate entity determined in step S714 are included in the at least one candidate entity and the identity information of at least one candidate entity obtained in step S705. This application does not make any limitations in this regard.

[0487] S715. The negotiator sends the federated learning task response message #1 to AF1. Correspondingly, AF1 receives the federated learning task response message #1 from the negotiator. The federated learning task response message #1 includes at least one candidate entity and the identity information of at least one candidate entity determined in step S714.

[0488] S716. AF1 determines the first entity and the first identity.

[0489] It should be understood that the first entity belongs to at least one candidate entity. For example, AF1 determines the first entity and the first identity from at least one candidate entity and the identity information of at least one candidate entity.

[0490] Exemplarily, if at least one candidate entity includes NWDAF and AF2, where NWDAF can provide labels and data in the execution of the first federated learning task, and AF2 can provide data in the execution of the first federated learning task, then AF1 can determine that the first entity includes NWDAF and AF2. Among them, the first identity of NWDAF can be the main participant. For example, NWDAF provides labels and data in the execution of the first federated learning task, and the first identity of AF2 can be the secondary participant. For example, AF2 provides data in the execution of the first federated learning task.

[0491] Exemplarily, if at least one candidate entity includes NWDAF and AF2, where both NWDAF and AF2 can provide labels and data in the execution of the first federated learning task, then AF1 can determine that the first entity includes NWDAF and AF2. Among them, the first identity of NWDAF can be the main participant, providing labels and data in the execution of the first federated learning task, and the first identity of AF2 can be the secondary participant, providing data in the execution of the first federated learning task.

[0492] Optionally, AF1 can participate in the execution of the first federated learning task or not participate in the execution of the federated learning task. This application does not make any limitations in this regard.

[0493] Optionally, after determining the first entity and the first identity, AF1 can finally notify the negotiator of the first entity and the first identity participating in the execution of the first federated learning task. For details, please refer to the following step S716.

[0494] S716. AF1 sends a first message to the negotiator. Correspondingly, the negotiator receives the first message from AF1.

[0495] The first message indicates the first entity and the first identity. It should be noted that if the first entity and the first identity finally determined by AF1 are exactly the same as the candidate entity and the identity information of the candidate entity sent by the negotiator in step S715, AF1 may not execute step S716. If the first entity and the first identity finally determined by AF1 are inconsistent with the candidate entity and the identity information of the candidate entity sent by the negotiator in step S715, AF1 needs to execute step S716 to notify the first entity and the first identity that will finally execute the first federated learning task.

[0496] Further, after the negotiator obtains the first entity and the first identity, the negotiator can trigger a request for the first entity to execute the first federated learning task, which specifically includes the following steps S717 - S720.

[0497] S717. The negotiator sends an initialization request message #1 to NWDAF. Correspondingly, NWDAF receives the initialization request message #1 from AF1.

[0498] The parameters included in the initialization request message #1 and their specific interpretations can refer to the description of the first request message of the above method 400.

[0499] S718. NWDAF sends an initialization response message #1 to AF1. Correspondingly, AF1 receives the initialization response message #1 from NWDAF.

[0500] The initialization response message #1 is used to indicate that AF1 confirms to execute the first federated learning task with the assigned identity next. The specific federated learning process can refer to the relevant description of the above method 300.

[0501] S719. The negotiator sends an initialization request message #2 to AF2. Correspondingly, AF2 receives the initialization request message #2 from the negotiator.

[0502] The parameters included in the initialization request message #2 and their specific interpretations can refer to the description of the first request message of the above method 400.

[0503] S720. AF2 sends an initialization response message #2 to AF1. Correspondingly, AF1 receives the initialization response message #2 from AF2.

[0504] Among them, the initialization response message #2 is used to instruct AF2 to confirm that the first federated learning task will be executed with the assigned identity next. For the specific federated learning process, reference can be made to the relevant description of the above method 300.

[0505] Based on the above solution, as the task initiator, AF1 triggers the negotiator to obtain at least one candidate entity and the identity information of at least one candidate entity, and the negotiator confirms with at least one candidate entity whether it agrees to execute the first federated learning task. Finally, AF1 determines the first entity and the first identity participating in the execution of the first federated learning task from at least one candidate entity and the identity information of at least one candidate entity. This implementation method allows the negotiator to obtain at least one candidate entity and the identity information of at least one candidate entity, and trigger a federated learning task initialization request. By informing the first entity of its first identity in the execution of the first federated learning task, the execution efficiency of the first federated learning task can be effectively improved.

[0506] It should be understood that the above Figure 5 is that NWDAF, as the task initiator, initiates the first federated learning task. That is, NWDAF determines the first entity and the first identity for executing the first federated learning task, and triggers the sending of a federated learning task initialization request to the first entity. Compared with Figure 5 Figure 8 it is that NWDAF, as the task initiator, initiates the first federated learning task, and at the same time introduces a third network element (such as a coordinator), and the coordinator organizes or coordinates the first entity to complete the first federated learning task.

[0507] Figure 8 is a schematic flowchart of the communication method 800 provided by an embodiment of the present application. As Figure 8 shown, with the first network element as the initiator, the second network element as the NRF, and the third network element as the coordinator as the execution subject for interaction, this method can be regarded as a further refinement of the above method 400. It should be understood that Figure 8 the embodiment shown in Figure 4 and the embodiment shown in Figure 4 can be coupled to each other and can be referred to each other. Therefore, the relevant descriptions in the above method 400 also apply to this implementation method. There may be the same or similar technical means between the two. The content already described in the embodiment shown in

[0508] will not be elaborated again. This method includes the following multiple steps. For the parts not elaborated in detail, reference can be made to the above method 400 or existing protocols.

[0509] In S802, the NWDAF sends an entity discovery request message (i.e., an example of the third request message) to the NRF. Correspondingly, the NRF receives the entity discovery request message from the NWDAF.

[0510] In S803, the NRF determines at least one candidate entity and the identity information of at least one candidate entity, such as AF1 and AF2, according to the first analysis ID and the first federated learning group ID.

[0511] Optionally, the NRF can also determine at least one candidate entity and the identity information of at least one candidate entity according to other parameters carried in the entity discovery request message, such as information on vertical federated alliances and / or the first interoperability identifier.

[0512] In S804, the NRF sends an entity discovery response message (i.e., an example of the third response message) to the NWDAF. Correspondingly, the NWDAF receives the entity discovery response message from the NRF.

[0513] Among them, the specific implementation manners of the above steps S801 - S804 can refer to the relevant descriptions of the above steps S501 - S504.

[0514] In this implementation manner, a coordinator (i.e., an example of the third network element) is introduced to organize or coordinate the first entity to execute the first federated learning task, thereby improving the execution efficiency of the first federated learning task in an orderly manner. Therefore, before executing the first federated learning task, the NWDAF can confirm with the coordinator whether it agrees or is willing to coordinate the first entity to execute the first federated learning task. For details, see the following steps S805 - S806.

[0515] In S805, the NWDAF sends a coordination request message (i.e., an example of the fifth request message) to the coordinator. Correspondingly, the coordinator receives the coordination request message from the NWDAF.

[0516] Among them, the parameters included in the coordination request message and their specific interpretations can refer to the relevant descriptions of the fifth request message of the above method 400.

[0517] In S806, the coordinator sends a coordination response message (i.e., an example of the fifth response message) to the NWDAF. Correspondingly, the NWDAF receives the coordination response message from the coordinator.

[0518] Among them, the parameters included in the coordination request message and their specific interpretations can refer to the relevant descriptions of the fifth response message of the above method 400.

[0519] Furthermore, the NWDAF can request at least one candidate entity (such as AF1 and AF2) to prepare to execute the first federated learning task. For details, see the following steps S807 - S814.

[0520] S807, the NWDAF sends a federated learning request task request message #1 (i.e., an example of the second request message) to AF1. Correspondingly, AF1 receives the federated learning request task request message #1 from the NWDAF.

[0521] Exemplarily, the NWDAF may send the federated learning request task request message #1 to AF1 through the NEF.

[0522] Among them, the parameters included in the federated learning request task request message #1 and their specific interpretations can refer to the relevant descriptions of the second request message in the above method 400.

[0523] S808, AF1 agrees to participate in the federated learning task according to the assigned identity.

[0524] S809, AF1 agrees to the designated negotiator, that is, AF1 agrees that the negotiator organizes or negotiates AF1 to execute the first federated learning task.

[0525] S810, AF1 sends a federated learning request task response message #1 (i.e., an example of the second response message) to the NWDAF. Correspondingly, the NWDAF receives the federated learning request task response message #1 from AF1.

[0526] Exemplarily, AF1 may send the federated learning request task response message #1 to the NWDAF through the NEF.

[0527] Among them, the parameters included in the federated learning request task request response #1 and their specific interpretations can refer to the relevant descriptions of the second response message in the above method 400.

[0528] S811, the NWDAF sends a federated learning request task request message #2 (i.e., an example of the second request message) to AF2. Correspondingly, AF2 receives the federated learning request task request message #2 from the NWDAF.

[0529] Exemplarily, the NWDAF may send the federated learning request task request message #2 to AF2 through the NEF.

[0530] Among them, the parameters included in the federated learning request task request message #2 and their specific interpretations can refer to the relevant descriptions of the second request message in the above method 400.

[0531] S812, AF2 agrees to participate in the federated learning task according to the assigned identity.

[0532] S813, AF2 agrees to the designated negotiator, that is, AF2 agrees that the negotiator organizes or negotiates AF2 to execute the first federated learning task.

[0533] S814. AF2 sends a federated learning request task response message #2 (i.e., an example of the second response message) to the NWDAF. Correspondingly, the NWDAF receives the federated learning request task response message #1 from AF2.

[0534] Exemplarily, AF2 may send the federated learning request task response message #2 to the NWDAF through the NEF.

[0535] Among them, the parameters included in the federated learning request task request response #2 and the specific interpretations can refer to the relevant descriptions of the second response message in the above method 400.

[0536] It should be noted that the specific implementation manners of the above steps S807 - S814 can refer to the relevant descriptions of steps S411 - S413 in the above method 400. For the sake of brevity, they will not be elaborated here.

[0537] S815. The NWDAF determines the first entity and the first identity.

[0538] Furthermore, after determining the first entity and the first identity, the NWDAF may trigger a request for the first entity to execute the first federated learning task, which specifically includes the following steps S816 - S820.

[0539] S816. The NWDAF sends the first entity and the first identity to the coordinator. Correspondingly, the coordinator receives the first entity and the first identity from the NWDAF.

[0540] S817. The NWDAF sends an initialization request message #1 (i.e., an example of the first request message) to AF1. Correspondingly, AF1 receives the initialization request message #1 from the NWDAF.

[0541] Among them, the parameters included in the initialization request message #1 and the specific interpretations can refer to the relevant descriptions of the first request message in the above method 400.

[0542] S818. AF1 sends an initialization response message #1 to the NWDAF. Correspondingly, the NWDAF receives the initialization response message #1 from AF1.

[0543] Among them, the initialization response message #1 is used to instruct AF1 to confirm that it will execute the first federated learning task with the assigned identity next. The specific federated learning process can refer to the relevant descriptions of the above method 300.

[0544] S819. The NWDAF sends an initialization request message #2 (i.e., an example of the first request message) to AF2. Correspondingly, AF2 receives the initialization request message #2 from the NWDAF.

[0545] Among them, the parameters included in the initialization request message #2 and their specific interpretations can refer to the relevant descriptions of the first request message in the above method 400.

[0546] S820, AF2 sends an initialization response message #2 to the NWDAF. Correspondingly, the NWDAF receives the initialization response message #2 from AF2.

[0547] Among them, the initialization response message #2 is used to indicate that AF2 confirms to execute the first federated learning task with the assigned identity next. The specific federated learning process can refer to the relevant descriptions of the above method 300.

[0548] Based on the above solution, as the task initiator, the NWDAF determines the first entity and the first identity participating in the execution of the first federated learning task by obtaining at least one candidate entity and the identity information of at least one candidate entity, and triggering the confirmation with the coordinator and at least one candidate entity on whether they agree to execute the first federated learning task. This implementation method triggers the initialization request of the federated learning task by the NWDAF, effectively improves the execution efficiency of the first federated learning task by informing the first entity of its first identity in the execution of the first federated learning task, and organizing or coordinating the first entity to execute the first federated learning task by the coordinator.

[0549] The above is combined with Figures 1 to 8 describes the method-side embodiments of the communication method of the present application. Next, the device-side embodiments of the communication method of the present application will be described in detail in combination with Figure 9 and Figure 10 It should be understood that the description of the device embodiments corresponds to the description of the method embodiments. Therefore, the parts not described in detail can refer to the previous method embodiments.

[0550] Figure 9 is a schematic diagram of a communication device 1000 provided by an embodiment of the present application. As Figure 9 shown, the communication device 1000 includes a communication module 1002 and a processing module 1001. The communication device 1000 can be the first network element, or a communication device applied to the first network element or used in matching with the first network element and capable of implementing the method executed by the first network element, such as a chip, a chip system or a circuit; or, the communication device 1000 can be the second network element, or a communication device applied to the second network element or used in matching with the second network element and capable of implementing the method executed by the second network element, such as a chip, a chip system or a circuit; or, the communication device 1000 can be the third network element, or a communication device applied to the third network element or used in matching with the third network element and capable of implementing the method executed by the third network element, such as a chip, a chip system or a circuit.

[0551] Among them, the communication module 1002 can also be referred to as a transceiver module, transceiver, transceiver unit, or transceiver device, etc. The processing module 1001 can also be referred to as a processor, processing board, processing unit, or processing device, etc. Optionally, the communication module 1002 is used to perform the sending operation and receiving operation of the first network element, the second network element, or the third network element in the above method. The device in the communication module 1002 for implementing the receiving function can be regarded as a receiving unit, and the device in the communication module 1002 for implementing the sending function can be regarded as a sending unit, that is, the communication module 1002 includes a receiving unit and a sending unit.

[0552] When the communication device 1000 is applied to the first network element, the processing module 1001 can be used to implement the processing function of the first network element in the above embodiments, and the communication module 1002 can be used to implement the transceiver function of the first network element in the above embodiments.

[0553] When the communication device 1000 is applied to the second network element, the processing module 1001 can be used to implement the processing function of the second network element in the above embodiments, and the communication module 1002 can be used to implement the transceiver function of the second network element in the above embodiments.

[0554] When the communication device 1000 is applied to the third network element, the processing module 1001 can be used to implement the processing function of the third network element in the above embodiments, and the communication module 1002 can be used to implement the transceiver function of the third network element in the above embodiments.

[0555] In addition, it should be noted that the foregoing communication module and / or processing module can be implemented by a virtual module. For example, the processing module can be implemented by a software functional unit or a virtual device, and the communication module can be implemented by a software function or a virtual device. Alternatively, the processing module or the communication module can also be implemented by a physical device. For example, if the device is implemented by a chip / circuit (such as an integrated circuit or a logic circuit, etc.). The communication module can be an input / output circuit and / or a communication interface, performing an input operation (corresponding to the foregoing receiving operation) and an output operation (corresponding to the foregoing sending operation); the processing module is an integrated processor or microprocessor or circuit (such as an integrated circuit or a logic circuit, etc.).

[0556] The division of modules in this application is illustrative, only a logical function division. In actual implementation, there may be other division methods. In addition, in each example of this application, each functional module can be integrated in one processor, or can exist separately physically, or two or more modules can be integrated in one module. The above integrated module can be implemented in the form of hardware or in the form of a software functional module.

[0557] Figure 10 It is a schematic diagram of another communication device 2000 provided by an embodiment of the present application. As Figure 10As shown, optionally, the communication device 2000 may be a chip or a chip system. Optionally, in this application, a chip system may be composed of chips, or may include chips and other discrete devices.

[0558] The communication device 2000 can be used to implement the functions of any network element (such as the first network element, the second network element, or the third network element) in the communication system described in the foregoing examples. The communication device 2000 may include a communication interface 2030 and a processor 2010. Among them, the communication device 2000 can interact with other devices through the communication interface 2030. Exemplarily, the communication interface 2030 may be a transceiver, a circuit, a bus, a module, a pin, or other types of communication interfaces. When the communication device 2000 is a chip-type device or circuit, the communication interface 2030 in the device 2000 may also be an input / output circuit, which can input information (or receive information) and output information (or send information). The processor 2010 is an integrated processor, a microprocessor, an integrated circuit, or a logic circuit, etc. The processor can determine the output information according to the input information.

[0559] Optionally, the processor 2010 is coupled to a memory. The memory may be located within the device, or the memory may be integrated with the processor, or the memory may also be located outside the device. For example, the communication device 2000 may further include at least one memory 2020. The memory 2020 stores the necessary computer programs, computer programs or instructions and / or data in any of the above examples; the processor 2010 may execute the computer programs stored in the memory 2020 to complete the methods in any of the above examples.

[0560] The coupling in this application is an indirect coupling or communication connection between devices, units or modules, which can be electrical, mechanical or other forms, and is used for information interaction between devices, units or modules. The processor 2010 may cooperate with the memory 2020 and the communication interface 2030. In this application, the specific connection medium between the above-mentioned processor 2010, memory 2020 and communication interface 2030 is not limited.

[0561] Optionally, as Figure 10 shown, the processor 2010, the memory 2020 and the communication interface 2030 are interconnected through a bus 2040. Optionally, the bus may include types of buses such as an address bus, a data bus, and a control bus. In addition, for ease of representation, Figure 10 only one bus 2040 is shown in, but it does not mean that there is only one bus or one type of bus.

[0562] It should be understood that the processor mentioned in the embodiments of the present application may be the following devices or a part of the circuit for processing functions in the following devices: a central processing unit (CPU), and may also 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 may be a microprocessor or the processor may also be any conventional processor, etc.

[0563] It should also be understood that the memory mentioned in the embodiments of the present application may be a volatile memory and / or a non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM). For example, the RAM may be used as an external cache. By way of example and not limitation, the RAM includes the following various forms: static random access memory (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0564] It should be noted that when the processor is a general-purpose processor, DSP, ASIC, FPGA, or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, the memory (storage module) may be integrated in the processor.

[0565] It should also be noted that the memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0566] An embodiment of the present application also provides a computer-readable storage medium, on which computer instructions for implementing the methods executed by at least one of the first network element, the second network element, or the third network element in the above method embodiments are stored.

[0567] An embodiment of the present application also provides a computer program product, including instructions, which when executed by a computer, implement the methods executed by at least one of the first network element, the second network element, or the third network element in the above method embodiments.

[0568] An embodiment of the present application also provides a communication system, which includes at least one of the first network element, the second network element, or the third network element in the above embodiments.

[0569] The explanations and beneficial effects of the relevant content in any of the above-mentioned devices can refer to the corresponding method embodiments provided above, and will not be described herein again.

[0570] For the convenience of understanding the above embodiments provided by the present application, the following points are explained:

[0571] 1) In the present application, if there is no special description and logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced to each other. The technical features in different embodiments can be combined to form new embodiments according to their internal logical relationships.

[0572] 2) In the present application, "at least one" means one or more, and "a plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone, where A and B can be singular or plural. In the text description of the present application, the character " / " generally represents an "or" relationship between the preceding and following associated objects. "At least one (item)" or its similar expression refers to any combination of these items, including any combination of single item (item) or plural items (items). For example, at least one (item) of a, b, and c can represent: a, or, b, or, c, or, a and b, or, a and c, or, b and c, or, a, b, and c. Where a, b, and c can be single or multiple respectively.

[0573] 3) In this application, "first", "second", and various numerical numbers (e.g., #1, #2, etc.) are used for distinction for convenience of description, and do not limit the scope of the embodiments of this application. For example, they are used to distinguish different messages, etc., rather than to describe a specific order or sequence. It should be understood that the objects described in this way can be interchanged under appropriate circumstances so as to be able to describe solutions other than the embodiments of this application.

[0574] 4) In this application, descriptions such as "when...", "in the case of...", and "if" all refer to the device making corresponding processing under a certain objective circumstance, which does not limit time, and does not require the device to have a judgment action when implemented, nor does it mean there are other limitations.

[0575] 5) In this application, "used to indicate" may include directly indicating and indirectly indicating. When describing that a certain indication information is used to indicate A, it may include that the indication information directly indicates A or indirectly indicates A, rather than meaning that A must be carried in the indication information.

[0576] The indication methods involved in the embodiments of this application should be understood to cover various methods that can enable the party to be indicated to obtain the indication information. The indication information can be sent as a whole, or divided into multiple sub-information and sent separately, and the sending periods and / or sending opportunities of these sub-information can be the same or different. This application does not limit, for example, the sending method.

[0577] The "indication information" in the embodiments of this application can be explicit indication, that is, directly indicated by signaling, or according to the parameters indicated by signaling, combined with other rules or combined with other parameters or obtained by derivation. It can also be implicit indication, that is, obtained according to rules or relationships, or according to other parameters, or by derivation. This application does not make specific limitations on this.

[0578] 6) In this application, "protocol" may refer to standard protocols in the communication field. For example, it may include 5G protocols, NR protocols, and related protocols applied to future communication systems. This application does not make limitations on this. "Predefined" may include predefined. For example, protocol definition. "Preconfigured" can be implemented by pre-saving corresponding codes, tables, or other ways that can be used to indicate relevant information in the device. This application does not limit its implementation method, for example.

[0579] 7) In this application, "communication" can also be described as "data transmission", "information transmission", "data processing", etc. "Transmission" includes "sending" and "receiving".

[0580] In this application, a configuration may refer to a signaling configuration, and may also be described as configuring signaling. For example, the signaling configuration may be configured by a network device sending signaling, and these signals may be radio resource control (RRC) messages, downlink control information (DCI), or system information block (SIB). For another example, the signaling configuration may be pre-configured, where the pre-configuration defines or configures the values of corresponding parameters in advance in a protocol manner, and may be stored in the device during communication. This application does not make any limitations in this regard.

[0581] In various embodiments of this application, the magnitudes of the sequence numbers of the above processes do 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.

[0582] In this application, on the premise of no logical contradiction, the examples can refer to each other. For example, the methods and / or terms between method embodiments can refer to each other. For example, the functions and / or terms between device embodiments can refer to each other. For example, the functions and / or terms between device examples and method examples can refer to each other.

[0583] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0584] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0585] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in electrical, mechanical, or other forms.

[0586] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0587] In addition, in each embodiment of this application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0588] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it 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 part of this 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 for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs that can store program codes.

[0589] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A communication method, characterized in that, including: A first network element determines a first entity and a first identity, where the first identity is used to indicate the identity of the first entity in performing a first federated learning task; The first network element sends a first request message to the first entity, where the first request message is used to request the first entity to perform the first federated learning task with the first identity.

2. The method according to claim 1, characterized in that, The first identity includes a main participant or a secondary participant, where The main participant supports providing labels in the first federated learning task, and the labels correspond to the first federated learning task; or, the main participant supports providing labels and data in the first federated learning task; The secondary participant supports providing data in the first federated learning task.

3. The method according to claim 1 or 2, characterized in that The method further includes: The first network element obtains at least one candidate entity and identity information of the at least one candidate entity, where the identity information of the at least one candidate entity is used to indicate the identity supported by the at least one candidate entity in the first federated learning task; The first network element determining the first entity and the first identity includes: The first network element determines the first entity and the first identity from the at least one candidate entity and the identity information of the at least one candidate entity.

4. The method according to claim 3, characterized in that, The first network element obtaining at least one candidate entity and identity information of the at least one candidate entity includes: The first network element receives the at least one candidate entity and the identity information of the at least one candidate entity from a second network element, and the second network element supports discovering entities that perform the first federated learning task.

5. The method according to claim 4, wherein The first network element receiving the at least one candidate entity and the identity information of the at least one candidate entity from the second network element includes: The first network element sends a third request message to the second network element, where the third request message is used to obtain entities that support performing the first federated learning task; The first network element receives a third response message from the second network element, and the third response message includes the at least one candidate entity and the identity information of the at least one candidate entity.

6. The method according to any one of claims 1 to 5, characterized in that, Before the first network element determines the first entity and the first identity, the method further includes: The first network element sends a second request message to the first entity, where the second request message is used to request the first entity to prepare to perform the first federated learning task; The first network element receives a second response message from the first entity, and the second response message is used to indicate that the first entity agrees to perform the first federated learning task with the first identity.

7. The method according to claim 6, wherein The second request message includes at least one identity supported by the first entity in the first federated learning task, and the at least one identity includes the first identity.

8. The method according to claim 6 or 7, characterized in that The second response message includes the first identity.

9. The method according to any one of claims 6 to 8, characterized in that The second request message further includes information about a third network element, and the third network element supports coordinating the first entity to perform the first federated learning task.

10. The method according to any one of claims 1 to 9, the method further includes: The first network element obtains information of a third network element and identity information of the third network element, where the identity information of the third network element is used to indicate that the third network element supports coordinating the first entity to execute the first federated learning task.

11. The method according to claim 10, where the first network element obtains information of a third network element and identity information of the third network element, including: The first network element receives the information of the third network element and the identity information of the third network element from a second network element.

12. The method according to claim 10 or 11, characterized in that The method further includes: The first network element sends a fifth request message to the third network element, where the fifth request message is used to request the third network element to coordinate the first entity to execute the first federated learning task; The first network element receives a fifth response message from the third network element, where the fifth response message is used to indicate that the third network element agrees to coordinate the first entity to execute the first federated learning task.

13. The method according to any one of claims 10 to 12, characterized in that, Before the first network element sends a first request message to the first entity, the method further includes: The first network element sends a sixth request message to the third network element, where the sixth request message is used to request to create or coordinate the first federated learning task, and the sixth request message includes the first entity and the first identity.

14. The method according to any one of claims 1 to 13, characterized in that, Before the first network element determines the first entity and the first identity, the method further includes: The first network element obtains a first analysis identifier, where the first analysis identifier corresponds to the first federated learning task; The first network element triggers the first federated learning task according to the first analysis identifier.

15. A communication method, characterized in that, including: The second network element receives a third request message, where the third request message is used to obtain an entity that supports executing the first federated learning task, and the second network element supports discovering an entity that executes the first federated learning task; The second network element determines at least one candidate entity and identity information of the at least one candidate entity, where the identity information of the at least one candidate entity is used to indicate an identity supported by the at least one candidate entity in the first federated learning task; The second network element sends a third response message, where the third response message includes the at least one candidate entity and the identity information of the at least one candidate entity.

16. The method according to claim 15, wherein The third request message includes multiple types, and the multiple types correspond to the at least one candidate entity.

17. The method according to claim 15 or 16, characterized in that, The method further includes: The second network element sends information of a third network element and identity information of the third network element to the first network element, where the identity information of the third network element is used to indicate that the third network element supports coordinating the first entity to execute the first federated learning task.

18. The method according to any one of claims 15 to 17, characterized in that, The method further includes: The second network element receives a registration request message of at least one second entity, where the registration request message includes identity information of the at least one second entity, the at least one second entity supports executing the first federated learning task, and the at least one second entity includes the at least one candidate entity.

19. The method according to any one of claims 15 to 18, characterized in that, The second network element receiving the third request message includes: The second network element receives the third request message from the first network element, where the first network element supports initiating the first federated learning task; or, The second network element receives the third request message from the third network element, where the third network element supports coordinating the first entity to execute the first federated learning task, and the at least one candidate entity includes the first entity.

20. A communication method, characterized in that, Comprising: The third network element obtains the first entity and the first identity, where the first identity is used to indicate the identity of the first entity in executing the first federated learning task, and the third network element supports coordinating the first entity to execute the first federated learning task; The third network element sends a seventh request message to the first entity, where the seventh request message is used to request to execute the first federated learning task, and the seventh request message includes the first identity.

21. The method according to claim 20, characterized in that, The third network element obtaining the first entity and the first identity includes: The third network element receives the first entity and the first identity from the first network element, where the first network element supports initiating the first federated learning task.

22. The method according to claim 21, wherein The third network element receiving the first entity and the first identity from the first network element includes: The third network element receives a sixth request message from the first network element, where the sixth request message is used to request to create or coordinate the first federated learning task, and the sixth request message includes the first entity and the first identity.

23. The method according to claim 21 or 22, characterized in that, Before the third network element receives the first entity and the first identity information from the first network element, the method further includes: The third network element sends the at least one candidate entity and the identity information of the at least one candidate entity to the first network element, where the identity information of the at least one candidate entity is used to indicate the identity supported by the at least one candidate entity in the first federated learning task, and the at least one candidate entity includes the first entity.

24. The method according to claim 23, wherein Before the third network element sends the at least one candidate entity and the identity information of the at least one candidate entity to the first network element, the method further includes: The third network element receives a fourth request message from the first network element, where the fourth request message is used to request to prepare for executing the first federated learning task; The third network element obtains the at least one candidate entity and the identity information of the at least one candidate entity; The third network element sends an eighth request message to the at least one candidate entity, where the eighth request message is used to request the at least one candidate entity to execute the first federated learning task, and the eighth request message includes the identity information of the at least one candidate entity; The third network element receives an eighth response message from the at least one candidate entity, where the eighth response message is used to indicate that the at least one candidate entity agrees to support executing the first federated learning task with the identity indicated by the identity information of the candidate entity.

25. The method according to claim 24, wherein The eighth request message further includes information of the third network element, and the eighth response message is further used to indicate agreement for the third network element to negotiate the first federated learning task.

26. The method according to claim 24 or 25, characterized in that, The third network element obtaining the at least one candidate entity and the identity information of the at least one candidate entity includes: The third network element receives the at least one candidate entity and the identity information of the at least one candidate entity from the second network element, where the second network element supports discovering entities for executing the first federated learning task.

27. The method according to claim 26, wherein The third network element receives the at least one candidate entity and the identity information of the at least one candidate entity from the second network element, including: The third network element sends a third request message to the second network element, where the third request message is used to obtain an entity that supports executing the first federated learning task; The third network element receives a third response message from the second network element, and the third response message includes the at least one candidate entity and the identity information of the at least one candidate entity.

28. The method according to any one of claims 21 to 27, characterized in that, Before the third network element receives the first entity and the first identity from the first network element, the method further includes: The third network element receives a fifth request message from the first network element, where the fifth request message is used to request the third network element to coordinate the first entity to execute the first federated learning task; The third network element sends a fifth response message to the first network element, and the fifth response message is used to indicate that the third network element agrees to coordinate the first entity to execute the first federated learning task.

29. A communication device, characterized in that, Including at least one module, where the at least one module is used to execute the method according to any one of claims 1 to 28.

30. A communication device, characterized in that, Including: A processor, where the processor is coupled to a memory; The processor is configured to execute a computer program stored in the memory, so that the device executes the method according to any one of claims 1 to 28.

31. A computer-readable storage medium, characterized in that, Computer program code or instructions are stored on the computer-readable storage medium, and when the computer program code or instructions run on a computer, the method according to any one of claims 1 to 28 is executed.

Images