Communication method and apparatus for vertical federated learning, and readable storage medium

Abstract

The present application relates to the technical field of communications, and particularly relates to a communication method and apparatus in vertical federated learning (VFL), and a readable storage medium. The method comprises: an AF sending a first request message to an NEF, wherein the first request message comprises a first identifier of a first NWDAF and an identifier of a VFL task; the NEF sending a second request message to a UDR, wherein the second request message comprises the identifier of the VFL task; the UDR returning a second response message to the NEF, wherein the second response message comprises a mapping relationship between a first identifier of each NWDAF among one or more NWDAFs and a second identifier of each MWDAF, the one or more NWDAFs participating in the VFL task, and the one or more NWDAFs comprising the first NWDAF; and the NEF sending a third request message to the first NWDAF, wherein the third request message comprises a second identifier of the first NWDAF. By means of the present application, not only can the service provided by an NWDAF be obtained, the security of the NWDAF can also be improved.

Description

Communication methods, devices, and readable storage media in vertical federated learning

[0001] This application claims priority to Chinese Patent Application No. 202411599556.7, filed on November 8, 2024, entitled “Communication Method, Apparatus and Readable Storage Medium in Vertical Federated Learning”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of communication technology, and in particular to a communication method, apparatus and readable storage medium in vertical federated learning. Background Technology

[0003] Federated learning (FL) can be divided into horizontal federated learning, vertical federated learning, and federated transfer learning. Vertical federated learning (VFL) is typically suitable for scenarios where participating entities (e.g., network elements in a vertical federation) have significant overlap in training samples but minimal overlap in data features. Currently, the 3rd Generation Partnership Project (3GPP) supports training and / or inference using vertical federated learning to enable mobile network-related analytics services.

[0004] An application function (AF) can act as a VFL server, and this AF can be trusted or untrusted. When an AF acting as a VFL server is untrusted, how to securely request services from a network data analytics function (NWDAF) acting as a VFL client is a problem currently being studied by those skilled in the art. Summary of the Invention

[0005] This application provides a communication method, apparatus, and readable storage medium in vertical federated learning, which can obtain services provided by VFL clients (e.g., NWDAF) and improve the security of VFL clients when the VFL server (e.g., AF) is untrusted.

[0006] The present application is described below from different aspects. It should be understood that the different implementation methods and beneficial effects described below can be referenced from each other.

[0007] Firstly, this application provides a communication method in vertical federated learning. This method can be executed by a communication device, which may be a network exposure function (NEF) or a chip within the NEF. For ease of description, the method is illustrated here using an NEF as the communication device. The method includes: the NEF receiving a first request message from an AF, the first request message including a first identifier of a first NWDAF and an identifier of a VFL task; the NEF sending a second request message to a data processing network element, the second request message including the identifier of the VFL task; the NEF receiving a second response message from the data processing network element, the second response message including a mapping relationship between the first identifier and the second identifier of each of one or more NWDAFs participating in the VFL task, and the one or more NWDAFs including the first NWDAF; and the NEF then sending a third request message to the first NWDAF according to the above mapping relationship. The VFL task is used to train a model. The first request message is used to request services provided by the model trained by the VFL task. Alternatively, the first request message is used to request services provided by the model corresponding to the identifier of the VFL task. For example, the first request message can be a VFL prediction / inference request or a sample alignment request. The second request message can be used to obtain the first and second identifiers of the NWDAFs participating in the VFL task. The third request message is also used to request services provided by the model, or the third request message is used to request services from the model corresponding to the identifier of the VFL task.

[0008] For example, the third request message includes a second identifier corresponding to the first identifier of the first NWDAF.

[0009] In this application, the first identifier can be an identifier used to identify the NWDAF outside the public land mobile network (PLMN), or an identifier open to the outside of the PLMN for identifying the NWDAF, or the first identifier used to identify the NWDAF for an AF outside the PLMN, or the first identifier used to identify the NWDAF for an untrusted AF. For example, the first identifier can be a temporary NF instance ID. Exemplarily, the same NWDAF can have multiple first identifiers, which can be used by different devices outside the PLMN, such as multiple different AFs identifying the NWDAF. These multiple first identifiers can also be used to identify the NWDAF for the same AF, with the AF using different first identifiers in different services or tasks (e.g., VFL tasks).

[0010] Correspondingly, the second identifier in this application can be an identifier used within the PLMN to identify the NWDAF, or in other words, the second identifier is used to identify the NWDAF for an AF within the PLMN, or the second identifier is used to identify the NWDAF for a trusted network element. For example, the second identifier can be an NF instance ID. Further explanation of the first and second identifiers will not be repeated below.

[0011] In this application, "data processing network element" can be a network element with data storage and / or data management functions. For example, the "data processing network element" in this application may include unified data repository (UDR) function and / or unified data management (UDM) function.

[0012] Upon receiving a service request (e.g., a VFL prediction / inference / sample alignment request) from an AF, the NEF of this application can use the VFL task identifier carried in the service request to query the data processing network element for the first and second identifiers corresponding to the VFL task identifier. The second identifier corresponding to the VFL task identifier includes the second identifier of the first NWDAF. Then, the NEF can send a service request (e.g., a VFL prediction / inference request) to the first NWDAF based on the queried second identifier of the first NWDAF. In this way, the AF can both obtain the services provided by the VFL client (e.g., the NWDAF) and improve the security of the VFL client.

[0013] In conjunction with the first aspect, in one possible implementation, before the NEF sends the second request message to the data processing network element, the method further includes: the NEF determining whether a second identifier corresponding to the first identifier of the first NWDAF exists locally. If no second identifier corresponding to the first identifier of the first NWDAF exists locally, the NEF sends the second request message to the data processing network element. Alternatively, the NEF may determine whether a first identifier and a second identifier corresponding to the identifier of the aforementioned VFL task exist locally. If no first identifier and a second identifier corresponding to the identifier of the VFL task exist locally, the NEF sends the second request message to the data processing network element.

[0014] When the NEF of this application does not have a second identifier corresponding to the first identifier of the first NWDAF locally, it queries the data processing network element for the second identifier corresponding to the first identifier of the first NWDAF so that the AF can obtain the services provided by the NWDAF.

[0015] In conjunction with the first aspect, in one possible implementation, after the NEF receives the second response message from the data processing network element, the method further includes: storing the mapping relationship between the first identifier of each NWDAF and the second identifier of each NWDAF in the one or more NWDAFs.

[0016] For example, after the NEF receives the second response message from the data processing network element, the method further includes: the NEF receiving a fourth request message from the AF, the fourth request message including the first identifier of the second NWDAF. If the one or more NWDAFs include the second NWDAF, it indicates that the NEF has a second identifier corresponding to the first identifier of the second NWDAF locally, and there is no need to query the data processing network element again. The NEF can send a fifth request message to the second NWDAF based on the second identifier of the second NWDAF. The fifth request message is used to request the services provided by the model trained by the above-mentioned VFL task. For example, the fifth request message includes the second identifier corresponding to the first identifier of the second NWDAF stored locally by the NEF. For example, the fourth request message can be a VFL prediction / inference request.

[0017] For the same VFL task, when the NEF receives the first service request (such as the first request message mentioned above), it indexes the first and second identifiers of all NWDAFs participating in the VFL task using the VFL task's identifier. When the NEF subsequently receives another service request for the same VFL task (such as the fourth request message mentioned above), it does not need to query the data processing network element again. Therefore, for multiple service requests for the same VFL task (such as the first and fourth request messages), this application only needs to query the data processing network element once to obtain the first and second identifiers of all NWDAFs participating in the VFL task, which can improve the efficiency of the NEF in processing service requests.

[0018] In conjunction with the first aspect, in one possible implementation, before the NEF receives the first request message from the AF, the method further includes: the NEF receiving a first message from the AF, the first message including a first identifier of the NWDAF participating in the VFL task and an identifier of the VFL task, wherein the NWDAF participating in the VFL task includes a first NWDAF; the NEF obtaining a second identifier of the NWDAF participating in the VFL task; and the NEF sending a storage request to the data processing network element, the storage request including the first identifier of the NWDAF participating in the VFL task, the identifier of the VFL task, and the second identifier of the NWDAF participating in the VFL task. The data processing network element stores the mapping relationship between the first identifier, the second identifier, and the identifier of the VFL task of the NWDAF participating in the VFL task. The NWDAF participating in the VFL task includes one or more of the aforementioned NWDAFs.

[0019] It is understood that the NWDAF participating in the VFL task in this application can be a VFL member determined by the AF for that VFL task, or it can be the NWDAF of the VFL task corresponding to the identifier participating in the VFL task determined by the AF. Further details will not be elaborated below. The VFL task includes one or more of the following: training task, prediction task, training task and prediction task, etc. For example, if the VFL task is a prediction task, then all NWDAFs participating in the VFL task can participate in training the model.

[0020] It can also be understood that there can be one or more NWDAFs participating in the VFL task. When there are multiple NWDAFs participating in the VFL task, the first message mentioned above includes a first identifier of each NWDAF participating in the VFL task and an identifier of the VFL task.

[0021] This application's NEF stores information about NWDAFs participating in VFL tasks in the data processing network element. When some NEFs are unavailable, the AF can send a first request message to other available NEFs, including a first identifier of the first NWDAF and an identifier of the VFL task. Other NEFs can use the identifier of the VFL task to query the data processing network element for a second identifier corresponding to the first identifier of the first NWDAF, and then send a third request message to the first NWDAF to request it to provide services. This solves the problem of network element services corresponding to the first identifier being unavailable when some NEF services are unavailable, reduces network failures, and improves the continuity of service requests based on anonymized network element identifiers.

[0022] In conjunction with the first aspect, in one possible implementation, before the NEF receives the first message from the AF, the method further includes: the NEF receiving a network element discovery request from the AF, the network element discovery request being used to discover candidate NWDAFs, the network element discovery request including an analysis identifier; the NEF assigning a first identifier to the candidate NWDAF; and the NEF sending a network element discovery response to the AF, the network element discovery response including the first identifier of the candidate NWDAF. The candidate NWDAFs include those participating in the aforementioned VFL task. The VFL task is used to train the model corresponding to the analysis identifier.

[0023] For example, the NWDAF participating in the VFL task also includes a second NWDAF.

[0024] Secondly, this application provides a communication method in vertical federated learning, which can be executed by a communication device, which can be a data processing network element or a chip within the data processing network element. For ease of description, this is illustrated using a data processing network element as an example. The method includes: the data processing network element receiving a second request message from a NEF, the second request message including an identifier of the VFL task, the second request message being used to obtain a first identifier and a second identifier of the NWDAFs participating in the VFL task; the data processing network element sending a second response message to a first NEF, the second response message including a mapping relationship between the first identifier and the second identifier of each of one or more NWDAFs participating in the VFL task. These one or more NWDAFs include a first NWDAF. The VFL task is used to train a model.

[0025] The "data processing network element" in this application can be a network element with data storage and / or data management functions. For example, the "data processing network element" in this application can include unified data repository (UDR) functions and / or unified data management (UDM) functions.

[0026] In conjunction with the second aspect, in one possible implementation, the first identifier mentioned above can be used to identify NWDAF outside the PLMN, and the second identifier mentioned above can be used to identify NWDAF inside the PLMN.

[0027] In conjunction with the second aspect, in one possible implementation, before the data processing network element receives the second request message from the NEF, the method further includes: the data processing network element receiving a storage request from the first NEF, the storage request including a first identifier and a second identifier of the NWDAFs participating in the aforementioned VFL task, and an identifier of the aforementioned VFL task, wherein the NWDAFs participating in the aforementioned VFL task include the first NWDAF; the data processing network element storing the mapping relationship between the first identifier, the second identifier, and the identifier of the aforementioned VFL task of the NWDAFs participating in the aforementioned VFL task. Wherein, the NWDAFs participating in the aforementioned VFL task include one or more of the aforementioned NWDAFs.

[0028] In conjunction with the second aspect, in one possible implementation, the method further includes: a data processing network element receiving a fifth request message from a second NEF, the fifth request message including an identifier for the aforementioned VFL task; the data processing network element sending a fifth response message to the second NEF, the fifth response message including a mapping relationship between a first identifier and a second identifier of each of one or more NWDAFs participating in the aforementioned VFL task, the one or more NWDAFs including a first NWDAF. The fifth request message can be used to obtain the first and second identifiers of the NWDAFs participating in the VFL task.

[0029] For example, the NWDAF participating in the above VFL task also includes a second NWDAF.

[0030] Thirdly, this application provides a communication method in longitudinal federated learning, which can be executed by a communication device, which may be an AF or a chip within the AF. For ease of description, the method is illustrated here using an AF as an example. The method includes: the AF generating a first request message and sending the first request message to NEF1. The first request message includes a first identifier of a first NWDAF and an identifier of a VFL task. The VFL task is used to train the model. The first request message is used to request services provided by the model. For example, the first request message may be a VFL prediction / inference request.

[0031] For example, the first identifier mentioned above may be used to identify NWDAF outside the PLMN.

[0032] In conjunction with the third aspect, in one possible implementation, after the AF sends the first request message to NEF1, the method further includes: the AF sending a fourth request message to NEF1, the fourth request message including the first identifier of the second NWDAF. This fourth request message is used to request services provided by the model. For example, the fourth request message could be a VFL prediction / inference request.

[0033] In conjunction with the third aspect, in one possible implementation, before the AF sends the first request message to NEF1, the method further includes: the AF sending a first message to NEF2, the first message including a first identifier of the NWDAF participating in the VFL task and an identifier of the VFL task. The NWDAF participating in the VFL task includes a first NWDAF. After receiving the first message, NEF2 can obtain a second identifier of the NWDAF participating in the VFL task; then it sends a storage request to the data processing network element, the storage request including the first identifier of the NWDAF participating in the VFL task, the identifier of the VFL task, and the second identifier of the NWDAF participating in the VFL task. The data processing network element stores the mapping relationship between the first identifier and the second identifier of the NWDAF participating in the VFL task and the identifier of the VFL task.

[0034] In conjunction with the third aspect, in one possible implementation, before the AF sends the first message to NEF2, the method further includes: the AF sending a network element discovery request to NEF2, the network element discovery request being used to discover candidate NWDAFs, the network element discovery request including an analysis identifier; and the AF receiving a network element discovery response from NEF2, the network element discovery response including a first identifier of the candidate NWDAF. The candidate NWDAFs include those participating in the aforementioned VFL task. This VFL task is used to train the model corresponding to the analysis identifier.

[0035] Fourthly, this application provides a communication device, which may be a NEF, a data processing network element, an AF, or a chip therein. The communication device includes units and / or modules for performing the methods provided by any one of the first to third aspects, or any possible implementation thereof, such as a transceiver module and / or a processing module. The transceiver module is used to transmit and receive various information or signaling, thus also achieving the beneficial effects (or advantages) of the methods provided by any one of the first to third aspects.

[0036] Fifthly, this application provides a communication method in vertical federated learning, which can be executed by a communication device, which may be a NEF or a chip within the NEF. For ease of description, the method is illustrated here using an NEF as the communication device. The method includes: the NEF receiving a first request message from an AF, the first request message including a first identifier of an NWDAF participating in the VFL task; the NEF sending a first network element discovery request to a network repository function (NRF), the first network element discovery request including the first identifier of the NWDAF participating in the VFL task; the NEF receiving a first network element discovery response from the NRF, the first network element discovery response including the first identifier of the NWDAF participating in the VFL task and its corresponding second identifier; and the NEF sending a second request message to the NWDAF participating in the VFL task based on the second identifier corresponding to the first identifier of the NWDAF participating in the VFL task. The VFL task is used to train a model. The first request message is used to request services provided by the model; for example, the first request message may be a VFL inference / prediction request. The second request message is also used to request services provided by the model.

[0037] For example, the second request message includes a second identifier corresponding to the first identifier of the NWDAF participating in the VFL task.

[0038] For example, the first identifier mentioned above may be used to identify NWDAF outside the PLMN, and the second identifier mentioned above may be used to identify NWDAF inside the PLMN.

[0039] After receiving a service request (e.g., a VFL prediction / inference request) from an AF, the NEF in this application can use a first identifier carried in the service request to query the NRF for a second identifier corresponding to the first identifier. Then, based on the retrieved second identifier, the NEF can send a service request (e.g., a VFL prediction / inference request) to the NWDAF participating in the VFL task. This allows the AF to obtain services provided by the VFL client (e.g., the NWDAF) while also improving the security of the VFL client.

[0040] In conjunction with the fifth aspect, in one possible implementation, before the NEF sends the first network element discovery request to the NRF, the method further includes: the NEF determining whether a second identifier corresponding to the first identifier of an NWDAF participating in the aforementioned VFL task exists locally. If no second identifier corresponding to the first identifier of an NWDAF participating in the VFL task exists locally, the NEF sends the first network element discovery request to the NRF.

[0041] In conjunction with the fifth aspect, in one possible implementation, the aforementioned first network element discovery request further includes first information. This first information indicates the type of the first identifier carried in the first network element discovery request, or indicates that the first identifier carried in the first network element discovery request is external to the PLMN and used to identify the NWDAF. After receiving the first network element discovery request, the NRF can determine, based on the first information, that the first identifier carried in the first request message is a temporary identifier, not an instance identifier. Thus, since the first identifier and the second identifier are constructed in the same way, it is easier for the NRF to distinguish between the first identifier and the second identifier, reducing the possibility of the NRF making an error in finding the second identifier based on the first identifier.

[0042] In conjunction with the fifth aspect, in one possible implementation, before the NEF receives the first request message from the AF, the method further includes: the NEF sending a second network element discovery request to the NRF, the second network element discovery request including second information and / or information of the AF, the second information being used to instruct (for candidate NWDAFs) to assign a first identifier; the second network element discovery request also including a VFL capability type, the second network element discovery request being used to discover candidate NWDAFs; the NEF receiving a second network element discovery response from the NRF, the second network element discovery response including a first identifier and a second identifier of the candidate NWDAF; the candidate NWDAF matching a VFL capability type; the NEF sending the first identifier of the candidate NWDAF to the AF, the candidate NWDAF including NWDAFs participating in VFL tasks.

[0043] Sixthly, this application provides a communication method in vertical federated learning, which can be executed by a communication device, which may be a network repository function (NRF) or a chip within the NRF. For ease of description, this is illustrated using an NRF as an example. The method includes: the NRF receiving a first network element discovery request from NEF1, the first network element discovery request including a first identifier of an NWDAF participating in the VFL task; the NRF determining a second identifier of an NWDAF participating in the VFL task based on the first identifier of the NWDAF participating in the VFL task; and the NRF sending a first network element discovery response to NEF1, the first network element discovery response including the first identifier of the NWDAF participating in the VFL task and its corresponding second identifier.

[0044] For example, the first identifier mentioned above may be used to identify NWDAF outside the PLMN, and the second identifier mentioned above may be used to identify NWDAF inside the PLMN.

[0045] In conjunction with the sixth aspect, in one possible implementation, the aforementioned first network element discovery request further includes first information. This first information indicates the type of the first identifier carried in the first network element discovery request, or indicates that the first identifier carried in the first network element discovery request is external to the PLMN and used to identify the NWDAF. After receiving the first network element discovery request, the NRF can determine, based on the first information, that the first identifier carried in the first request message is a temporary identifier, not an instance identifier. Thus, since the first identifier and the second identifier are constructed in the same way, it is easier for the NRF to distinguish between the first identifier and the second identifier, reducing the possibility of the NRF making an error in finding the second identifier based on the first identifier.

[0046] In conjunction with the sixth aspect, in one possible implementation, before the NRF receives the first network element discovery request from NEF1, the method further includes: the NRF receiving a second network element discovery request from NEF2, the second network element discovery request including second information and / or AF information, the second information being used to indicate the allocation of a first identifier; the second network element discovery request also including a VFL capability type, the second network element discovery request being used to discover candidate NWDAFs; the NRF determining the candidate NWDAF and its second identifier based on the VFL capability type; the candidate NWDAF matching the VFL capability type; the NRF determining to allocate a first identifier to the candidate NWDAF based on the second information or the AF information; the NRF sending a second network element discovery response to NEF2, the second network element discovery response including the first and second identifiers of the candidate NWDAF. Wherein, the candidate NWDAF includes the aforementioned NWDAFs participating in the VFL task.

[0047] In this application, the NRF assigns a first identifier to the candidate NWDAF, stores the first identifier and second identifier of the candidate NWDAF, and sends the first identifier of the candidate NWDAF to the AF through the NEF. In this way, even if some NEFs are unavailable, the AF can still obtain the services provided by the NWDAF corresponding to the first identifier through other NEFs. Furthermore, since the first identifier is used externally by the PLMN to identify the NWDAF, the security of the NWDAF can be improved.

[0048] Seventhly, this application provides a communication device, which may be a NEF or NRF, or a chip therein. The communication device includes units and / or modules for performing the methods provided by any possible implementation of the fifth aspect, or the sixth aspect, or any of the aspects described above, such as a transceiver module and / or a processing module. The transceiver module is used to send and receive various information or signaling, thus also achieving the beneficial effects (or advantages) of the methods provided by the fifth or sixth aspect.

[0049] Eighthly, this application provides a communication method in vertical federated learning. This method can be executed by a communication device, which may be an NRF or a chip within the NRF. For ease of description, an NRF is used as an example for illustration. The method includes: the NRF receiving a network element discovery request from a NEF, the network element discovery request including information about an AF (Active Element), the network element discovery request being used to discover candidate NWDAFs; the NRF determining the candidate NWDAF based on the network element discovery request; and the NRF sending a network element discovery response to the NEF, the network element discovery response including information about the candidate NWDAF. The candidate NWDAF is allowed to be discovered by the AF. Alternatively, the candidate NWDAF can provide services or VFL (Virtual Functional Element) services to the AF.

[0050] For example, the information of an AF may include one or more of the following: the identifier of the AF, the type of the AF, or the domain name of the AF.

[0051] In the network element discovery process of this application, the NEF discovers candidate NWDAFs to the NRF and informs the NRF of the AF information. The NRF filters NWDAFs based on the AF information, which can filter out NWDAFs that include AFs in the allowed NFc, thereby improving the security of NWAF.

[0052] In conjunction with aspect eight, in one possible implementation, the information of the aforementioned AF is the type, identifier, or domain name of the AF. The allowed NFc list of the aforementioned candidate NWDAFs includes the identifier of the AF. Alternatively, the allowed NF domains of the aforementioned candidate NWDAFs include the domain name of the AF. Alternatively, the allowed NF types of the aforementioned candidate NWDAFs include the type of the AF.

[0053] For example, the information of the AF mentioned above is the identifier of the AF. NRF can verify whether the allowed NFc list in the locally stored NWDAF profile includes the identifier of the AF. If the allowed NFc list in the profile of an NWDAF includes the identifier of the AF, NRF determines that the NWDAF is a candidate NWDAF.

[0054] For another example, the information of the AF mentioned above is the type of the AF. The NRF can verify whether the allowed NFc types of the locally stored NWDAF include the type of the AF. If the allowed NFc types of an NWDAF include the type of the AF, the NRF determines that the NWDAF is a candidate NWDAF.

[0055] For another example, the information of the AF mentioned above is the domain name of the AF. The NRF can verify whether the allowed NFc Domains of the locally stored NWDAF include the domain name of the AF. If the allowed NFc Domains of an NWDAF include the domain name of the AF, the NRF determines that the NWDAF is a candidate NWDAF.

[0056] In conjunction with aspect eight, in one possible implementation, the aforementioned network element discovery request may further include one or more of the following: analysis identifier, or VFL capability type. The NRF can determine a candidate NWDAF based on the AF information and VFL capability type in the aforementioned network element discovery request. The candidate NWDAF also matches the VFL capability type. The analysis identifier supported by the candidate NWDAF is the same as the analysis identifier in the network element discovery request.

[0057] In conjunction with aspect eight, in one possible implementation, the aforementioned network element discovery request may further include indication information, which can be used to indicate the allocation of a first identifier. The NRF can determine, based on the indication information in the network element discovery request, to allocate a first identifier to a candidate NWDAF. The information of the candidate NWDAF includes the first identifier and the second identifier of the candidate NWDAF.

[0058] For example, the first identifier may be used to identify the NWDAF outside the PLMN, and the second identifier may be used to identify the NWDAF inside the PLMN.

[0059] Ninthly, this application provides a communication method in vertical federated learning, which can be executed by a communication device, which may be a NEF or a chip within the NEF. For ease of description, the method is illustrated here using an NEF as the communication device. The method includes: the NEF sending a network element discovery request to the NRF, the network element discovery request including information about the AF (Active Functional Element), the network element discovery request being used to discover candidate NWDAFs; the NEF receiving a network element discovery response from the NRF, the network element discovery response including information about the candidate NWDAF. The allowed NF consumers of the candidate NWDAF include the AF.

[0060] For example, the information of an AF may include one or more of the following: the identifier of the AF, the type of the AF, or the domain name of the AF.

[0061] In conjunction with aspect nine, in one possible implementation, the information of the aforementioned AF is the type, identifier, or domain name of the AF. The allowed NFc list of the aforementioned candidate NWDAF includes the identifier of the AF. Alternatively, the allowed NF Domains of the aforementioned candidate NWDAF includes the domain name of the AF. Alternatively, the allowed NF types of the aforementioned candidate NWDAF includes the type of the AF.

[0062] In conjunction with aspect nine, in one possible implementation, the aforementioned network element discovery request may further include one or more of the following: an analysis identifier, or a VFL capability type. The candidate NWDAF also matches the VFL capability type. The analysis identifier supported by the candidate NWDAF is the same as the analysis identifier in the network element discovery request.

[0063] In conjunction with aspect nine, in one possible implementation, the aforementioned network element discovery request may further include indication information, which may be used to indicate the allocation of a first identifier. The information of the aforementioned candidate NWDAF includes the first identifier and the second identifier of the candidate NWDAF.

[0064] For example, the first identifier may be used to identify the NWDAF outside the PLMN, and the second identifier may be used to identify the NWDAF inside the PLMN.

[0065] Tenthly, this application provides a communication device, which may be a NEF or NRF, or a chip therein. The communication device includes units and / or modules for performing the methods provided by any possible implementation of the eighth aspect, or the ninth aspect, or any of these aspects, such as a transceiver module and / or a processing module. The transceiver module is used to transmit and receive various information or signaling, thus also achieving the beneficial effects (or advantages) of the methods provided by the eighth or ninth aspect.

[0066] Eleventhly, this application provides a communication method in longitudinal federated learning, which can be executed by a communication device, which may be an AF or a chip within the AF. For ease of description, the communication device is exemplified as an AF. The method includes: the AF sending a VFL training request to the NEF, the VFL training request including a first identifier of the NWDAF participating in the VFL task, the VFL task being used to train a model, and the VFL training request being used to request the NWDAF participating in the VFL task to train the model; the AF receiving a VFL training response from the NEF; the AF storing a mapping relationship between the identifier of the NEF and the identifier of the VFL task; and the AF sending a VFL inference request to the NEF corresponding to the identifier of the VFL task according to the mapping relationship. The VFL inference request includes a first identifier of the NWDAF participating in the VFL task. The VFL inference request is used to request services provided by the model.

[0067] For example, the first identifier may be used to identify the NWDAF outside the PLMN.

[0068] This application reduces the likelihood of VFL inference failures and improves network stability by constraining the VFL training and inference processes to use the same NEF.

[0069] In a twelfth aspect, this application provides a communication device, which may be an AF or a chip in an AF. The communication device includes units and / or modules for performing the methods provided in the eleventh aspect or any possible implementation thereof, such as a transceiver module and / or a processing module. The transceiver module is used to send and receive various information or signaling, and thus can also achieve the beneficial effects (or advantages) of the methods provided in the eleventh aspect.

[0070] In a thirteenth aspect, this application provides a communication device comprising a processor configured to execute the method shown in any one of the first to third aspects, or the fifth aspect, or the sixth aspect, or the eighth aspect, or the ninth aspect, or the eleventh aspect, or any possible implementation thereof. Alternatively, the processor may execute a program stored in a memory, wherein when the program is executed, the method shown in any one of the first to third aspects, or the fifth aspect, or the sixth aspect, or the eighth aspect, or the ninth aspect, or the eleventh aspect, or any possible implementation thereof, is executed.

[0071] In conjunction with aspect thirteen, in one possible implementation, the memory is located outside the aforementioned communication device.

[0072] In conjunction with aspect thirteen, in one possible implementation, the memory is located within the aforementioned communication device.

[0073] In this application, the processor and memory can also be integrated into a single device, that is, the processor and memory can be integrated together.

[0074] In conjunction with aspect thirteen, in one possible implementation, the communication device further includes a transceiver for sending or receiving various types of information.

[0075] In a fourteenth aspect, this application provides a communication device, which can be implemented as a chip or as a device. The communication device may include a processor and an interface circuit, the processor and the interface circuit being connected. The interface circuit is used for exchanging (or sending and receiving, or inputting and outputting) information or data, and the processor is used to execute program instructions, causing the communication device to perform the methods described in any one of the first to third aspects, or the fifth aspect, or the sixth aspect, or the eighth aspect, or the ninth aspect, or the eleventh aspect, or any possible implementation thereof. The interface circuit may be a communication interface or a transceiver. The transceiver may be a radio frequency module in the communication device, or a combination of a radio frequency module and an antenna, or an input / output interface of a chip or circuit.

[0076] In conjunction with the fourteenth aspect, in one possible implementation, the communication device further includes a memory that is circuitically connected to the processor. Exemplarily, the processor and the memory may be physically independent units, or the memory may be integrated with the processor.

[0077] In a fifteenth aspect, this application provides a readable storage medium storing program instructions that, when executed on a communication device, cause the communication device to perform the method described in any one of the first to third aspects, or the fifth aspect, or the sixth aspect, or the eighth aspect, or the ninth aspect, or the eleventh aspect, or any possible implementation of any one of these aspects.

[0078] In a sixteenth aspect, this application provides a program product containing instructions that, when executed, cause the method described in any one of the first to third aspects, or the fifth aspect, or the sixth aspect, or the eighth aspect, or the ninth aspect, or the eleventh aspect, or any possible implementation thereof, to be performed.

[0079] In a seventeenth aspect, this application provides a communication system comprising a network element (NEF) and a data processing network element. The NEF can be used to perform the method described in the first aspect or any possible implementation thereof, and the data processing network element can be used to perform the method described in the second aspect or any possible implementation thereof. The communication system also includes an array element (AF). The AF can be used to perform the method described in the third aspect or any possible implementation thereof.

[0080] In an eighteenth aspect, this application provides a communication system comprising an NRF and a NEF. The NEF can be used to perform the methods described in the fifth aspect or any possible implementation thereof, and the NRF can be used to perform the methods described in the sixth aspect or any possible implementation thereof.

[0081] In a nineteenth aspect, this application provides a communication system including an NRF. The NRF can be used to perform the methods described in the eighth aspect or any possible implementation thereof. The communication system also includes a NEF, which can be used to perform the methods described in the ninth aspect or any possible implementation thereof.

[0082] In a twentieth aspect, this application provides a communication system including an AF. The AF can be used to perform the method described in the eleventh aspect or any of the possible implementations of the eleventh aspect.

[0083] The technical effects achieved in the above aspects can be referred to each other or to the beneficial effects in the method embodiments shown below, which will not be repeated here. Attached Figure Description

[0084] Figure 1 is a schematic diagram of the 5G system architecture provided in an embodiment of this application;

[0085] Figure 2 is a schematic diagram of the NF service registration process provided in an embodiment of this application;

[0086] Figure 3 is a schematic diagram of a VFL client discovery process provided in an embodiment of this application;

[0087] Figure 4 is a schematic diagram of the first communication method in vertical federated learning provided in an embodiment of this application.

[0088] Figure 5 is a schematic diagram of a second type of communication method in vertical federated learning provided in an embodiment of this application;

[0089] Figure 6 is a schematic diagram of the third communication method in vertical federated learning provided in the embodiments of this application;

[0090] Figure 7 is a schematic diagram of the fourth communication method in vertical federated learning provided in the embodiments of this application;

[0091] Figure 8 is a schematic diagram of the fifth communication method in vertical federated learning provided in the embodiments of this application;

[0092] Figure 9 is a structural schematic diagram of a communication device provided in an embodiment of this application;

[0093] Figure 10 is another structural schematic diagram of the communication device provided in an embodiment of this application;

[0094] Figure 11 is another structural schematic diagram of the communication device provided in the embodiment of this application. Detailed Implementation

[0095] In the description of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. "At least one of the following" or "one or more of the following" or similar expressions refer to any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can represent: a, b, c; a and b; a and c; b and c; or a and b and c. Where a, b, and c can be single or multiple.

[0096] In the description of this application, the terms "first," "second," etc., are used only to distinguish different objects and do not limit the quantity or order of execution, nor do they imply that they are necessarily different. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0097] In this application, the words "exemplary" or "for example" are used to indicate that something is an example, illustration, or illustration. Any embodiment or design described as "exemplary," "for example," or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of the words "exemplary," "for example," or "for example" is intended to present the relevant concepts in a specific manner.

[0098] It should be understood that in this application, "when," "if," and "if" all refer to the device making a corresponding action under certain objective circumstances, and are not time-limited, nor do they require the device to make a judgment when it is implemented, nor do they imply any other limitations.

[0099] In this application, the use of singular designations for elements is intended to represent "one or more" rather than "one and only one," unless otherwise specified.

[0100] In this application, the indication includes direct indication (also known as explicit indication) and implicit indication. Direct indication information A refers to information A that is included; implicit indication information A refers to information A that is indicated through the correspondence between information A and information B, and the direct indication information B. The correspondence between information A and information B can be predefined, pre-stored, pre-burned, or pre-configured.

[0101] In this application, expressions such as "C corresponds to D" and "C corresponds / is associated with D" all indicate that there is a correspondence / mapping relationship between C and D, and that D can be determined based on C. Determining D based on C includes determining D solely based on C, as well as determining D based on C and other information. Furthermore, the use of information C to determine information D can also include indirect determination, such as when information D is determined based on information E, and information E is determined based on information C.

[0102] In this application, "send" and "receive" can indicate the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which can include direct transmission via the air interface or indirect transmission via the air interface from other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which can include direct reception from YY via the air interface or indirect reception from YY via the air interface from other units or modules. Information may undergo necessary processing between the source and destination, such as format changes, but the destination can understand the valid information from the source. Similar expressions in this application can be interpreted similarly and will not be elaborated further here.

[0103] In this application, "transmitting" can also be understood as "output" of the chip interface, and "receiving" can also be understood as "input" of the chip interface. In other words, transmitting and receiving can occur between devices, such as between network devices and terminal devices, or within a device, such as between components, modules, chips, software modules, or hardware modules within the device via a bus, wiring, or interface.

[0104] The following is a brief description of the network elements and system architecture involved in this application.

[0105] The technical solutions provided in this application can be applied to various communication systems that deploy virtual network functions. Examples include: 5th generation (5G) communication systems or new radio (NR), LTE networks equipped with network functions virtualization infrastructure (NFVI) or other Long Term Evolution (LTE) networks with virtual network functions, MulteFire networks (creating new wireless networks by independently operating LTE technology on unlicensed spectrum (e.g., global 5GHz unlicensed spectrum), or home base station networks, mobile networks with wireless fidelity (Wi-Fi) access, wideband code division multiple access (WCDMA) networks, fixed-mobile converged networks (fixed access networks accessing mobile networks), and other future communication systems.

[0106] For example, the technical solution provided in this application can be applied to the 5G system architecture defined in 3GPP TS 23.288.

[0107] Referring to Figure 1, Figure 1 is a schematic diagram of the 5G system architecture provided in an embodiment of this application. As shown in Figure 1, the 5G system architecture 100 includes, but is not limited to: user equipment (UE) 110, access network, core network (CN), data network (DN) 140, and application function (AF) 141. The access network can be used to implement functions related to radio access, and it may include a radio access network (RAN) 120. The core network may include various network functions (NFs) or network elements, such as including all or some of the following logical functions: user plane function (UPF)130, network exposure function (NEF)131, network repository function (NRF)132, policy control function (PCF)133, unified data management (UDM) function134, unified data repository (UDR) function135, network data analytics function (NWDAF)136, authentication server function (AUSF)137, access and mobility management function (AMF)138, or session management function (SMF)139, etc.

[0108] In one possible implementation, the UE can access the data network by establishing a session between the UE and the RAN, then to the UPF, and finally to the data network (DN), namely a protocol data unit (PDU) session.

[0109] In this context, UE can be a terminal device, such as a mobile phone, IoT terminal device, smart terminal, in-vehicle terminal, in-vehicle equipment, wearable device, multimedia device, streaming media device, etc. Terminal devices can be widely used in various scenarios, such as device-to-device (D2D), vehicle-to-everything (V2X) communication, machine-type communication (MTC), Internet of Things (IoT), virtual reality, augmented reality, industrial control, autonomous driving, telemedicine, smart grids, smart furniture, smart offices, smart wearables, smart transportation, smart cities, etc. Terminal devices can include mobile phones, tablets, computers with wireless transceiver capabilities, wearable devices, vehicles, urban air mobility vehicles (such as drones, helicopters, etc.), ships, robots, robotic arms, smart home devices, etc.

[0110] The RAN (Radio Access Network) provides wireless access for terminal devices, including but not limited to: 5G base stations (Next-Generation node B, gNB), base stations in future mobile communication systems, radio base stations (evolved Node B, eNodeB or eNB) in LTE networks, wireless fidelity access points (Wi-Fi APs), worldwide interoperability for microwave access base stations (WiMAX BS), relay stations, etc.; or modules or units that perform some of the functions of a base station. In the 5G RAN architecture, a gNB can include a centralized unit (CU) and a distributed unit (DU). A gNB can also include a radio unit (RU). CUs and DUs can be understood as a logical functional division of the base station. CUs and DUs can be physically separated or deployed together. For example, multiple DUs can share a single CU, or a single DU can connect to multiple CUs. CUs and DUs can be connected via an F1 interface.

[0111] The AMF (Advanced Mobility Management) is primarily responsible for mobility management in mobile networks, such as user location updates, network registration, and handover. Additionally, the AMF is responsible for transmitting user policies between terminal devices and the PCF (Programming Provider Function). The SMF (Supply Provider Function) is primarily responsible for session management in mobile networks, such as session establishment, modification, and release. Specific functions include assigning Internet Protocol (IP) addresses to users and selecting the UPF (User Packet Forwarder) to provide packet forwarding. The PCF provides policies to the AMF and SMF, such as Quality of Service (QoS) policies and slice selection policies. The UPF (User Packet Processor) is primarily responsible for processing user packets, such as forwarding and accounting.

[0112] NWDAF has functions such as data collection, model training, data analysis, and model inference. NWDAF supports collecting data from other network functions and AFs, collecting data from operation administration and maintenance (OAM) network elements, and providing analytical information to other network functions and AFs.

[0113] NWDAF can be used to collect relevant data from network functions, third-party service servers, terminal devices, or network management systems. Based on this data, it performs data analysis to obtain results, which are then provided to these network functions, third-party service servers, terminal devices, or network management systems. These results can assist the network in selecting service quality parameters, performing traffic routing, or selecting background data transmission strategies. Furthermore, NWDAF can also collect relevant data from network functions, third-party service servers, terminal devices, or network management systems, and use this data to train models to obtain artificial intelligence (AI) or machine learning (ML) models. These AI / ML models can then be provided to other NWDAFs. These AI / ML models can assist NWDAFs in generating data analysis results based on relevant data.

[0114] 3GPP separates the training and inference functions of an NWDAF (Non-Dual-Way AF). An NWDAF can support only model training, only data inference, or both. An NWDAF supporting model training can be called a training NWDAF, or an NWDAF containing a model training logical function (MTLF) (abbreviated as MTLF). A training NWDAF can train a model based on acquired data to obtain a trained model. An NWDAF supporting data inference can be called an inference NWDAF, or an NWDAF containing an analytics logical function (AnLF) (abbreviated as AnLF). An inference NWDAF can input input data into the trained model to obtain analysis results or inference data. In this embodiment, a training NWDAF refers to an NWDAF that at least supports model training. As one possible implementation, a training NWDAF can also support data inference. An inference NWDAF refers to an NWDAF that at least supports data inference. As one possible implementation, an inference NWDAF can also support model training. If an NWDAF simultaneously supports model training and data inference functions, then the NWDAF can be called a training NWDAF, an inference NWDAF, a training-inference NWDAF, or simply an NWDAF. In various embodiments of this application, an NWDAF can be a separate network element or can be co-located with other network elements; this application does not impose any restrictions.

[0115] The NWDAF in this application can simultaneously support model training and / or model (data) inference functions.

[0116] UDM network elements can be responsible for the storage, management, and access control of user data, such as user identification, access authorization, registration, mobility management, subscription management, and SMS management. UDR network elements support the storage and retrieval of multiple network functions (NFs).

[0117] NRF provides registration and discovery functions, enabling network functions (NFs) to discover each other and communicate through an application programming interface (API). NRF can also provide other network element management services, such as network element updates, deregistration, and network element status subscription and push.

[0118] NEF can be used to support the opening of capabilities and events. For example, it can open the capabilities of various network functions (NFs) and transform internal and external information. NEF can provide security guarantees to ensure the security of external applications to the 3GPP network, and provide functions such as opening up the ability to customize the quality of service (QoS) of external applications, subscription to mobility state events, and distribution of AF requests.

[0119] Application Providers (AFs) can be used to convey application-side requests to the network side, such as Quality of Service (QoS) requirements or user state event subscriptions. An AF can be a third-party functional entity or an application server deployed by an operator; this application's embodiments are not limited to these.

[0120] It is understood that the above network elements are examples of one implementation method. This application does not exclude the possibility that network elements or devices with the above network element functions may have other names or other forms in 6G or future wireless communication systems.

[0121] It is understood that the various network functions shown in Figure 1 can be referenced in relevant protocols or standards, and will not be elaborated upon in this application. It is also understood that "Nnef", "Nnrf", "Npcf", "Nudm", "Nudr", "Nnwdaf", "Naf", "Nausf", "Namf", and "Nsmf" in Figure 1 represent the names of service interfaces, which can be used to call corresponding service operations. For details, please refer to the relevant descriptions in the 3GPP standard protocols; they will not be elaborated upon here.

[0122] It should also be understood that N1, N2, N3, N4, N6, etc., shown in Figure 1 are all interface sequence numbers. The meanings of these interface sequence numbers are as follows: N1 can represent the interface between the AMF and the terminal equipment, and can be used to transmit non-access stratum (NAS) signaling (such as QoS rules from the AMF) to the terminal equipment. N2 can represent the interface between the AMF and the RAN, and can be used to transmit radio bearer control information from the core network side to the RAN. N3 can represent the interface between the RAN and the UPF, mainly used to transmit uplink and downlink user plane data between the RAN and the UPF. N4 can represent the interface between the SMF and the UPF, and can be used to transmit information between the control plane and the user plane, including the distribution of forwarding rules, QoS rules, traffic statistics rules, etc., from the control plane to the user plane, as well as the reporting of user plane information. N6 can represent the interface between the UPF and the DN, and can be used to transmit uplink and downlink user data streams between the UPF and the DN. For example, the meanings of the above interface sequence numbers can also be found in the definitions in the 3GPP standard protocols; this application does not limit the meanings of the above interface sequence numbers.

[0123] In one possible implementation, the network element or function in this application embodiment can be a network component in a hardware device, a software function running on dedicated hardware, or a virtualization function instantiated on a platform (e.g., a cloud platform). As one possible implementation, the network element or function in this application embodiment can be implemented by one device, multiple devices working together, or a functional module within a single device; this application embodiment does not specifically limit this.

[0124] To better understand the technical solutions of the embodiments of this application, some related contents involved in this application will be briefly described below.

[0125] I. Network Functions (NF) Service Registration

[0126] Referring to Figure 2, which is a schematic diagram of the NF service registration process provided in an embodiment of this application, the NF service registration process includes, but is not limited to, the following steps:

[0127] Step 1: The NF service consumer (NF instance) sends an NF registration request (e.g., Nnrf_NF Management_NF Register_request) to the NRF to notify the NRF of its NF profile when the NF service consumer becomes operational for the first time. The NF profile of the NF service consumer can be configured by the OAM system.

[0128] Step 2: The NRF stores the NF profile of the NF service consumer and marks the NF service consumer as available. Whether the NF profile sent by the NF service consumer to the NRF needs to be protected for integrity by the NF service consumer and verified by the NRF can be determined by the Standalone (SA) network (SA3).

[0129] Step 3: NRF confirms that the NF registration has been accepted via an NF registration response (e.g., Nnrf_NF Management_NF Register_response).

[0130] It is understood that "NF service consumer" in Figure 2 above can refer to the consumer of NRF services and should not be confused with the role of NF (consumer or producer).

[0131] II. VFL Client Discovery

[0132] In one possible implementation, the AF acts as the VFL server, and the AF can be trusted or untrusted. The NWDAF acts as the VFL client, and there can be one or more of them. Refer to Figure 3, which is a schematic diagram of a VFL client discovery process provided in an embodiment of this application. As shown in Figure 3, when the VFL server is an untrusted AF, the VFL client discovery process includes steps 1 to 7 as described below.

[0133] Step 1: If an untrusted AF acts as the VFL server, the AF determines that VFL operation is required and that an NWDAF is needed as the VFL client. The AF sends a VFL client discovery request to the NEF, providing the selection criteria for the VFL client in the request, such as the analysis identifier (ID), the required NF type (i.e., NWDAF type), and the VFL capability type (i.e., VFL client).

[0134] Step 2: NEF checks whether AF has the right to request the analysis ID from the VFL client according to the configured policy.

[0135] In steps 3-4, the NEF, representing the AF, discovers VFL clients (i.e., NWDAFs) from the NRF. For example, the NEF sends a network element discovery request (e.g., Nnrf_NF Discovery_Request) to the NRF, and the NRF returns a network element discovery response (e.g., Nnrf_NF Discovery_Response) to the NEF. This network element discovery response includes one or more candidate NWDAFs.

[0136] Step 5: NEF selects NWDAFs that can act as VFL clients and meet the received selection criteria. NEF anonymizes the NWDAF's instance identifier (NF instance ID) and assigns a temporary VFL client ID to each selected NWDAF. NEF stores the temporary VFL client ID and NF instance ID of the NWDAF.

[0137] Step 6: The NEF sends a VFL client discovery response to the AF. This VFL client discovery response includes the temporary VFL client ID of the NWDAF.

[0138] Step 7: The AF stores the received temporary VFL client ID. This temporary VFL client ID is used by the AF to indicate the target VFL client (i.e., the NWDAF) during subsequent interactions with the NEF.

[0139] As shown in Figure 3, when the VFL server is a trusted AF, the discovery process of the VFL client may include steps 8 to 9.

[0140] Step 8: If the AF acting as the VFL server is trusted, the AF can invoke the Network Element Discovery Request (e.g., Nnrf_NF Discovery_Request) service to discover the NWDAF as a VFL Client from the NRF. The selection criteria for the VFL Client may include: analysis ID, required NF type (i.e., NWDAF type), VFL capability type (i.e., VFL client), etc.

[0141] Step 9: The NRF returns one or more NWDAFs as VFL Clients to the AF via a network element discovery response (e.g., Nnrf_NF Discovery_Response). This response includes the NF instance ID of the NWDAFs. The NF instance ID of the NWDAF can be used to identify the NWDAF during subsequent interactions between the AF and the NEF.

[0142] It is understandable that before the VFL client discovery process shown in Figure 3 above, the NWDAF registers with the NRF using its NF Profile, i.e., performs the NF service registration process. The NWDAF can register the analytics IDs it supports. Similarly, the AF can register with the NRF using its NF Profile, i.e., perform the NF service registration process.

[0143] In this application, the NF instance ID can represent an identifier provided by an NF service consumer (such as a consumer of an NRF service), which is globally unique within the public land mobile network (PLMN) to which the NRF to which the NF is registered belongs.

[0144] In this application, "Trusted" can be understood as being trusted by the PLMN or being within the PLMN, while "Untrusted" can be understood as not being trusted by the PLMN or being outside the PLMN. For example, an Untrusted AF (Trusted 3rd party Application Functions) is an AF that does not belong to the 3GPP operator domain, while a Trusted AF is an AF that belongs to the 3GPP operator domain.

[0145] The technical solution provided in this application will be described in detail below with reference to more accompanying drawings.

[0146] The technical solutions provided in this application are illustrated through multiple embodiments, with specific reference to the descriptions of the various embodiments below. The same or similar parts between the various embodiments or implementations can be referred to mutually. In the various embodiments of this application, and in the various implementation methods / methods / implementations within each embodiment, unless otherwise specified or logically conflicting, the terminology and / or descriptions between different embodiments and between the various implementation methods / methods / implementations within each embodiment are consistent and can be mutually referenced. The technical features in different embodiments and the various implementation methods / methods / implementations within each embodiment can be combined according to their inherent logical relationships to form new embodiments, implementation methods, methods, or implementation approaches. The embodiments described below do not constitute a limitation on the scope of protection of this application.

[0147] In one possible implementation, the "data processing network element" of this application can be a network element with data storage and / or data management functions. For example, the "data processing network element" of this application can be a UDR and / or a UDM.

[0148] In the following embodiments of this application, the first identifier can be an identifier used outside the PLMN to identify the NWDAF, or the first identifier can be an identifier opened to the outside of the PLMN (such as an AF) to identify the NWDAF, or the first identifier can be used to identify the NWDAF for an AF outside the PLMN, or the first identifier can be used to identify the NWDAF for an untrusted AF. It can be understood that the NWDAF is a network element of the core network, a network entity within a 3GPP management domain. For example, the first identifier can be a temporary NF ID. Exemplarily, the same NWDAF can have multiple first identifiers, which can be used by different devices outside the PLMN, such as multiple different AFs identifying the NWDAF. These multiple first identifiers can also be used to identify the NWDAF for the same AF, with the AF using different first identifiers in different services or tasks (e.g., VFL tasks).

[0149] Correspondingly, in the following embodiments of this application, the second identifier can be an identifier used within the PLMN to identify the NWDAF, or the second identifier can be used to identify the NWDAF within the PLMN (such as the CN, data processing network element UDM, and / or UDR, etc.), or the second identifier can be used to identify the NWDAF for trusted network elements. For example, the second identifier can be an NF instance ID. For instance, the second identifier identifies the NWDAF within the 3GPP operator domain.

[0150] For clarity, the following explanation will use the example of the first identifier being the temporary NF ID and the second identifier being the NF instance ID. It should be understood that the temporary NF ID can be replaced by the first identifier, and the NF instance ID can be replaced by the second identifier.

[0151] Referring to Figure 4, which is a schematic flowchart of a first communication method in vertical federated learning provided in an embodiment of this application, in this method, the NWDAF acts as a VFL client and the AF acts as a VFL server. For example, this AF is untrusted.

[0152] As shown in Figure 4, the communication method in this vertical federated learning includes, but is not limited to, the following steps:

[0153] S101, AF sends the first message to NEF1. This first message includes the first identifier of K NWDAFs and the identifier of the VFL task. The VFL task includes one or more of the following: training task, prediction task, training and prediction task, etc. K is a positive integer.

[0154] In one possible implementation, before step S101, the NWDAF (which can be one or more) acting as a VFL client can register its NF Profile with the NRF. This NF Profile includes VFL information supported by the NWDAF, such as the analytics ID. The specific registration process can be found in existing technologies, as shown in Figure 2 above, and will not be elaborated here.

[0155] When the AF (Active Element) acting as the VFL server determines that VFL operation is required and that an NWDAF (Non-WDAF) is needed as a VFL client, the AF can send a Network Element Discovery Request 1 to NEF1. This AF discovery request 1 can be used to discover candidate NWDAFs. It can include VFL client selection information, such as one or more of the following: analysis ID, required NF type (i.e., NWDAF type), VFL capability type (i.e., VFL client), etc. NEF1 sends a Network Element Discovery Request 2 (e.g., Nnrf_NF Discovery_Request) to the NRF to discover VFL clients (i.e., NWDAFs) from the NRF. The NRF returns a Network Element Discovery Response 2 (e.g., Nnrf_NF Discovery_Response) to NEF1. This response includes the profiles of the candidate NWDAFs. A candidate NWDAF's profile includes a second identifier (e.g., NF instance ID). After receiving the Network Element Discovery Response 2, NEF1 can assign a first identifier (e.g., temporary NWDAF ID) to one or more candidate NWDAFs. The first identifier of the candidate NWDAF can be an anonymized version of the second identifier of the candidate NWDAF (such as the NF instance ID), for example, some fields of the NF instance ID are hidden. For instance, after receiving the network element discovery response 2, NEF1 can also store the first identifier (such as a temporary NWDAF ID) and the second identifier (NF instance ID) of the candidate NWDAF. NEF1 can then send a network element discovery response 1 to the AF, which may include the first identifier (such as a temporary NWDAF ID) of each of the one or more candidate NWDAFs. The AF stores the received first identifier (such as the temporary NWDAF ID). At this point, the VFL client's discovery process ends. For instance, the AF can also store the producer information of the received first identifier (such as the NEF1 ID). It is understood that the above VFL client discovery process is an illustration of the AF discovering a candidate NWDAF through one NEF. In practical applications, the AF can discover candidate NWDAFs through multiple NEFs. This application embodiment does not limit this.

[0156] In one possible implementation, the aforementioned network element discovery request 2 may further include AF information. The NRF can determine candidate NWDAFs based on the AF information in the network element discovery request 2. For example, if the AF information is the AF's identifier, the NRF can verify whether the allowed NFc list in the locally stored NWDAF profile includes the AF's identifier; if the allowed NFc list in the profile of an NWDAF includes the AF's identifier, the NRF determines that the NWDAF is a candidate NWDAF. As another example, if the AF information is the AF's type, the NRF can verify whether the allowed NFc types in the locally stored NWDAF list include the AF's type; if the allowed NFc types in the allowed NFc list of an NWDAF include the AF's type, the NRF determines that the NWDAF is a candidate NWDAF. For another example, if the information of an AF is its domain name, the NRF can verify whether the allowed NFc Domains of the locally stored NWDAF include the domain name of that AF. If the allowed NFc Domains of an NWDAF include the domain name of that AF, the NRF determines that the NWDAF is a candidate NWDAF. Then, the NRF can send a network element discovery response 2 to NEF1.

[0157] In some scenarios, after NEF1 assigns a first identifier to one or more candidate NWDAFs, it can send a storage request 2 to UDR / UDM. This storage request 2 can include the first identifier (e.g., a temporary NWDAF ID) and the second identifier (e.g., an NF instance ID) of each candidate NWDAF. Upon receiving the storage request 2, UDR / UDM can store the first and second identifiers of all candidate NWDAFs in the storage request 2, thereby establishing a mapping relationship between the first and second identifiers of the candidate NWDAFs. For example, in this embodiment of the application, each network element can use associated storage when storing the first and second identifiers of the same NWDAF.

[0158] In one possible implementation, after the VFL client's discovery process is completed (e.g., after the AF receives the first identifier of a candidate NWDAF), the AF can determine VFL members for a VFL task, i.e., determine K NWDAFs that participate in the VFL task. The K NWDAFs participating in the VFL task are all or some of the aforementioned candidate NWDAFs. Then, the AF can send a first message to NEF1. This VFL task can be used to train the model corresponding to the aforementioned analysis ID. The first message may include the first identifier (e.g., a temporary NWDAF ID) of each of the K NWDAFs participating in the VFL task, and information about the VFL task. For example, the information about the VFL task includes the identifier of the VFL task, and optionally, the analysis ID. For ease of description, the identifier of the VFL task can be represented by a VFL correlation ID; however, other identifiers can also be used, as long as they can identify the VFL task once. This application is not limited to these.

[0159] S102, NEF1 sends storage request 1 to UDR / UDM. Storage request 1 includes the first identifier of the aforementioned K NWDAFs, the identifier of the VFL task, and the second identifier of the K NWDAFs.

[0160] S103, UDR / UDM stores the mapping relationship between the first and second identifiers of K NWDAFs and the identifier of the VFL task.

[0161] In one possible implementation, after receiving the first message, NEF1 can send a storage request 1 to UDR / UDM. This storage request 1 may include the first identifier (e.g., a temporary NWDAF ID) of each of the K NWDAFs and the identifier of the VFL task (e.g., a VFL correlation ID). It is understood that in the VFL client discovery process, if NEF1 has stored the first and second identifiers of candidate NWDAFs with UDR / UDM, then storage request 1 may not carry the second identifiers of the K NWDAFs; however, storage request 1 may also carry the second identifiers of the K NWDAFs, and this embodiment is not limited. It is understood that if K is greater than 1, then storage request 1 may include the first identifier of each of the K NWDAFs and the identifier of the VFL task.

[0162] In another possible implementation, after receiving the first message, NEF1 can use the first identifiers of the K NWDAFs to index the second identifiers (such as NF instance IDs) corresponding to the first identifiers of the K NWDAFs in local storage. Then, NEF1 can send storage request 1 to UDR / UDM. Storage request 1 includes the first identifiers of the K NWDAFs, the identifier of the VFL task, and the second identifiers of the K NWDAFs. It can be understood that if K is greater than 1, storage request 1 can include the first identifier of each of the K NWDAFs, the second identifier of each NWDAF, and the identifier of the VFL task.

[0163] In one possible implementation, after receiving the storage request 1, the UDR / UDM can store the first and second identifiers of the K NWDAFs, as well as the identifier of the VFL task, thereby establishing a mapping relationship between the first and second identifiers of each of the K NWDAFs and the identifier of the VFL task. In some scenarios, after storage is complete, the UDR / UDM can return a data storage success response message to NEF1. For example, the content stored in the UDR / UDM can be as shown in Table 1 below.

[0164] Table 1

[0165] The index (key) in Table 1 above can be an identifier for the VFL task. For example, the temporary NF ID and NF instance ID can be obtained through the VFL correlation ID, and the NF instance ID can also be obtained through the temporary NF ID.

[0166] S104, the AF sends a first request message to NEF1. This first request message includes the first identifier of the first NWDAF and the identifier of the VFL task. This VFL task is used to train the model. The first request message requests the services provided by the model corresponding to the identifier of the VFL task.

[0167] In one possible implementation, the first request message may include a first identifier of the first NWDAF (such as a temporary NWDAF ID) and an identifier of the VFL task (such as a VFL correlation ID). The first NWDAF may be one of the NWDAFs participating in the VFL task. The VFL task may be used to train a model. The first request message may be used to request services provided by the model. For example, the first request message may be a VFL prediction / inference request.

[0168] S105, NEF1 sends a second request message to UDR / UDM. This second request message includes the identifier of the VFL task. This second request message can be used to obtain the first identifier and second identifier corresponding to the identifier of the VFL task.

[0169] S106, UDR / UDM sends a second response message to NEF1. This second response message includes the mapping relationship between the first identifier and the second identifier of each of the aforementioned K NWDAFs. K is a positive integer. These K NWDAFs participate in this VFL task. These K NWDAFs include the first NWDAF.

[0170] S107, NEF1 sends a third request message to the first NWDAF. This third request message is used to request services provided by the model. For example, the third request message includes a second identifier of the first NWDAF.

[0171] In one possible implementation, after receiving the first request message, NEF1 can send a second request message to UDR / UDM. For example, before sending the second request message to UDR / UDM, NEF1 can also check in its local storage whether there is a second identifier (such as an NF instance ID) corresponding to the first identifier (such as a temporary NWDAF ID) of the first NWDAF. Alternatively, before sending the second request message to UDR / UDM, NEF1 can also check in its local storage whether the first and second identifiers corresponding to the identifier of the VFL task in the first request message exist. If there is no second identifier corresponding to the first identifier of the first NWDAF locally, or if there is no first and second identifier corresponding to the identifier of the VFL task locally, NEF1 can send a second request message to UDR / UDM. This second request message may include the identifier of the aforementioned VFL task. After receiving the second request message, UDR / UDM can search locally for the first and second identifiers corresponding to the identifier of the VFL task, thereby finding the first and second identifiers of K NWDAFs. UDR / UDM can then return a second response message to NEF1. The second response message may include the first and second identifiers of the K NWDAFs and their mapping relationship. These K NWDAFs participate in the VFL task. These K NWDAFs include the aforementioned first NWDAF. These K NWDAFs are also the NWDAFs that the AF determined in step S101 above to participate in the VFL task. After receiving the second response message, NEF1 can store the first identifier and the second identifier of each of the K NWDAFs and their mapping relationship, and can send a third request message to the first NWDAF based on the second identifier of the first NWDAF (such as the NF instance ID). This third request message can be used to request services provided by the model trained for the VFL task. This third request message may include the second identifier of the first NWDAF.

[0172] In one possible implementation, if a second identifier corresponding to the first identifier of the first NWDAF exists locally, NEF1 can send a third request message to the first NWDAF. This third request message is used to request services provided by the model trained on the aforementioned VFL task. This third request message may include the second identifier of the first NWDAF (such as an NF instance ID).

[0173] In this embodiment of the application, after receiving a service request (e.g., a VFL prediction / inference request) from an AF, the NEF checks whether the local storage contains the NF instance ID (i.e., the second identifier) ​​corresponding to the temporary NWDAF ID (i.e., the first identifier) ​​of the first NWDAF in the service request. If not, the NEF uses the identifier of the VFL task carried in the service request to query the UDR / UDM for the temporary NWDAF ID and NF instance ID corresponding to the identifier of the VFL task. The NF instance ID corresponding to the identifier of the VFL task includes the NF instance ID of the first NWDAF. Then, the NEF can send a service request (e.g., a VFL prediction / inference request) to the first NWDAF based on the queried NF instance ID of the first NWDAF. In this way, the AF can both obtain the services provided by the VFL client (e.g., the NWDAF) and improve the security of the VFL client.

[0174] In one possible implementation, the actions / behaviors performed by NEF1 in steps S104 to S107 can also be performed by other NEFs, such as NEF2. For example: AF sends a request message to NEF2, which includes the first identifier of the third NWDAF and the identifier of the VFL task. The third NWDAF participates in the VFL task. NEF2 sends a fifth request message to UDR / UDM, which includes the identifier of the VFL task. UDR / UDM sends a fifth response message to NEF2, which includes the mapping relationship between the first identifier and the second identifier of each of the K NWDAFs, which participate in the VFL task, and these K NWDAFs include the third NWDAF. NEF2 sends a request message to the third NWDAF, which includes the second identifier of the third NWDAF.

[0175] Furthermore, in some scenarios, NEF services may be unavailable, such as due to short-term high NEF load or NEF maintenance. This can cause the network element service corresponding to the temporary NWDAF ID allocated by the NEF to become unavailable, resulting in network failure. Therefore, in this embodiment, after the AF determines the VFL member for a VFL task, it stores the VFL member's information (temporary NWDAF ID, NWDAF instance ID, and the identifier of the VFL task) in the UDR / UDM through the NEF. When some NEFs are unavailable, the AF can send a first request message to other NEFs, including the temporary NWDAF ID and the identifier of the VFL task. Other NEFs can use the identifier of the VFL task to query the NF instance ID corresponding to the temporary NWDAF ID from the UDR / UDM, and then send a third request message to the NWDAF identified by the NF instance ID, requesting the NWDAF to provide services. This solves the problem of network element services corresponding to temporary NWDAF IDs being unavailable when the service of a certain NEF is unavailable, reduces network failures, and improves the continuity of service requests based on anonymized network element identifiers.

[0176] In one possible implementation, after step S107, the method shown in Figure 4 above further includes: the AF sending a fourth request message (e.g., a VFL prediction / inference request) to NEF1. The AF sending the fourth request message can be the same as or different from the AF sending the first request message; this embodiment is not limited. The fourth request message may include the first identifier of the second NWDAF (e.g., a temporary NWDAF ID) and the identifier of the VFL task. NEF1 can search in local storage for a second identifier (e.g., an NF instance ID) corresponding to the first identifier of the second NWDAF. If the aforementioned K NWDAFs include the second NWDAF, NEF1 can find the second identifier corresponding to the first identifier of the second NWDAF locally, and NEF1 does not need to query the UDR / UDM again. NEF1 can send a fifth request message to the second NWDAF. This fifth request message can be used to request services provided by the model trained for the VFL task. This fifth request message may include the second identifier corresponding to the first identifier of the second NWDAF stored locally.

[0177] In this embodiment, the first and second identifiers of all NWDAFs participating in the VFL task are indexed by the identifier of the VFL task. For the same VFL task, NEF only needs to query UDR / UDM once to obtain the first and second identifiers of all NWDAFs participating in the VFL task, which can improve the efficiency of NEF in processing service requests.

[0178] In one possible implementation, after step S107, NEF1 can send a second message to the UDR / UDM based on information from the AF. The second message includes the identifier of the aforementioned VFL task. The second message can be used to update / delete / unregister the information associated with the identifier of the VFL task in the UDR / UDM. If the second message is used to update the information associated with the identifier of the VFL task in the UDR / UDM, the second message may also include the updated information.

[0179] In this embodiment of the application, the NEF determines the usage status of the VFL task identifier based on information from the AF and maintains the VFL task identifier-related information in the UDR / UDM. This can prevent the NEF from maintaining invalid VFL task identifiers and reduce the waste of system resources.

[0180] Referring to Figure 5, which is a second flowchart illustrating the communication method in vertical federated learning provided in this embodiment of the application, in this method, the NWDAF acts as a VFL client, and the AF acts as a VFL server. For example, this AF is untrusted.

[0181] As shown in Figure 5, the communication method in this vertical federated learning includes, but is not limited to, the following steps:

[0182] S201, AF sends a network element discovery request to NEF1. This network element discovery request is used to discover candidate NWDAFs.

[0183] S202, NEF1 obtains information about candidate NWDAFs from NRF.

[0184] S203, NEF1 assigns the first identifier to the candidate NWDAF.

[0185] S204, NEF1 sends a network element discovery response to the AF. This network element discovery response includes the first identifier of the candidate NWDAF.

[0186] In one possible implementation, one or more NWDAFs acting as VFL clients can register their NF Profile with the NRF. This NF Profile includes VFL information supported by the NWDAF, such as the analytics ID. The specific registration process can be found in existing technologies, as shown in Figure 2 above, and will not be elaborated here.

[0187] When an AF (Active Element) acting as a VFL server determines that VFL operation is required and that an NWDAF (Network Element AF) is needed as a VFL client, the AF can send a network element discovery request to NEF1. This request may include VFL client selection information, such as one or more of the following: analysis ID, required NF type (i.e., NWDAF type), VFL capability type (i.e., VFL client), etc. NEF1 sends a network element discovery request A (e.g., Nnrf_NF Discovery_Request) to the NRF to discover VFL clients (i.e., NWDAFs) from the NRF. The NRF returns a network element discovery response B (e.g., Nnrf_NF Discovery_Response) to NEF1. This response B includes the profiles of candidate NWDAFs. A candidate NWDAF's profile includes a second identifier (e.g., NF instance ID) for that NWDAF. NEF1 assigns a first identifier (e.g., temporary NWDAF ID) to one or more candidate NWDAFs. The first identifier of the candidate NWDAF can be an anonymized version of the second identifier of the candidate NWDAF (such as the NF instance ID), for example, some fields of the NF instance ID are hidden. For example, NEF1 can also store the first and second identifiers of the candidate NWDAF. NEF1 can then send a network element discovery response to the AF, which may include the first identifiers of one or more candidate NWDAFs (such as temporary NWDAF IDs). The AF stores the received first identifiers (such as temporary NWDAF IDs). For example, the AF can also store the producer information (e.g., NEF1 ID) of the received first identifiers.

[0188] In one possible implementation, the aforementioned network element discovery request A may also include information about the AF (Active Functional Attachment). The NRF can determine candidate NWDAFs based on the AF information in the network element discovery request A. For example, if the AF information is the AF's identifier, the NRF can verify whether the allowed NFc list in the locally stored NWDAF profile includes the AF's identifier; if the allowed NFc list in the profile of an NWDAF includes the AF's identifier, the NRF determines that the NWDAF is a candidate NWDAF. As another example, if the AF information is the AF's type, the NRF can verify whether the allowed NFc types in the locally stored NWDAF list include the AF's type; if the allowed NFc types in the allowed NFc list of an NWDAF include the AF's type, the NRF determines that the NWDAF is a candidate NWDAF. For another example, if the information of an AF is its domain name, the NRF can verify whether the allowed NFc Domains of the locally stored NWDAF include the domain name of that AF. If the allowed NFc Domains of an NWDAF include the domain name of that AF, the NRF determines that the NWDAF is a candidate NWDAF. Then, the NRF can send a network element discovery response B to NEF1.

[0189] S205, NEF1 sends Storage Request 3 to UDR / UDM. Storage Request 3 includes the first and second identifiers of the candidate NWDAF.

[0190] S206, UDR / UDM stores the first and second identifiers of the candidate NWDAF and their mapping relationship.

[0191] In one possible implementation, after NEF1 assigns a first identifier to one or more candidate NWDAFs, it can send a storage request 3 to UDR / UDM. This storage request 3 can include the first and second identifiers of all candidate NWDAFs. Upon receiving the storage request 3, UDR / UDM can store the first and second identifiers of all candidate NWDAFs in the storage request 3, thereby establishing a mapping relationship between the first and second identifiers of the candidate NWDAFs.

[0192] In one possible implementation, after step S204, the AF can determine the VFL members for a VFL task, that is, determine the N NWDAFs participating in the VFL task. These N NWDAFs are all or some of the candidate NWDAFs mentioned above. The AF can send the first identifiers of the N NWDAFs participating in the VFL task to NEF1. Then, NEF1 sends storage request 3 to UDR / UDM. This storage request 3 includes the first and second identifiers of these N NWDAFs. UDR / UDM can then store the first and second identifiers of these N NWDAFs.

[0193] For example, in this application embodiment, each network element can use associated storage when storing the first and second identifiers of the same NWDAF. In some scenarios, after storage is completed, the UDR / UDM can return a data storage success response message to NEF1.

[0194] In one possible implementation, the storage request 3 may further include one or more of the following: producer information of the first identifier (e.g., NEF1 ID), or user information of the first identifier (e.g., AF ID). Accordingly, the UDR / UDM may also store the NEF1 ID and / or AF ID. For example, the NEF1 ID and / or AF ID stored in the UDR / UDM are associated with the first identifier / second identifier, i.e.: first identifier + second identifier + NEF1 ID and / or AF ID.

[0195] For example, the contents stored in UDR / UDM can be as shown in Table 2 below.

[0196] Table 2

[0197] The index (key) for the content shown in Table 2 above can be a combination of one or more of the following: temporary NF ID (i.e., the first identifier), the producer identifier of temporary NF ID (i.e., the first identifier), or the consumer identifier of temporary NF ID (i.e., the first identifier). For example, the index (key) for the content shown in Table 2 above could be temporary NF ID (i.e., the first identifier), or temporary NF ID + producer identifier of temporary NF ID, or temporary NF ID + producer identifier of temporary NF ID + consumer identifier of temporary NF ID, etc., etc., which are not listed here.

[0198] It is understood that the NEF in this application embodiment stores the producer identifier and consumer identifier of the first identifier (such as a temporary NWDAF ID) in the UDR / UDM, which can improve the management efficiency of the first identifier (such as a temporary NWDAF ID) and reduce the usage errors caused by the duplication of the first identifier (such as a temporary NWDAF ID).

[0199] S207, AF sends a first request message to NEF2. The first request message includes the first identifier of the first NWDAF.

[0200] In one possible implementation, after receiving the first identifier of a candidate NWDAF, the AF can determine the VFL members for a VFL task, that is, determine the N NWDAFs participating in the VFL task. These N NWDAFs are all or some of the aforementioned candidate NWDAFs. The VFL task can be used to train a model. If the NEF1 service is unavailable at this time, such as due to a short-term high load on NEF1 or NEF1 maintenance, the AF can send a first request message to NEF2. The first request message may include the first identifier of the first NWDAF (such as a temporary NWDAF ID). The first NWDAF may be a candidate NWDAF discovered by NEF1. In other words, the first identifier of the first NWDAF may be assigned by NEF1. The first request message can be used to request services provided by the model trained by the VFL task. For example, the first request message may be a VFL prediction / inference request.

[0201] In one possible implementation, the first request message may further include one or more of the following: producer information of the first identifier of the first NWDAF (e.g., NEF1 ID), or user information of the first identifier of the first NWDAF (e.g., AF ID).

[0202] S208, NEF2 sends a second request message to UDR / UDM. This second request message includes the first identifier of the first NWDAF.

[0203] S209, UDR / UDM sends a second response message to NEF2. This second response message includes the second identifier of the first NWDAF.

[0204] S210, NEF2 sends a third request message to the first NWDAF. This second request message includes the second identifier of the first NWDAF.

[0205] In one possible implementation, after receiving the first request message, NEF2 can send a second request message to UDR / UDM. For example, before sending the second request message to UDR / UDM, NEF2 can also search its local storage for a second identifier (e.g., NF instance ID) corresponding to the first identifier of the first NWDAF (e.g., a temporary NWDAF ID). If no second identifier corresponding to the first identifier of the first NWDAF is found locally, NEF2 can send the second request message to UDR / UDM. This second request message may include the first identifier of the first NWDAF. For example, the second request message may also include one or more of the following: producer information of the first identifier of the first NWDAF (e.g., NEF1 ID), or user information of the first identifier of the first NWDAF (e.g., AF ID). After receiving the second request message, UDR / UDM can search its local storage for the second identifier corresponding to the first identifier of the first NWDAF. UDR / UDM can then return a second response message to NEF1. This second response message may include the second identifier of the first NWDAF. After receiving the second response message, NEF2 can store the first identifier and second identifier of the first NWDAF and their mapping relationship, and can send a third request message to the first NWDAF based on the second identifier of the first NWDAF (such as the NF instance ID). This third request message can be used to request services provided by the model trained for the VFL task. The third request message may include the second identifier of the first NWDAF. It is understood that if NEF2 has a second identifier corresponding to the first identifier of the first NWDAF locally, NEF2 can send the third request message to the first NWDAF.

[0206] In this embodiment, after the NEF assigns a first identifier to a candidate NWDAF, it stores the first and second identifiers of the candidate NWDAF in the UDR / UDM. When the NEF that assigned the first identifier (e.g., NEF1) becomes unavailable, the AF can send a first request message to other NEFs (e.g., NEF2), which includes the first identifier of a candidate NWDAF. Other NEFs (e.g., NEF2) can use the first identifier of the candidate NWDAF to query the second identifier corresponding to the first identifier from the UDR / UDM, and then send a third request message to the NWDAF identified by the second identifier, requesting that NWDAF to provide services. This solves the problem that when the service of a certain NEF is unavailable, the network element service corresponding to the first identifier is also unavailable, reducing network failures and improving the continuity of service requests based on anonymized network element identifiers.

[0207] In one possible implementation, after step S209, NEF1 can send a second message to the UDR / UDM based on information from the AF. This second message includes the first identifiers (such as temporary NWDAF IDs) of M NWDAFs. M is a positive integer. This second message can be used to update / delete / unregister information in the UDR / UDM associated with the first identifiers of these M NWDAFs. If the second message is used to update information in the UDR / UDM associated with the first identifiers of these M NWDAFs, it may also include updated information, such as the identifiers of the VFL tasks in which these M NWDAFs participate.

[0208] In this embodiment of the application, the NEF determines the usage of the temporary NWDAF ID based on the information from the AF and maintains the information related to the temporary NWDAF ID in the UDR / UDM. This can prevent the NEF from maintaining invalid temporary NWDAF IDs and reduce the waste of system resources.

[0209] Referring to Figure 6, which is a third flowchart illustrating the communication method in vertical federated learning provided in an embodiment of this application, in this method, the NWDAF acts as a VFL client and the AF acts as a VFL server. For example, this AF is untrusted.

[0210] As shown in Figure 6, the communication method in this vertical federated learning includes, but is not limited to, the following steps:

[0211] S301, NEF1 sends a second network element discovery request to the NRF. This second network element discovery request includes second information and / or AF information. The second information is used to indicate the allocation of a first identifier. The second network element discovery request also includes the VFL capability type. The second network element discovery request is used to discover candidate NWDAFs.

[0212] In one possible implementation, one or more NWDAFs acting as VFL clients can register their NF Profile with the NRF. This NF Profile includes VFL information supported by the NWDAF, such as the analytics ID. The specific registration process can be found in existing technologies, as shown in Figure 2 above, and will not be elaborated here.

[0213] When the AF (Active Element) acting as the VFL server determines that VFL operation is required and that an NWDAF (Near-Wide AF) is needed as a VFL client, the AF can send a Network Element Discovery Request 1 to NEF1. This AF discovery request 1 can be used to discover candidate NWDAFs. It may include VFL client selection information, such as one or more of the following: analysis ID, required NF type (i.e., NWDAF type), VFL capability type (i.e., VFL client), etc. NEF1 can check whether the AF has the right to request an analysis ID from the VFL client according to the configured policy. NEF1 can also send a second network element discovery request to the NRF (Network Element Release Center). This second network element discovery request may include second information and / or AF information, and may also include the VFL capability type. The second information may be used to instruct the allocation of a first identifier (such as a temporary NWDAF ID) (for candidate NWDAFs). This second network element discovery request can be used to discover candidate NWDAFs.

[0214] S302, the NRF determines candidate NWDAFs and their second identifiers based on the VFL capability type. The candidate NWDAFs match the VFL capability type.

[0215] S303, the NRF determines, based on the second information or the information from the AF, to assign a first identifier to the candidate NWDAF.

[0216] In one possible implementation, after receiving the aforementioned second network element discovery request, the NRF can determine candidate NWDAFs (which may be one or more) based on the VFL capability type. The candidate NWDAF matches the VFL capability type. Furthermore, since the NRF stores the NWDAF profile, and the NWDAF profile includes a second identifier (such as an NF instance ID), the NRF can determine the second identifier of the candidate NWDAF. The NRF can then determine, based on the second information or AF information in the second network element discovery request, to assign a first identifier (such as a temporary NWDAF ID) to the candidate NWDAF. For example, the aforementioned second information is used to indicate the assignment of the first identifier. As another example, the AF information includes the AF ID and / or the AF type, etc. The NRF can determine whether the AF is trustworthy based on the AF information. If the NRF confirms that the AF is untrustworthy, then the NRF assigns a first identifier to the candidate NWDAF. Conversely, if the NRF confirms that the AF is trustworthy, then it is not necessary to assign a first identifier to the candidate NWDAF.

[0217] In one possible implementation, the information of the aforementioned AF can also be used to filter candidate NWDAFs. The aforementioned second network element discovery request includes AF information and VFL capability type. The NRF can determine candidate NWDAFs (which may be one or more) based on the VFL capability type and AF information. Specifically, the allowed network function consumers (NF consumers) of the candidate NWDAF include the AF, and the candidate NWDAF matches the VFL capability type. For example, if the AF information is the AF's identifier, then the NRF-determined list of allowed NFcs for candidate NWDAFs includes the AF's identifier. Alternatively, if the AF information is the AF's type, then the NRF-determined list of allowed NFc types for candidate NWDAFs includes the AF's type. Or, if the AF information is the AF's domain name, then the NRF-determined list of allowed NFc Domains for candidate NWDAFs includes the AF's domain name.

[0218] For example, the second network element discovery request also includes an analysis ID. The analysis identifier supported by the candidate NWDAF is the same as the analysis identifier in the second network element discovery request.

[0219] S304, NRF sends a second network element discovery response to NEF1. This second network element discovery response includes the first and second identifiers of the candidate NWDAF.

[0220] In one possible implementation, after the NRF assigns a first identifier to a candidate NWDAF, it can send a second network element discovery response to NEF1. This second network element discovery response may include the first and second identifiers of the candidate NWDAF. For example, the second identifier of the candidate NWDAF may be carried in the candidate NWDAF's profile. In other words, the second network element discovery response may include the candidate NWDAF's first identifier (such as a temporary NWDAF ID) and its profile.

[0221] S305, NEF1 sends the first identifier of the candidate NWDAF to AF.

[0222] In one possible implementation, after receiving the second network element discovery response, NEF1 can send the first identifier of the candidate NWDAF to AF. AF can store the received first identifier of the candidate NWDAF.

[0223] For example, after receiving the second network element discovery response, NEF1 can also store the first identifier and profile of the aforementioned candidate NWDAF. Of course, NEF1 may also choose not to store them, and this embodiment of the application is not limited thereto.

[0224] In one possible implementation, if the aforementioned NRF does not filter candidate NWDAFs based on the AF information, or if the aforementioned second network element discovery request does not include AF information, NEF1, upon receiving the aforementioned second network element discovery response, can further filter candidate NWDAFs based on the AF information and the NWDAF profile in the second network element discovery response. The specific filtering method can be referred to the preceding description and will not be repeated here. Then, NEF1 can send the first identifier of the further filtered NWDAF to the AF. The AF is included in the allowed NF consumers for the further filtered NWDAF.

[0225] S306, AF sends a first request message to NEF1. This first request message includes the first identifier of the NWDAF participating in the VFL task. The VFL task is used to train the model. This first request message is used to request services provided by the model.

[0226] In one possible implementation, after receiving the first identifiers of candidate NWDAFs (which may be one or more), the AF can determine VFL members for a VFL task, that is, determine the NWDAFs (which may be one or more) participating in the VFL task. The VFL task can be used to train the model corresponding to the aforementioned analysis ID. The NWDAFs participating in the VFL task are all or some of the aforementioned candidate NWDAFs. Then, the AF can send a first request message to NEF1. The first request message may include the first identifiers (such as temporary NWDAF IDs) of the NWDAFs participating in the VFL task. The first request message can be used to request services provided by the model. For example, the first request message may be a VFL prediction / inference request.

[0227] In one possible implementation, if the NEF1 service is unavailable at this time, such as due to a short-term high load on NEF1 or maintenance of NEF1, AF can send a first request message to NEF2.

[0228] The following explanation uses the example of AF sending a first request message to NEF1.

[0229] S307, NEF1 sends a first network element discovery request to the NRF. This first network element discovery request includes the first identifier of the NWDAF participating in the VFL task.

[0230] S308, NRF sends a first network element discovery response to NEF1. This first network element discovery response includes the first and second identifiers of the NWDAFs participating in the VFL task.

[0231] In one possible implementation, after receiving the first request message, NEF1 can send a first network element discovery request to NRF. For example, before sending the first network element discovery request to NRF, NEF1 can also check if it has a second identifier (e.g., NF instance ID) corresponding to the first identifier (e.g., temporary NWDAF ID) of the NWDAF participating in the VFL task. If it does not have a second identifier corresponding to the first identifier of the NWDAF participating in the VFL task, NEF1 can send the first network element discovery request to NRF. This first network element discovery request may include the first identifier of the NWDAF participating in the VFL task (which may be one or more). NRF can search locally for the second identifier corresponding to the first identifier of the NWDAF participating in the VFL task. Then, NRF can return a first network element discovery response to NEF1. This first network element discovery response may include the first identifier and the second identifier of the NWDAF participating in the VFL task. If there is only one NWDAF participating in the VFL task, the first network element discovery response may not include the first identifier of the NWDAF participating in the VFL task, but may include the second identifier of the NWDAF participating in the VFL task. If there are multiple NWDAFs participating in the VFL task, the first network element discovery response includes the first identifier and the second identifier of each NWDAF participating in the VFL task. Of course, regardless of the number of NWDAFs participating in the VFL task, the first network element discovery response can also include the first identifier of the NWDAFs participating in the VFL task. For example, after receiving the first network element discovery response, NEF1 can also store the first identifier and the second identifier of the NWDAFs participating in the VFL task, thereby establishing a mapping relationship between the first identifier and the second identifier of the NWDAFs participating in the VFL task. For example, in this embodiment of the application, each network element can use associated storage when storing the first identifier and the second identifier of the same NWDAF.

[0232] For example, the second identifier of the NWDAF participating in the VFL task can be carried in the NWDAF's profile. In other words, the first network element discovery response can include the first identifier (such as a temporary NWDAF ID) and profile of the NWDAF participating in the VFL task.

[0233] In one possible implementation, if a second identifier corresponding to the first identifier of the NWDAF participating in the VFL task exists locally, NEF1 can send a second request message to the NWDAF participating in the VFL task. This second request message can be used to request services provided by the model trained for the VFL task. The second request message may include the second identifier of the NWDAF participating in the VFL task (such as an NF instance ID).

[0234] In one possible implementation, the aforementioned first network element discovery request may further include first information. This first information may be used to indicate the type of the first identifier carried in the first network element discovery request, or to indicate that the first identifier carried in the first network element discovery request is used externally by the PLMN to identify the NWDAF, or to indicate that the first identifier carried in the first network element discovery request is used by an AF external to the PLMN to identify the NWDAF. After receiving the first request message, NEF1 can determine from the first information that the first identifier carried in the first request message is a temporary NWDAF ID, not an NF instance ID. Thus, since the first identifier and the second identifier are constructed in the same way, it is easier for the NRF to distinguish between the first identifier and the second identifier, reducing the possibility of the NRF making an error in finding the second identifier based on the first identifier.

[0235] S309, NEF1 sends a second request message to the NWDAF participating in the VFL task. The second request message is used to request the services provided by the aforementioned model. The second request message includes a second identifier of the NWDAF participating in the VFL task.

[0236] In one possible implementation, after receiving the first network element discovery response, NEF1 can send a second request message to the NWDAFs participating in the VFL task based on their second identifier (such as the NF instance ID). This second request message can be used to request services provided by the model trained for the VFL task. The second request message may include the second identifier of the NWDAF participating in the VFL task. It is understood that if there are multiple NWDAFs participating in the VFL task, NEF1 can send a second request message to each of them separately. Each second request message includes the second identifier of one NWDAF participating in the VFL task.

[0237] In this embodiment, the NRF assigns a first identifier to the NWDAF and stores the first and second identifiers of the NWDAF. The first identifier of the NWDAF is then sent to the AF via the NEF. Even if a certain NEF is unavailable, the AF can still obtain the services provided by the NWDAF corresponding to the first identifier through other NEFs. Furthermore, because the first identifier is used externally by the PLMN to identify the NWDAF, the security of the NWDAF is improved.

[0238] Referring to Figure 7, Figure 7 is a schematic flowchart of the fourth communication method in vertical federated learning provided in this application embodiment. As shown in Figure 7, the communication method in vertical federated learning includes, but is not limited to, the following steps:

[0239] S401, AF sends Network Element Discovery Request 1 to NEF. This Network Element Discovery Request 1 is used to discover candidate NWDAFs.

[0240] S402, NEF sends Element Discovery Request 2 to NRF. Element Discovery Request 2 includes information about the AF. Element Discovery Request 2 is used to discover candidate NWDAFs.

[0241] In one possible implementation, the aforementioned network element discovery request 1 may include network element selection information, such as one or more of the following: analysis ID, required NF type (i.e., NWDAF type), etc. After receiving network element discovery request 1, the NEF can send network element discovery request 2 to the NRF. Network element discovery request 2 may include AF information, such as the AF identifier, AF type, or AF domain name, etc. For example, if the AF wants the NWDAF as a VFL client, network element discovery request 1 may also include one or more of the following: analysis identifier, or VFL capability type. Correspondingly, network element discovery request 2 may also include analysis ID and / or VFL capability type.

[0242] S403, the NRF determines candidate NWDAFs based on Network Element Discovery Request 2. The candidate NWDAF may be allowed to be discovered by the AF. Alternatively, the candidate NWDAF may provide services or VFL services to the AF.

[0243] S404, the NRF sends a Network Element Discovery Response 2 to the NEF. This Network Element Discovery Response 2 includes information about candidate NWDAFs.

[0244] In one possible implementation, the NRF determines candidate NWDAFs based on the AF information in the aforementioned Network Element Discovery Request 2. For example, if the AF information is its identifier, the NRF can verify whether the allowed NFc list in the locally stored NWDAF profile includes the AF identifier; if the allowed NFc list in the profile of an NWDAF includes the AF identifier, the NRF determines that NWDAF as a candidate NWDAF. As another example, if the AF information is its type, the NRF can verify whether the allowed NFc types in the locally stored NWDAF list include the AF type; if the allowed NFc types in the allowed NFc list of an NWDAF include the AF type, the NRF determines that NWDAF as a candidate NWDAF. As yet another example, if the AF information is its domain name, the NRF can verify whether the allowed NFc Domains in the locally stored NWDAF list include the AF domain name; if the allowed NFc Domains in the allowed NFc list of an NWDAF include the AF domain name, the NRF determines that NWDAF as a candidate NWDAF. The NRF can then send a Network Element Discovery Response 2 to the NEF. This Network Element Discovery Response 2 may include the profile of the candidate NWDAF.

[0245] In one possible implementation, the aforementioned network element discovery request 2 may further include a VFL capability type. The NRF can determine a candidate NWDAF based on the AF information and VFL capability type in the network element discovery request 2. The candidate NWDAF matches the VFL capability type. If the aforementioned network element discovery request 2 may further include an analysis ID, then the analysis identifier supported by the candidate NWDAF is the same as the analysis identifier in the network element discovery request 2.

[0246] In one possible implementation, the aforementioned network element discovery request 2 may further include indication information, which can be used to indicate the allocation of a first identifier (such as a temporary NWDAF ID). The NRF can determine whether to allocate a first identifier for the candidate NWDAF based on the indication information in the network element discovery request 2. Then, the NRF can send a network element discovery response 2 to the NEF. This response 2 may include the candidate NWDAF's profile and its first identifier. The candidate NWDAF's profile includes its second identifier (NF instance ID).

[0247] S405, NEF sends Network Element Discovery Response 1 to AF. This Network Element Discovery Response 1 includes the identifier of the candidate NWDAF. The identifier of the candidate NWDAF can be a second identifier (such as NF instance ID) or a first identifier (such as temporary NWDAF ID).

[0248] In one possible implementation, if the AF is trusted, the identifier of the candidate NWDAF in the above-mentioned network element discovery response 1 is the second identifier; if the AF is untrusted, the identifier of the candidate NWDAF in the above-mentioned network element discovery response 1 is the first identifier.

[0249] In this embodiment of the application, during the network element discovery process, the NEF, representing the AF, discovers candidate NWDAFs to the NRF and informs the NRF of the AF's information. The NRF then filters NWDAFs based on the AF's information. This allows the NRF to filter out NWDAFs that are allowed to be discovered by the AF, preventing the AF from discovering NWDAFs that it is not allowed to discover. For example, filtering out NWDAFs whose allowed NFc includes the AF improves the security of the NWAF.

[0250] Referring to Figure 8, which is a fifth flowchart illustrating the communication method in vertical federated learning provided in this application embodiment, in this method, the NWDAF acts as a VFL client and the AF acts as a VFL server. For example, this AF is untrusted.

[0251] As shown in Figure 8, the communication method in this vertical federated learning includes, but is not limited to, the following steps:

[0252] S501, the AF sends a VFL training request to the NEF. This VFL training request includes the first identifier of the NWDAF participating in the VFL task. The VFL task is used to train the model. This VFL training request is used to request the NWDAF participating in the VFL task to train the model.

[0253] S502, NEF sends the VFL training response to AF.

[0254] In one possible implementation, after the VFL client discovery process is completed, the AF can determine the VFL members for a VFL task, that is, determine the NWDAFs participating in the VFL task. This VFL task can be used to train a model. For example, the VFL client discovery process in this embodiment can refer to existing technology, as shown in Figure 3 above; or it can refer to the relevant descriptions of any of the preceding embodiments, which will not be repeated here. Then, the AF can send a VFL training request to the NEF. This VFL training request may include a first identifier (such as a temporary NWDAF ID) of the NWDAFs participating in the VFL task. This VFL training request can be used to request the NWDAFs participating in the VFL task to train the model. The NEF can then send VFL training requests to each NWDAF participating in the VFL task, requesting each NWDAF participating in the VFL task to train the model. The NEF can then send a VFL training response to the AF.

[0255] S503, AF stores the mapping relationship between the NEF identifier and the VFL task identifier.

[0256] S504, the AF sends a VFL inference / prediction request to the NEF corresponding to the identifier of the VFL task. This VFL inference / prediction request includes the first identifier of the NWDAF participating in the VFL task. This VFL inference / prediction request is used to request the services provided by the aforementioned model.

[0257] In one possible implementation, after receiving the VFL training response, the AF determines that VFL training is complete. The AF can store the mapping relationship between the NEF's identifier and the VFL task's identifier. Then, if the AF determines that VFL inference / prediction is needed, it can send a VFL inference / prediction request to the NEF corresponding to the VFL task's identifier based on the stored mapping relationship. This VFL inference / prediction request may include the first identifier of the NWDAF participating in the VFL task. This VFL inference / prediction request can be used to request services provided by the model trained for the VFL task.

[0258] It is understood that by constraining the VFL training and inference processes to use the same NEF, the embodiments of this application can reduce the possibility of VFL inference process failure and improve network stability.

[0259] The foregoing details the method of this application. To facilitate better implementation of the above-described solutions in the embodiments of this application, corresponding apparatus or devices are also provided.

[0260] This application embodiment can divide the network elements of this application into functional modules according to the above method example. The network elements can also be divided into functional modules according to the above method example. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing module. The integrated modules can be implemented in hardware or as software functional modules. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation. The communication device of this application embodiment will be described in detail below with reference to Figures 9 to 11.

[0261] Referring to Figure 9, which is a schematic diagram of a communication device provided in an embodiment of this application, the communication device includes a transceiver module 10 and a processing module 20. The transceiver module 10 can implement corresponding communication functions, and the processing module 20 is used for data processing. The transceiver module 10 can also be referred to as a communication interface or a communication unit, etc.

[0262] In some embodiments of this application, the communication device may be the NEF shown above. That is, the communication device shown in FIG9 may be used to perform the steps or functions performed by the NEF in the above method embodiments. Exemplarily, the communication device may be the NEF or a chip or functional module configured in the NEF, etc., and this application embodiment does not limit this. The transceiver module 10 is used to perform NEF transceiver-related operations in the above method embodiments, and the processing module 20 is used to perform NEF processing-related operations in the above method embodiments.

[0263] In one design, transceiver module 10 is configured to receive a first request message, which includes a first identifier of a first Network Data Analysis Function (NWDAF) and an identifier of a Vertical Federated Learning (VFL) task used to train a model, and the first request message is used to request services provided by the model. Transceiver module 10 is further configured to send a second request message to a UDR / UDM, which includes the identifier of the VFL task. Transceiver module 10 is also configured to receive a second response message from the UDR / UDM, which includes a mapping relationship between the first identifier of one or more NWDAFs and the second identifier of each NWDAF, wherein the one or more NWDAFs participate in the VFL task and include the first NWDAF. Transceiver module 10 is further configured to send a third request message to the first NWDAF based on the second identifier of the first NWDAF, the third request message being used to request services provided by the model. For example, the third request message includes the second identifier corresponding to the first identifier of the first NWDAF.

[0264] The processing module 20 is used to generate the second request message and the third request message.

[0265] For example, the processing module 20 is further configured to determine that there is no second identifier corresponding to the first identifier of the first NWDAF locally.

[0266] For example, the processing module 20 is also used to store the mapping relationship between the first identifier of each NWDAF and the second identifier of each NWDAF in the one or more NWDAFs.

[0267] For example, the transceiver module 10 is further configured to receive a fourth request message, the fourth request message including a first identifier of the second NWDAF; the transceiver module 10 is further configured to, when one or more of the aforementioned NWDAFs include the second NWDAF, send a fifth request message to the second NWDAF according to the second identifier of the second NWDAF, the fifth request message being used to request the services provided by the model. For example, the fifth request message includes a second identifier corresponding to the first identifier of the second NWDAF stored locally.

[0268] For example, the transceiver module 10 is further configured to receive a first message, the first message including a first identifier of the one or more NWDAFs and an identifier of the VFL task; the processing module 20 is further configured to obtain a second identifier of the one or more NWDAFs; the transceiver module 10 is further configured to send a storage request to the UDR / UDM, the storage request including the first identifier of the one or more NWDAFs, the identifier of the VFL task, and the second identifier of the one or more NWDAFs.

[0269] For example, the transceiver module 10 is further configured to receive a network element discovery request, which is used to discover candidate NWDAFs, and the network element discovery request includes an analysis identifier; the processing module 20 is further configured to assign a first identifier to the candidate NWDAF; the transceiver module 10 is further configured to send a network element discovery response to the application function AF, which includes the first identifier of the candidate NWDAF. The candidate NWDAF includes one or more of the aforementioned NWDAFs, and the VFL task is used to train the model corresponding to the analysis identifier.

[0270] In this embodiment of the application, for specific descriptions of each request message, response message, first identifier, second identifier, storage request, first message, network element discovery request, and network element discovery response, please refer to the method embodiment shown in Figure 4 above, which will not be described in detail here.

[0271] It is understood that the specific descriptions of the transceiver module and processing module shown in the embodiments of this application are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the method embodiment shown in Figure 4 above, which will not be described in detail here. In addition, the technical effects of the embodiments of this application are the same as those in the method embodiment shown in Figure 4 above, and will not be repeated here for the sake of brevity.

[0272] In another design, the transceiver module 10 is used to receive a first request message, which includes a first identifier of the Network Data Analysis Function (NWDAF) participating in the Vertical Federated Learning (VFL) task used to train a model, and the first request message is used to request services provided by the model. The transceiver module 10 is also used to send a first network element discovery request to the NRF, which includes the first identifier of the NWDAF participating in the VFL task. The transceiver module 10 is also used to receive a first network element discovery response, which includes a second identifier corresponding to the first identifier of the NWDAF participating in the VFL task. The transceiver module 10 is further used to send a second request message to the NWDAF participating in the VFL task based on the second identifier corresponding to the first identifier of the NWDAF participating in the VFL task, and the second request message is used to request services provided by the model. For example, the second request message includes the second identifier corresponding to the first identifier of the NWDAF participating in the VFL task.

[0273] The processing module 20 is used to generate the first network element discovery request and the second request message.

[0274] For example, the processing module 20 is further configured to determine that there is no second identifier corresponding to the first identifier of the NWDAF participating in the VFL task locally.

[0275] For example, the transceiver module 10 is further configured to send a second network element discovery request to the NRF, the second network element discovery request including second information and / or AF information, the second information being used to indicate the allocation of a first identifier; the second network element discovery request also includes a VFL capability type, the second network element discovery request being used to discover candidate NWDAFs; the transceiver module 10 is further configured to receive a second network element discovery response, the second network element discovery response including a first identifier and a second identifier of the candidate NWDAF; the candidate NWDAF matching the VFL capability type; the transceiver module 10 is further configured to send the first identifier of the candidate NWDAF to the AF, the candidate NWDAF including the NWDAF participating in the VFL task.

[0276] In this embodiment of the application, the specific descriptions of each request message, the first identifier, the second identifier, each network element discovery request, and each network element discovery response can be found in the method embodiment shown in Figure 6 above, and will not be described in detail here.

[0277] It is understood that the specific descriptions of the transceiver module and processing module shown in the embodiments of this application are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the method embodiment shown in Figure 6 above, which will not be described in detail here. In addition, the technical effects of the embodiments of this application are the same as those in the method embodiment shown in Figure 6 above, and will not be repeated here for the sake of brevity.

[0278] Reusing Figure 9, in some other embodiments of this application, the communication device may be the UDR / UDM shown above. That is, the communication device shown in Figure 9 can be used to perform the steps or functions performed by the UDR / UDM in the above method embodiments. For example, the communication device may be a UDR / UDM or a chip or functional module configured in the UDR / UDM, etc., and this application embodiment does not limit this. The transceiver module 10 is used to perform UDR / UDM transmission and reception related operations in the above method embodiments, and the processing module 20 is used to perform UDR / UDM processing related operations in the above method embodiments.

[0279] For example, the transceiver module 10 is used to receive a second request message, which includes an identifier of a VFL task used to train a model; the transceiver module 10 is also used to send a second response message to a first NEF, which includes a mapping relationship between a first identifier of each NWDAF in one or more NWDAFs and a second identifier of each NWDAF, wherein the one or more NWDAFs participate in the VFL task, and the one or more NWDAFs include a first NWDAF.

[0280] The processing module 20 is used to generate the second response message.

[0281] For example, the transceiver module 10 is further configured to receive a first storage request from a first NEF, the first storage request including the first and second identifiers of the one or more NWDAFs and the identifier of the VFL task; the processing module 20 is further configured to store the mapping relationship between the first and second identifiers of the one or more NWDAFs and the identifier of the VFL task.

[0282] For example, the transceiver module 10 is further configured to receive a fifth request message from the second NEF, the fifth request message including the identifier of the aforementioned VFL task; the transceiver module 10 is further configured to send a fifth response message to the second NEF, the fifth response message including a mapping relationship between a first identifier and a second identifier of each of one or more NWDAFs. These one or more NWDAFs participate in the aforementioned VFL task, and these one or more NWDAFs include a first NWDAF.

[0283] In this embodiment of the application, the specific descriptions of the second request message, the second response message, the first identifier, the second identifier, the first storage request, and the second storage request can be referred to the method embodiment shown in Figure 4 above, and will not be described in detail here.

[0284] It is understood that the specific descriptions of the transceiver module and processing module shown in the embodiments of this application are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the method embodiment shown in Figure 4 above, which will not be described in detail here. In addition, the technical effects of the embodiments of this application are the same as those in the method embodiment shown in Figure 4 above, and will not be repeated here for the sake of brevity.

[0285] Reusing Figure 9, in some other embodiments of this application, the communication device may be the NRF shown above. That is, the communication device shown in Figure 9 may be used to perform the steps or functions performed by the NRF in the above method embodiments. For example, the communication device may be a UDR / UDM or a chip or functional module configured in the NRF, etc., and this application embodiment does not limit this. The transceiver module 10 is used to perform NRF transceiver-related operations in the above method embodiments, and the processing module 20 is used to perform NRF processing-related operations in the above method embodiments.

[0286] In one design, transceiver module 10 is used to receive a first network element discovery request, which includes a first identifier of an NWDAF participating in a VFL task, the VFL task being used to train a model; processing module 20 is used to determine a second identifier of the NWDAF participating in the VFL task based on the first identifier of the NWDAF participating in the VFL task; transceiver module 10 is also used to send a first network element discovery response to a first NEF, which includes a second identifier corresponding to the first identifier of the NWDAF participating in the VFL task.

[0287] For example, the transceiver module 10 is further configured to receive a second network element discovery request, which includes second information and / or AF information, the second information being used to indicate the allocation of a first identifier; the second network element discovery request also includes a VFL capability type, which is used to discover candidate NWDAFs; the processing module 20 is further configured to determine the candidate NWDAF and its second identifier based on the VFL capability type; the candidate NWDAF matches the VFL capability type; the processing module 20 is further configured to determine, based on the second information or the AF information, to allocate a first identifier to the candidate NWDAF; the transceiver module 10 is further configured to send a second network element discovery response to the second NEF, which includes the first identifier and second identifier of the candidate NWDAF. The candidate NWDAF includes the NWDAF participating in the VFL task.

[0288] In this embodiment of the application, the specific descriptions of the first identifier, the second identifier, the discovery requests of each network element, and the discovery responses of each network element can be found in the method embodiment shown in Figure 6 above, and will not be described in detail here.

[0289] It is understood that the specific descriptions of the transceiver module and processing module shown in the embodiments of this application are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the method embodiment shown in Figure 6 above, which will not be described in detail here. In addition, the technical effects of the embodiments of this application are the same as those in the method embodiment shown in Figure 6 above, and will not be repeated here for the sake of brevity.

[0290] In another design, the transceiver module 10 is used to receive a network element discovery request, which includes information about the AF (Active Function) and is used to discover candidate NWDAFs. The processing module 20 is used to determine the candidate NWDAF based on the network element discovery request, and the allowed network function consumers of the candidate NWDAF include the AF. The transceiver module 10 is also used to send a network element discovery response to the NEF (Network Element Discovery Response), which includes information about the candidate NWDAF.

[0291] For example, the network element discovery request further includes indication information, which is used to indicate the allocation of a first identifier. The processing module 20 is further configured to determine, based on the indication information, to allocate a first identifier to the candidate NWDAF; the information of the candidate NWDAF includes the first identifier and the second identifier of the candidate NWDAF.

[0292] In this application embodiment, the specific descriptions of the first identifier, the second identifier, the network element discovery request, and the network element discovery response can be found in the method embodiment shown in Figure 7 above, and will not be described in detail here.

[0293] It is understood that the specific descriptions of the transceiver module and processing module shown in the embodiments of this application are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the method embodiment shown in Figure 7 above, which will not be described in detail here. In addition, the technical effects of the embodiments of this application are the same as those in the method embodiment shown in Figure 7 above, and will not be repeated here for the sake of brevity.

[0294] Reusing Figure 9, in some further embodiments of this application, the communication device may be the AF shown above. That is, the communication device shown in Figure 9 may be used to perform the steps or functions performed by the AF in the above method embodiments. For example, the communication device may be a UDR / UDM or a chip or functional module configured in the AF, etc., and this application embodiment does not limit this. The transceiver module 10 is used to perform AF transmission and reception related operations in the above method embodiments, and the processing module 20 is used to perform AF processing related operations in the above method embodiments.

[0295] For example, the transceiver module 10 is used to send a VFL training request to the NEF, the VFL training request including a first identifier of the NWDAF participating in the VFL task, the VFL task being used to train the model, and the VFL training request being used to request the NWDAF participating in the VFL task to train the model; the transceiver module 10 is also used to receive a VFL training response from the NEF; the processing module 20 is used to save the mapping relationship between the identifier of the NEF and the identifier of the VFL task; the transceiver module 10 is also used to send a VFL inference request to the NEF corresponding to the identifier of the VFL task, the VFL inference request including the first identifier, and the VFL inference request being used to request the services provided by the model.

[0296] In this embodiment of the application, the specific descriptions of the VFL task identifier, the first identifier, the VFL training request, and the VFL inference request can be found in the method embodiment shown in Figure 8 above, and will not be described in detail here.

[0297] It is understood that the specific descriptions of the transceiver module and processing module shown in the embodiments of this application are merely examples. For the specific functions or execution steps of the transceiver module and processing module, please refer to the method embodiment shown in Figure 8 above, which will not be described in detail here. In addition, the technical effects of the embodiments of this application are the same as those in the method embodiment shown in Figure 8 above, and will not be repeated here for the sake of brevity.

[0298] The communication device of the present application embodiments has been described above. The following describes possible product forms of the communication device. It should be understood that any product possessing the functions of the communication device described in FIG9 above falls within the protection scope of the present application embodiments. It should also be understood that the following description is merely illustrative and does not limit the product form of the communication device of the present application embodiments to this.

[0299] In one possible implementation, in the communication device shown in FIG9, the processing module 20 can be one or more processors, the transceiver module 10 can be a transceiver, or the transceiver module 10 can also be a transmitting module and a receiving module. The transmitting module can be a transmitter, and the receiving module can be a receiver. The transmitting module and the receiving module are integrated into one device, such as a transceiver. In the embodiments of this application, the processor and the transceiver can be coupled, etc., and the connection method between the processor and the transceiver is not limited in the embodiments of this application. In the process of executing the above method, the process of sending information in the above method can be understood as the process of the processor outputting the above information. When outputting the above information, the processor outputs the above information to the transceiver so that the transceiver can transmit it. After the above information is output by the processor, it may need to undergo other processing before reaching the transceiver. Similarly, the process of receiving information in the above method can be understood as the process of the processor receiving the input above information. When the processor receives the input information, the transceiver receives the above information and inputs it into the processor. Furthermore, after the transceiver receives the above information, the above information may need to undergo other processing before being input into the processor.

[0300] Referring to Figure 10, which is another structural schematic diagram of the communication device provided in an embodiment of this application, the communication device provided in this application embodiment can be used to implement the methods described in the above method embodiments, as can be seen from the description in the above method embodiments. The communication device can be the aforementioned NEF, UDR / UDM, NRF, or AF, or a chip or circuit thereof. Exemplarily, the communication device includes one or more processors 1001 and transceivers 1002. The communication device may further include a memory 1003. In one implementation, the communication device also includes an input / output device (not shown in the figure).

[0301] The processor 1001 is mainly used to process communication protocols and communication data, control the entire communication device, execute software programs, and process the data of the software programs. The memory 1003 is mainly used to store software programs and data. The transceiver 1002 may include control circuitry and an antenna. The control circuitry is mainly used for converting baseband signals to radio frequency signals and processing radio frequency signals. The antenna is mainly used for transmitting and receiving radio frequency signals in the form of electromagnetic waves. Input / output devices, such as touchscreens, displays, and keyboards, are mainly used to receive user input data and output data to the user.

[0302] When the communication device is powered on, the processor 1001 can read the software program in the memory 1003, interpret and execute the instructions of the software program, and process the data of the software program. When data needs to be transmitted wirelessly, the processor 1001 performs baseband processing on the data to be transmitted and outputs the baseband signal to the radio frequency (RF) circuit. The RF circuit processes the baseband signal and transmits the RF signal outward in the form of electromagnetic waves through the antenna. When data is sent to the communication device, the RF circuit receives the RF signal through the antenna, converts the RF signal into a baseband signal, and outputs the baseband signal to the processor 1001. The processor 1001 converts the baseband signal into data and processes the data.

[0303] In another implementation, the radio frequency circuitry and antenna can be set up independently of the processor performing baseband processing. For example, in a distributed scenario, the radio frequency circuitry and antenna can be arranged remotely, independent of the communication device.

[0304] The processor 1001, transceiver 1002, and memory 1003 can be connected via a communication bus.

[0305] For example, when the communication device is used to perform the steps, methods, or functions performed by NEF1 in the method embodiment shown in FIG4 above, the processor 1001 may be used to generate a storage request, a second request message, and a third request message, and / or to perform other processes of the technology described herein; the transceiver 1002 may be used to perform steps S102, S105, and S107 in FIG4, and / or to perform other processes of the technology described herein.

[0306] For example, when the communication device is used to perform the steps, methods, or functions performed by UDR / UDM in the method embodiment shown in FIG4 above, the processor 1001 may be used to perform step S103 in FIG4, and / or to perform other processes of the technology described herein; the transceiver 1002 may be used to perform step S106 in FIG4, and / or to perform other processes of the technology described herein.

[0307] For example, when the communication device is used to perform the steps, methods, or functions performed by NEF1 in the method embodiment shown in FIG5 above, the processor 1001 may be used to perform steps S202 and S203 in FIG5, and / or to perform other processes of the technology described herein; the transceiver 1002 may be used to perform steps S204 and S205 in FIG5, and / or to perform other processes of the technology described herein.

[0308] For example, when the communication device is used to perform the steps, methods, or functions executed by UDR / UDM in the method embodiment shown in FIG5 above, the processor 1001 may be used to execute step S206 in FIG5, and / or to execute other processes of the technology described herein; the transceiver 1002 may be used to execute step S209 in FIG5, and / or to execute other processes of the technology described herein.

[0309] For example, when the communication device is used to perform the steps, methods, or functions performed by NEF2 in the method embodiment shown in FIG5 above, the processor 1001 may be used to generate a second request message and a third request message, and / or to perform other processes of the technology described herein; the transceiver 1002 may be used to perform steps S208 and S210 in FIG5, and / or to perform other processes of the technology described herein.

[0310] For example, when the communication device is used to execute the steps, methods, or functions performed by NEF1 in the method embodiment shown in FIG6 above, the processor 1001 can be used to generate a first network element discovery request, a second network element discovery request, and a second request message, and / or to execute other processes of the technology described herein; the transceiver 1002 can be used to execute steps S301, S305, S307, and S309 in FIG6, and / or to execute other processes of the technology described herein.

[0311] For example, when the communication device is used to perform the steps, methods, or functions executed by the NRF in the method embodiment shown in FIG6 above, the processor 1001 may be used to execute steps S302 and S303 in FIG6, and / or to execute other processes of the technology described herein; the transceiver 1002 may be used to execute steps S304 and S308 in FIG6, and / or to execute other processes of the technology described herein.

[0312] For example, when the communication device is used to perform the steps, methods, or functions performed by NEF in the method embodiment shown in FIG7 above, the processor 1001 may be used to generate network element discovery request 2, and / or to perform other processes of the technology described herein; the transceiver 1002 may be used to perform steps S402 and S405 in FIG7, and / or to perform other processes of the technology described herein.

[0313] For example, when the communication device is used to perform the steps, methods, or functions performed by the NRF in the method embodiment shown in FIG7 above, the processor 1001 may be used to perform step S403 in FIG7, and / or to perform other processes of the technology described herein; the transceiver 1002 may be used to perform step S404 in FIG7, and / or to perform other processes of the technology described herein.

[0314] For example, when the communication device is used to perform the steps, methods, or functions executed by AF in the method embodiment shown in FIG8 above, the processor 1001 may be used to execute step S503 in FIG8, and / or to execute other processes of the technology described herein; the transceiver 1002 may be used to execute steps S501 and S504 in FIG8, and / or to execute other processes of the technology described herein.

[0315] In any of the above implementations, the processor 1001 may include a transceiver for implementing receiving and transmitting functions. For example, the transceiver may be a transceiver circuit, an interface, or an interface circuit. The transceiver circuit, interface, or interface circuit for implementing receiving and transmitting functions may be separate or integrated. The aforementioned transceiver circuit, interface, or interface circuit can be used for reading and writing code / data, or it can be used for transmitting or relaying signals.

[0316] In any of the above implementations, the processor 1001 may store instructions, which may be computer programs. These computer programs, running on the processor 1001, cause the communication device to execute the methods described in the above method embodiments. The computer program may be embedded in the processor 1001; in this case, the processor 1001 may be implemented in hardware.

[0317] In one implementation, the communication device may include circuitry capable of performing the functions of transmitting, receiving, or communicating as described in the foregoing method embodiments. The processor and transceiver described in this application can be implemented on integrated circuits (ICs), analog ICs, radio frequency integrated circuits (RFICs), mixed-signal ICs, application-specific integrated circuits (ASICs), printed circuit boards (PCBs), electronic devices, etc. The processor and transceiver can also be manufactured using various IC process technologies, such as complementary metal oxide semiconductors (CMOS), n-metal-oxide-semiconductor (NMOS), positive channel metal oxide semiconductors (PMOS), bipolar junction transistors (BJTs), bipolar CMOS (BiCMOS), silicon germanium (SiGe), gallium arsenide (GaAs), etc.

[0318] It is understood that the communication device shown in the embodiments of this application may have more components than those in Figure 10, and the embodiments of this application do not limit this. The methods performed by the processor and transceiver shown above are merely examples, and the specific steps performed by the processor and transceiver can be referred to the description of the method embodiments above.

[0319] In another possible implementation, in the communication device shown in FIG9, the processing module 20 can be one or more processors, and the transceiver module 10 can be an input / output interface, or a communication interface, or an interface circuit, or an interface, etc. Alternatively, the transceiver module 10 can also be a transmitting module and a receiving module. The transmitting module can be an output interface, and the receiving module can be an input interface. The transmitting module and the receiving module are integrated into one module, such as an input / output interface. Referring to FIG11, FIG11 is another structural schematic diagram of the communication device provided in the embodiments of this application. As shown in FIG11, the communication device can be any of the aforementioned NEF, UDR / UDM, NRF, and AF. The communication device shown in FIG11 includes a processor 901 and an interface circuit 902. That is, the above-mentioned processing module 20 can be implemented with the processor 901, and the transceiver module 10 can be implemented with the interface circuit 902. Among them, the processor 901 can be a chip, processing circuit, integrated circuit, or system on chip (SoC) chip, etc., and the interface circuit 902 can be a communication interface circuit, an input / output interface circuit, pins, etc. For example, Figure 11 shows the communication device described above as a chip, which includes a processor 901 and an interface circuit 902.

[0320] In this embodiment, the processor and interface circuit can also be coupled to each other. The specific connection method between the processor and interface circuit is not limited in this embodiment. For example, the communication device shown in FIG11 can be used to implement the method described in any of the above method embodiments; please refer to the description in the above method embodiments.

[0321] It is understood that the communication device shown in the embodiments of this application can implement the method provided in the embodiments of this application in hardware form or in software form, etc., and the embodiments of this application do not limit it in this way.

[0322] For specific implementations of the various embodiments shown in Figure 11, please refer to the above embodiments, which will not be described in detail here.

[0323] This application also provides a communication system comprising a NEF and a UDR / UDM, which can be used to perform the methods in the method embodiments shown in FIG4 or FIG5. Alternatively, the communication system comprises a NEF and an NRF, which can be used to perform the methods in the method embodiments shown in FIG6 or FIG7. Alternatively, the communication system comprises a NEF and an AF, which can be used to perform the methods in the method embodiments shown in FIG8.

[0324] In addition, this application also provides a computer program for implementing the operations and / or processes performed by the aforementioned network elements in the method provided in this application.

[0325] This application also provides a computer-readable storage medium storing computer code, which, when executed on a computer, causes the computer to perform the operations and / or processes performed by the aforementioned network elements in the method provided in this application.

[0326] This application also provides a computer program product, which includes computer code or a computer program, which, when run on a computer, causes the operations and / or processes performed by the aforementioned network elements in the method provided in this application to be executed.

[0327] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, or it may be an electrical, mechanical, or other form of connection.

[0328] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the technical effects of the solutions provided in the embodiments of this application.

[0329] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0330] If the integrated unit is implemented as 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 all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned readable storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0331] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A communication method in vertical federated learning, characterized in that, include: Receive a first request message, the first request message including a first identifier of a first network data analysis function NWDAF and an identifier of a longitudinal federated learning VFL task, the VFL task being used to train a model, the first request message being used to request services provided by the model; Send a second request message to the data processing network element, the second request message including the identifier of the VFL task; Receive a second response message from the data processing network element. The second response message includes a mapping relationship between a first identifier of each NWDAF and a second identifier of each NWDAF in one or more NWDAFs. The one or more NWDAFs participate in the VFL task. The one or more NWDAFs include the first NWDAF. Based on the second identifier of the first NWDAF, a third request message is sent to the first NWDAF, the third request message being used to request the services provided by the model.

2. The method according to claim 1, characterized in that, The first identifier is used to identify NWDAF outside the Public Land Mobile Network (PLMN), and the second identifier is used to identify NWDAF inside the PLMN.

3. The method according to claim 1 or 2, characterized in that, Before sending the second request message to the data processing network element, the method further includes: It is determined that there is no second identifier corresponding to the first identifier of the first NWDAF locally.

4. The method according to any one of claims 1 to 3, characterized in that, After receiving the second response message from the data processing network element, the method further includes: Store the mapping relationship between the first identifier and the second identifier of each NWDAF in the one or more NWDAFs.

5. The method according to claim 4, characterized in that, After receiving the second response message from the data processing network element, the method further includes: Receive a fourth request message, the fourth request message including the first identifier of the second NWDAF; If the one or more NWDAFs include the second NWDAF, then a fifth request message is sent to the second NWDAF according to the second identifier of the second NWDAF, the fifth request message being used to request the services provided by the model.

6. The method according to any one of claims 1 to 5, characterized in that, Before receiving the first request message, the method further includes: Receive a first message, the first message including a first identifier of each NWDAF in the one or more NWDAFs and an identifier of the VFL task; Obtain a second identifier for participation in one or more NWDAFs; A storage request is sent to the data processing network element. The storage request includes a first identifier of the one or more NWDAFs, an identifier of the VFL task, and a second identifier of the one or more NWDAFs.

7. The method according to claim 6, characterized in that, Before receiving the first message, the method further includes: Receive a network element discovery request, the network element discovery request being used to discover candidate NWDAFs, the network element discovery request including an analysis identifier; Assign a first identifier to the candidate NWDAF; Send a network element discovery response to the application function AF, the network element discovery response including the first identifier of the candidate NWDAF; The candidate NWDAFs include one or more NWDAFs, and the VFL task is used to train the model corresponding to the analysis identifier.

8. A communication method in vertical federated learning, characterized in that, include: Receive a second request message, the second request message including an identifier of a longitudinal federated learning (VFL) task, the VFL task being used to train the model; Send a second response message to the first Network Open Function (NEF). The second response message includes a mapping relationship between the first identifier of each NWDAF and the second identifier of each NWDAF in one or more Network Data Analysis Functions (NWDAFs). The one or more NWDAFs participate in the VFL task. The one or more NWDAFs include the first NWDAF.

9. The method according to claim 8, characterized in that, The first identifier is used to identify NWDAF outside the Public Land Mobile Network (PLMN), and the second identifier is used to identify NWDAF inside the PLMN.

10. The method according to claim 8 or 9, characterized in that, Before receiving the second request message, the method further includes: Receive a storage request from the first NEF, the storage request including a first identifier and a second identifier of each of the one or more NWDAFs, and an identifier of the VFL task; The system stores a mapping relationship between the first and second identifiers of each NWDAF participating in the one or more NWDAFs, and the identifier of the VFL task.

11. The method according to any one of claims 9 to 10, characterized in that, The method further includes: Receive a fifth request message from the second NEF, the fifth request message including the identifier of the VFL task; Send a fifth response message to the second NEF. The fifth response message includes a mapping relationship between a first identifier and a second identifier of each NWDAF in one or more NWDAFs. The one or more NWDAFs participate in the VFL task. The one or more NWDAFs include a first NWDAF.

12. The method according to claim 10 or 11, characterized in that, The one or more NWDAFs also include a second NWDAF.

13. A communication method in vertical federated learning, characterized in that, include: Receive a network element discovery request, the network element discovery request including information of application function AF, the network element discovery request is used to discover candidate network data analysis function NWDAF; Based on the network element discovery request, candidate NWDAFs are determined, and the candidate NWDAFs are able to provide services to the AF; Send a network element discovery response to the Network Open Function (NEF), the network element discovery response including information about the candidate NWDAF.

14. The method according to claim 13, characterized in that, The network element discovery request may also include one or more of the following: analysis identifier, or vertical federated learning VFL capability type; the candidate NWDAF matches the VFL capability type, and the analysis identifier supported by the candidate NWDAF is the same as the analysis identifier in the network element discovery request.

15. The method according to claim 13 or 14, characterized in that, The information of the AF is the type, identifier, or domain name of the AF; The candidate list of allowed network function consumers for NWDAFs includes the identifier of the AF; or the candidate list of allowed network function consumer domain names for NWDAFs includes the domain name of the AF; or the candidate list of allowed network function consumer types for NWDAFs includes the type of the AF.

16. The method according to any one of claims 13 to 15, characterized in that, The network element discovery request also includes indication information, which is used to indicate the allocation of a first identifier; After receiving the network element discovery request, the method further includes: Based on the indicated information, a first identifier is assigned to the candidate NWDAF; The information of the candidate NWDAF includes the first identifier and the second identifier of the candidate NWDAF.

17. The method according to claim 16, characterized in that, The first identifier is used to identify NWDAF outside the Public Land Mobile Network (PLMN), and the second identifier is used to identify NWDAF inside the PLMN.

18. A communication method in vertical federated learning, characterized in that, include: The Application Function (AF) sends a Vertical Federated Learning (VFL) training request to the Network Open Function (NEF). The VFL training request includes a first identifier of the Network Data Analysis Function (NWDAF) participating in the VFL task. The VFL task is used to train a model. The VFL training request is used to request the NWDAF participating in the VFL task to train the model. The AF receives the VFL training response from the NEF; The AF stores the mapping relationship between the NEF identifier and the VFL task identifier; The AF sends a VFL inference request to the NEF corresponding to the identifier of the VFL task. The VFL inference request includes the first identifier and is used to request the services provided by the model.

19. A communication device, characterized in that, Includes modules for performing the method according to any one of claims 1 to 18.

20. A readable storage medium, characterized in that, The readable storage medium stores program instructions that, when executed on the communication device, cause the communication device to perform the method as described in any one of claims 1 to 18.

21. A computer program product, characterized in that, When the computer program product is executed, the method as described in any one of claims 1 to 18 is performed.

22. A communication system, characterized in that, It includes one or more of the following means: means for performing the method of any one of claims 1 to 7, means for performing the method of any one of claims 8 to 12, means for performing the method of any one of claims 13 to 17, or means for performing the method of claim 18.

Images