Network element registration method, model determination method, network element, and readable storage medium

Abstract

The present application discloses a network element registration method, a model determination method, an apparatus, a network element, a communication system, and a storage medium. The network element registration method of an embodiment of the present application includes a first network element sending a registration request to a second network element, the registration request is used to request that the federated learning capability information of the first network element be registered with the second network element, and the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

Description

[Technical field]

[0001] The present application relates to the field of communications technology, and specifically to a network element registration method, a model determination method, an apparatus, a network element, a communication system, and a storage medium. [Background technology]

[0002] With the continuous development of communication technology, federated learning technology has emerged, which allows multiple clients to cooperate with each other under the coordination of one central server to obtain a complete machine learning model. However, in the conventional technology, participants such as network elements, base stations, and user equipment (UE), which can be trained by default, can all participate in federated learning without additional reinforcement or discrimination of an artificial intelligence algorithm called federated learning, and such unfiltered selection of federated learning targets may cause problems such as missing federated learning targets and reducing learning efficiency due to various reasons (e.g., UE network signal is relatively poor, etc.). Therefore, how to reinforce the network element registration flow in federated learning is an issue that needs to be resolved as soon as possible. Summary of the Invention [Problem to be solved by the invention]

[0003] The embodiments of the present application provide a network element registration method, a model determination method, an apparatus, a network element, a communication system, and a storage medium, which can solve the problem of how to enhance the network element registration flow in federated learning. [Means for solving the problem]

[0004] According to a first aspect, a network element registration method is provided, the method including: a first network element sending a registration request to a second network element, the registration request being used to request that federated learning capability information of the first network element be registered with the second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0005] According to a second aspect, there is provided a network element registration device for use in a first network element, the network element registration device including: a sending module, the sending module being used for sending a registration request to a second network element, the registration request being used for requesting that federated learning capability information of the first network element be registered in the second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0006] According to a third aspect, a network element registration method is provided, the method including: a second network element receiving a registration request sent by a first network element, the registration request being used to request federated learning capability information of the first network element to be registered with the second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0007] According to a fourth aspect, there is provided a network element registration device for use in a second network element, the network element registration device including a receiving module, the receiving module being used for receiving a registration request sent by a first network element, the registration request being used for requesting to register federated learning capability information of the first network element with the second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0008] According to a fifth aspect, a model determination method is provided, the method including: a third network element sending a search request to a second network element, the search request being used to request a search for network elements capable of federated learning training, the search request including first information, the first information including at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information being used to indicate a network element type corresponding to a network element capable of federated learning training that needs to be searched, the network element type including a federated learning server network element type and / or a federated learning client network element type, and the network element number request information being used to indicate a request for the number of network elements capable of federated learning training that needs to be searched.

[0009] According to a sixth aspect, there is provided a model determination device, the model determination device including a sending module, the sending module is used for sending a search request to a second network element, the search request is used for requesting a search of a network element capable of federated learning training, the search request includes first information, the first information includes at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information is used for indicating a network element type corresponding to a network element capable of federated learning training that needs to be searched, the network element type includes a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used for indicating a request for the number of network elements capable of federated learning training that needs to be searched.

[0010] According to a seventh aspect, there is provided a network element comprising a processor and a memory, the memory storing a program or instructions operable on the processor, the program or instructions, when executed by the processor, realising the steps of the network element registration method described in the first aspect.

[0011] According to an eighth aspect, there is provided a network element including a processor and a communication interface, wherein the communication interface is used for sending a registration request to a second network element, the registration request is used for requesting that federated learning capability information of a first network element be registered with the second network element, and the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0012] According to a ninth aspect, there is provided a network element comprising a processor and a memory, the memory storing a program or instructions operable on the processor, the program or instructions, when executed by the processor, realising the steps of the network element registration method described in the third aspect.

[0013] According to a tenth aspect, a network element is provided, the network element including a processor and a communication interface, wherein the communication interface is used for receiving a registration request sent by a first network element, the registration request is used for requesting registration of federated learning capability information of the first network element with a second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0014] According to an eleventh aspect, there is provided a network element including a processor and a memory, the memory storing a program or instructions operable on the processor, the program or instructions, when executed by the processor, implementing the steps of the model determination method described in the fifth aspect.

[0015] According to a twelfth aspect, a network element is provided, the network element including a processor and a communication interface, wherein the communication interface is used for sending a search request to a second network element, the search request is used for requesting a search for network elements capable of federated learning training, the search request including first information, the first information including at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information is used to indicate a network element type corresponding to a network element capable of federated learning training that needs to be searched, the network element type including a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used to indicate a request for the number of network elements capable of federated learning training that needs to be searched.

[0016] According to a thirteenth aspect, there is provided a communication system, comprising a network element registration device according to the second aspect, e.g. a network element registration device according to the fourth aspect and a model determination device according to the sixth aspect, or comprising a network element according to the seventh, ninth and eleventh aspects, or comprising a network element according to the eighth, tenth and twelfth aspects, wherein said network element is adapted for performing the steps of the method according to the first, third and fifth aspects.

[0017] According to a fourteenth aspect, there is provided a readable storage medium, the readable storage medium storing a program or instructions which, when executed by a processor, perform the steps of the method according to the first aspect, or perform the steps of the method according to the third aspect, or perform the steps of the method according to the fifth aspect.

[0018] According to a fifteenth aspect, there is provided a chip, the chip including a processor and a communication interface, the communication interface coupled to the processor, the processor being adapted to run a program or instructions to implement the method of the first aspect, or to implement the method of the third aspect, or to implement the method of the fifth aspect.

[0019] According to a sixteenth aspect, there is provided a computer program / program product, the computer program / program product being stored in a storage medium, and the computer program / program product being executed by at least one processor to realize steps of the network element registration method described in the first aspect, or to realize steps of the network element registration method described in the third aspect, or to realize steps of the model determination method described in the fifth aspect. Effect of the Invention

[0020] In an embodiment of the present application, a first network element can register the federated learning capability information of the first network element to a second network element through a registration request, thereby realizing the registration of the first network element to the second network element, so that other network elements can search for the first network element capable of federated learning training by the second network element, and solve the problem of how the first network element can register its federated learning capability information to the second network element and be searched by other network elements.

[0021] By receiving the registration request sent by the first network element, the second network element can register the federated learning capability information of the first network element with the second network element, thereby realizing the registration of the first network element with the second network element, so that other network elements can search for the first network element capable of federated learning training, and solve the problem of how the first network element can register its federated learning capability information with the second network element and be searched by other network elements.

[0022] The third network element can search for a network element capable of associative learning training in the second network element through a search request, and since the search request includes the first information, the third network element can search for a target network element matching the first information, and perform associative learning training with the searched target network element capable of associative learning training. [Brief description of the drawings]

[0023] [Figure 1] 1 is a schematic architecture diagram of a wireless communication system according to an embodiment of the present application; [Diagram 2] FIG. 1 is a schematic diagram of the principle of horizontal association learning according to an embodiment of the present application. [Diagram 3] FIG. 2 is a schematic diagram of a neural network according to an embodiment of the present application. [Figure 4] FIG. 2 is a schematic diagram of a neuron according to an embodiment of the present application. [Diagram 5] 2 is a flowchart of a network element registration method according to an embodiment of the present application; [Figure 6] 2 is a second flowchart of a network element registration method according to an embodiment of the present application. [Figure 7] 3 is a third flowchart of a network element registration method according to an embodiment of the present application. [Figure 8] 4 is a fourth flowchart of a network element registration method according to an embodiment of the present application. [Figure 9] 1 is a flowchart of a model determination method according to an embodiment of the present application. [Figure 10] 2 is a second flowchart of a model determination method according to an embodiment of the present application. [Figure 11] 11 is a third flowchart of a model determination method according to an embodiment of the present application. [Figure 12] 4 is a fourth flowchart of a model determination method according to an embodiment of the present application. [Figure 13] 2 is a flowchart of a network element registration method and a model determination method according to an embodiment of the present application; [Figure 14] FIG. 2 is a structural schematic diagram of a network element registration device according to an embodiment of the present application; [Figure 15] FIG. 2 is a second structural schematic diagram of a network element registration device according to an embodiment of the present application; [Figure 16] FIG. 2 is a structural schematic diagram of a model determining device according to an embodiment of the present application; [Figure 17] 1 is a hardware structure schematic diagram of a communication device according to an embodiment of the present application; [Figure 18]FIG. 2 is a hardware structure schematic diagram of a network element according to an embodiment of the present application; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] The following clearly describes the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present application are all within the scope of protection of the present application.

[0025] The terms "first," "second," etc. in the specification and claims of the present application are intended to distinguish between similar objects and are not intended to describe a particular order or sequence. It is to be understood that the terms so used are interchangeable where appropriate, such that the embodiments of the present application may be performed in an order other than that shown or described herein, and that the objects distinguished by "first" and "second" are generally of the same type and do not limit the number of objects, e.g., the first object may be one or more. It is to be noted that "and / or" in the specification and claims represents at least one of the objects connected, and the character " / " generally represents an "or" relationship between the related objects.

[0026] It should be noted that the techniques described in the embodiments of the present application are not limited to Long Term Evolution (LTE) / LTE-Advanced (LTE-A) systems, but can also be applied to other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency Division Multiple Access (SC-FDMA), and other systems. The terms "system" and "network" in the embodiments of the present application are always used interchangeably, and the techniques described may be used in the above-mentioned systems and radio technologies, or in other systems and radio technologies. Although the following description describes a New Radio (NR) system for illustrative purposes and uses NR terminology in most of the description below, these techniques may be applied to applications other than NR system applications, such as 6th Generation (6G) communication systems.

[0027] 1 is a block diagram of a wireless communication system to which the embodiment of the present application can be applied. The wireless communication system includes a terminal 11 and a network side device 12. Here, the terminal 11 may be a terminal device such as a mobile phone, a tablet personal computer, a laptop computer (also called a notebook computer), a personal digital assistant (PDA), a palmtop computer, a netbook, an ultra-mobile personal computer (UMPC), a mobile Internet device (MID), an augmented reality (AR) / virtual reality (VR) device, a robot, a wearable device, a vehicle-mounted equipment (VUE), a pedestrian terminal (PUE), a smart home (home equipment having a wireless communication function, such as a refrigerator, a television, a washing machine, or furniture), a game console, a personal computer (PC), a deposit machine, or a self-service machine, and the wearable device includes a smart watch, a smart wristband, a smart earphone, a smart glasses, a smart accessory (a smart bracelet, a smart hand chain, a smart ring, a smart necklace, a smart ankle bracelet, a smart anklet, etc.), a smart band, a smart clothing, etc. It should be noted that the terminal 11 in the embodiment of the present application is not limited to a specific type. The network side equipment 12 may include an access network equipment or a core network equipment, where the access network equipment 12 may be called a radio access network equipment, a radio access network (Radio Access Network, RAN), a radio access network function or a radio access network unit.The access network equipment 12 may include a base station, a WLAN access point, or a WiFi node, and the base station may be called a Node B, an evolved Node B (eNB), an access point, a base transceiver station (BTS), a radio base station, a radio transceiver, a basic service set (BSS), an extended service set (ESS), a home B node, a home evolved B node, a transmitting receiving point (TRP), or any other suitable term in the art. As long as the same technical effect is achieved, the base station is not limited to a specific technical term, and it should be explained that in the embodiments of the present application, only base stations in an NR system are introduced as examples, and the specific type of the base station is not limited.The core network devices include a core network node, a core network function, a mobility management entity (MME), an access and mobility management function (AMF), a session management function (SMF), a user plane function (UPF), a policy control function (PCF), a policy and charging rules function (PCRF), an edge application server discovery function (EASDF), a unified data management (UDM), a unified data repository (UDR), a home subscriber server (HSS), a centralized network configuration (CNC), a network repository function (NRF), a network exposure function (NEF), a local NEF (Local NEF or L-NEF), a binding support function (BSF), an application function (Application Response Function), a network QoS function (QoS function ... It should be noted that the embodiments of the present application only take the core network device in the NR system as an example, and do not limit the specific type of the core network device.

[0028] Hereinafter, some concepts and / or terms related to the network element registration method, model determination method, device, network element, communication system, and storage medium according to the embodiments of the present application will be explained and described.

[0029] 1. Horizontal associative learning: Application scenario: The essence of horizontal federation learning is the federation of samples. It is applied to scenarios where the service types between participants are the same but the clients are different, i.e., there is a lot of overlap in features and little overlap in users, such as the same service (e.g., video services, voice services, and over-the-top (OTT) services of other Internet companies) for different users (e.g., each UE, i.e., different samples) in different cities in the 5th Generation (5G) communication system. By federating the same data features of different samples of participants, horizontal federation obtains a better model by increasing the number of training samples.

[0030] As shown in Figure 2, Server A is the coordinator, which sends information such as federated learning tasks and models to other clients (members), collects information such as models, parameters, gradients, and change rates fed back from clients, updates the initial model, and then distributes the updated model. Figure 2 shows the principle of horizontal federated learning.

[0031] Here, {1} is Sending encrypted gradients, {2}Secure aggregation, {3} Sending back model updates, {4} Updating models.

[0032] 2. Artificial Intelligence (AI) and AI models: Artificial intelligence is now widely applied in various fields. AI models have various algorithm implementation methods, such as neural networks, decision trees, support vector machines, Bayesian classifiers, etc. This application takes neural networks as an example for explanation, but does not limit the specific type of AI module.

[0033] As shown in Figure 3, it is a schematic diagram of a neural network. X1, X2...Xn are input values, Y is the output result, and each JPEG2025512876000002.jpg44 indicates that each neuron is a place where an operation is performed, and the result is passed on to the next layer. The input layer, hidden layer, and output layer, which are made up of a large number of neurons, are one neural network. The number of hidden layers and the number of neurons in each layer are the "network structure" of the neural network.

[0034] Here, the neural network is composed of neurons, and as shown in FIG. 4, is a schematic diagram of a neuron. Here, a1, a2, ... aK (i.e., X1 ... Xn above) are inputs, w is a weight (multiplication coefficient), b is a bias (addition coefficient), σ(z) is an activation function, and z is an output value. Common activation functions include Sigmoid, tanh, ReLU (Rectified Linear Unit), etc. The parameter information of each neuron combined with the algorithm used is the "parameter information" of the entire network, and is also an important part of the AI ​​model file.

[0035] In actual use, an AI model is a file containing elements such as network structure and parameter information. The trained AI model can be directly reused by the framework platform to directly perform intelligent functions such as judgment and identification without repeated construction or learning.

[0036] 3. Network Data Analytics Function (NWDAF) The network element can be decomposed into two parts: NWDAF (AnLF) and NWDAF (MTLF). The former is a network element responsible for the inference function (Analytics Logical Function, AnLF), and the latter is a network element responsible for the training function (Model Training Logical Function, MTLF).

[0037] In the following, the network element registration method and model determination method according to the embodiments of the present application will be described in detail with reference to several embodiments and application scenarios thereof in conjunction with the drawings.

[0038] In a 5G network, there are multiple NWDAF-MTLFs, such as NWDAF-MTLF(coordinator), NWDAF-MTLF1, and NWDAF-MTLF2, where: 1) NWDAF-MTLF (server A, coordinator) is responsible for data analysis for the entire province of Jiangsu, but does not have data for the entire province (because all data is distributed and stored in various cities in Jiangsu). 2) NWDAF-MTLF1 (Database B1, participant 1) can collect data in the Suzhou area and train the AI ​​model locally in Suzhou. 3) NWDAF-MTLF2 (Database B2, participant 2) can collect data in the Xuzhou area and train the AI ​​model locally in Suzhou.

[0039] Therefore, in order to obtain an AI model trained based on data from the entire Jiangsu province, it is necessary to use horizontal federated learning, which means that NWDAF-MTLF0 does not need to obtain data from the entire province.

[0040] Therefore, the flow and details of how to perform federated inference based on the model obtained by federated learning in a communication network are still unclear, such as which network elements participate in the federated inference, how information interactions are performed in the process, and what necessary information is interacted.

[0041] In the conventional network element registration and discovery technology, there is only a demand for the identifier, service and some capability information of the network element, but the registration and discovery about the federated learning capability is not yet within the standard scope.

[0042] In the embodiment of the present application, devices capable of participating in federated learning in different domains, such as an NWDAF network element, a base station, and a UE, need to initiate a registration process with a network element in charge of capability storage, such as an NRF, and instruct it to have capability information to participate in federated learning. When a federated learning initiator (e.g., an MTLF network element) initiates federated learning training, the federated learning initiator requests information from a network element, such as an NRF, to find an appropriate federated learning client (e.g., an NWDAF network element). After the federated learning training process is completed, the federated learning initiator MTLF network element may send a model and federated learning results to an AnLF network element, and at the same time, indicate participant information participating in federated learning.

[0043] Specifically, as one solution, a network element such as an NWDAF registers federated learning capability information with an NRF network element through a registration request, thereby realizing the registration of a network element such as an NWDAF with an NRF network element, and allowing other network elements to search for network elements capable of federated learning training by the NRF network element, thereby solving the problem of how a network element such as an NWDAF registers federated learning capability information with an NRF network element and can be searched by other network elements.

[0044] As another solution, the NRF network element receives a registration request sent by a network element such as the NWDAF, thereby registering the federated learning capability information of the network element such as the NWDAF to the NRF network element, thereby realizing the registration of the network element such as the NWDAF to the NRF network element, and allowing other network elements to search for network elements capable of federated learning training by the NRF network element, thereby solving the problem of how a network element such as the NWDAF can register its federated learning capability information to the NRF network element and be searched by other network elements.

[0045] In another method, the MTRL network element can search for a target network element capable of associative learning training in the NRF network element through a search request, and since the search request includes first information, the MTRL network element can search for a target network element matching the first information, and perform associative learning training with the searched target network element capable of associative learning training.

[0046] An embodiment of the present application provides a network element registration method, and FIG. 5 shows a flowchart of the network element registration method according to an embodiment of the present application. As shown in FIG. 5, the network element registration method according to an embodiment of the present application may include the following steps 201 and 202:

[0047] Step 201, a first network element sends a registration request to a second network element.

[0048] In an embodiment of the present application, the registration request is used to request the federated learning capability information of the first network element to be registered with the second network element.

[0049] Step 202, the second network element receives a registration request sent by the first network element.

[0050] In an embodiment of the present application, the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0051] Optionally, in the embodiments of the present application, the first network element may be an NWDAF network element or an NWDAF containing MTLF network element, and the second network element is a network element responsible for capability storage, which may be an NRF network element, a UDM network element, or a Data Collection Application Function (DCAF) network element.

[0052] Optionally, in an embodiment of the present application, the first network element may send a Nnrf_NFManagement_NFRegister Register request to the second network element to request the first network element to register with the second network element.

[0053] It should be explained that the type information of the federated learning training supported by the first network element is used to indicate the AI ​​model training algorithm type supported by the first network element.

[0054] Optionally, in an embodiment of the present application, the type information of federated learning training supported by the first network element includes at least one of indication information of whether the first network element supports federated learning training, a horizontal federated learning training type, a vertical federated learning training type, a federated learning server capability, and a federated learning client capability.

[0055] It should be noted that the federated learning server capability is used to indicate whether the first network element has federated learning server capability or whether the first network element supports being a federated learning server.

[0056] Optionally, in an embodiment of the present application, the first network element may be a federated learning server, which has the ability to aggregate local model training information provided by each federated learning client to generate a global model, or coordinate a federated learning process.

[0057] It should be noted that the federated learning client capability is used to indicate whether the first network element has federated learning client capability or whether the first network element supports being a federated learning client.

[0058] Optionally, in an embodiment of the present application, the first network element may be a federated learning client, which has the capability to participate in federated learning, train a local model, and provide local model training information.

[0059] It should be noted that the time information that the first network element supports the associative learning training is used to indicate the time to perform the associative learning training supported by the first network element, and performing the associative learning training within this time period, for example, from 10 p.m. to 6 a.m. the following morning, will result in relatively good performance.

[0060] It should be noted that the metadata information possessed by the first network element is used to indicate data information that the first network element can cover, obtain and provide.

[0061] Optionally, in an embodiment of the present application, the metadata information possessed by the first network element includes at least one of an input data type, an output data type, a data amount, and a data range.

[0062] It should be explained that the above input data type is the input data type of the associative learning training, and the above output data type is the output data type of the associative learning training.

[0063] It should be noted that the above input data type is used to indicate the data type that the first network element can collect or obtain to perform the model training process as input data.

[0064] Optionally, in the embodiment of the present application, the data type refers to the presence or absence of obvious characteristics in the data within the region, for example, the data is concentrated in the morning and scattered in the evening.

[0065] Optionally, in the embodiment of the present application, the data type may be a data type, such as location information of the UE, time information of the UE, load information of a network element, network status information, resource information of a network element, and so on.

[0066] Optionally, in an embodiment of the present application, the data scope is a service scope of the first network element.

[0067] Optionally, in an embodiment of the present application, the data range includes at least one of a service area of ​​the first network element, an area from which the first network element can collect data, and an object from which the first network element can collect data.

[0068] Optionally, in an embodiment of the present application, the collectable data of the first network element includes metadata information and training data for model training.

[0069] Optionally, in an embodiment of the present application, the area in which the first network element can collect data may be any one of the first network element service area, a sub-area range under the first network element service area, and a target of finer granularity of collectable data (e.g., a single UE list).

[0070] Optionally, in the embodiment of the present application, the objects from which the first network element can collect data may include one or more specific network elements, or one or more UEs.

[0071] Optionally, in an embodiment of the present application, the registration request further includes at least one of identifier information of the first network element, data analysis identifier information supported by the first network element, model filter information supported by the first network element, and network element type information of the first network element, where the network element type includes a federated learning server network element type and / or a federated learning client network element type.

[0072] Optionally, in an embodiment of the present application, the data analysis identifier information supported by the first network element may be user mobility trajectory (UE mobility) information.

[0073] Optionally, in an embodiment of the present application, the data analysis identifier information corresponds to type information of the associative learning training.

[0074] Optionally, in an embodiment of the present application, the data analysis identifier information corresponds to federated learning capability information of the first network element.

[0075] Optionally, in an embodiment of the present application, the federated learning capability information of the first network element corresponds to data analysis identifier information, for example, the federated learning capability information of the first network element may correspond to a certain data analysis identifier, i.e., the first network element has federated learning capability for the task of this data analysis identifier.

[0076] Optionally, in the embodiments of the present application, there is a mapping relationship between the data analysis identifier information and the association learning capability information of the first network element, and this mapping relationship may be one-to-one, many-to-one or one-to-many, where when the mapping relationship is one-to-one, for different data analysis identifier information, the association learning capability information of the first network element is correspondingly different.

[0077] For example, a first network element supports 10 data analysis identifier information, where data analysis identifiers 1 to 3 correspondingly support horizontal federated learning training, data analysis identifiers 4 to 6 correspondingly support vertical federated learning training, and data analysis identifiers 7 to 10 correspondingly do not support federated learning training.

[0078] Optionally, in the embodiment of the present application, the identifier information of the first network element may be one of: Fully Qualified Domain Name (FQDN) information and IP address information.

[0079] Optionally, in an embodiment of the present application, the entire qualified domain name information is used to indicate the location of the first network element and the connectivity of the first network element.

[0080] It should be noted that the model filter information may also be referred to as model coverage information, which is used to indicate the coverage of the model generated by the first network element training.

[0081] Optionally, in an embodiment of the present application, the model scope information includes the slice to which the model is applied, the external data network (Data Network Name, DNN) to which the model is applied, the region to which the model is applied, the time to which the model is applied, etc.

[0082] Optionally, in the embodiments of the present application, the time period to which the model is applied may be a large time range, for example within a year, or may be a periodic time period, for example from 9:00 to 10:00 every day.

[0083] Optionally, in an embodiment of the present application, the model filter information supported by the first network element includes model filter information corresponding to a model generated by the first network element through federated learning.

[0084] Optionally, in an embodiment of the present application, the federated learning capability information of the first network element is: Algorithm information underlying federated learning training of the first network element; model accuracy information achievable by the federated learning training of the first network element; federated learning training speed information of the first network element; Model description method information supported by a first network element; Model shareable information supported by a first network element; Manufacturer information of the first network element; and a number of federated learning models supported by the first network element.

[0085] It should be noted that the algorithm information underlying the federated learning training of the first network element is the specific algorithm used in the training model, such as linear regression, neural network.

[0086] It should be noted that the model accuracy information achievable by the federated learning training of the first network element is used to indicate the accuracy degree of the model output result.

[0087] It is to be explained that the speed information of the associative learning training of the first network element includes information of the length of time it takes for the first network element to reach a target accuracy of the model based on the associative learning training model.

[0088] It should be noted that the model description method information supported by the first network element is used to indicate that the first network element supports representing a model based on the model description method corresponding to the model description method information.

[0089] Optionally, in the embodiment of the present application, the above model description method information may be referred to as model description method request information or model description method expectation information.

[0090] Optionally, in the embodiments of the present application, the model description method information may be a model expression language, such as Open Neural Network Exchange (ONNX), or a model framework, such as TensorFlow, Pytorch, etc.

[0091] It should be explained that the model shareability information supported by the first network element is used to indicate that the first network element supports a sharing model, where shareable means interoperable, or shareable means mutually understandable, and shareable means operable.

[0092] Optionally, in the embodiment of the present application, the model sharable information may be referred to as model sharable request information or model sharable expectation information.

[0093] Specifically, when the model shareable information supported by a first network element includes one or more manufacturer information, other network elements belonging to this manufacturer information may interact with this first network element to obtain, for example, the model and model information of the first network element training.

[0094] Specifically, when the model shareable information supported by a first network element includes one or more model information, other network elements may interact with the model trained by the first network element corresponding to this model information, for example, to obtain the model and model information trained by the first network element.

[0095] It should be noted that the above model shareable information may be based on data analysis task granularity, i.e., different types of data analysis tasks have different shareable information (such as different manufacturer information, model information, etc.).

[0096] It is to be explained that the manufacturer information of the first network element is used to indicate a manufacturer corresponding to the first network element.

[0097] It should be noted that the manufacturer information of the first network element is subsequently used by the second network element to determine whether the model generated by the first network element based on the federated learning training can be shared with other network elements.

[0098] For example, if the manufacturer information of another network element is the same as the manufacturer information of the first network element, the other network element may share the model generated by the first network element based on the federated learning training.

[0099] Optionally, in the embodiment of the present application, the manufacturer information of the first network element may be used by the second network element to select network elements that can participate in the federated learning training.

[0100] It should be noted that the number information of federated learning models supported by the first network element is used to indicate the number information, e.g., maximum number and minimum number, of models based on federated learning training supported by each data analysis identifier, respectively, for the first network element.

[0101] Optionally, in an embodiment of the present application, the registration request further includes at least one of type information of the first network element and a service name supported by the first network element.

[0102] For example, the first network element is an NWDAF network element, and the service name supported by the NWDAF network element is Nnwdaf_AnalyticsInfo_Request.

[0103] Optionally, in an embodiment of the present application, the type information of the first network element is used to indicate what type of network element is currently registered, for example, if the type information of the first network element is an NWDAF network element type, it indicates that the currently registered network element is an NWDAF network element.

[0104] Optionally, the embodiment of the present application further includes, after the above step 202, the following steps 203 and 204.

[0105] Step 203, the second network element sends a registration response to the first network element.

[0106] In the embodiment of the present application, the registration response is used to indicate successful capability registration of the first device.

[0107] Optionally, in an embodiment of the present application, the second network element may send a registration response Nnrf_NFManagement_NFRegister response to the second network element to notify the first network element of successful registration.

[0108] Step 204, the first network element receives the registration response sent by the second network element.

[0109] An embodiment of the present application provides a network element registration method, in which a first network element can register the federated learning capability information of the first network element to a second network element through a registration request, thereby realizing the registration of the first network element to the second network element, so that other network elements can search for the first network element capable of federated learning training by the second network element, and solve the problem of how the first network element can register its federated learning capability information to the second network element and be searched by other network elements.

[0110] Optionally, in an embodiment of the present application, as shown in FIG. 6 in combination with FIG. 5, after the above step 202, the network element registration method according to the embodiment of the present application further includes the following steps 205 and 206.

[0111] Step 205: the third network element sends a search request to the second network element.

[0112] In the embodiment of the present application, the search request is used to request a search for network elements capable of federated learning training.

[0113] Step 206, the second network element receives the search request sent by the third network element.

[0114] In an embodiment of the present application, the search request includes first information, which includes at least one of data analysis identifier information corresponding to the target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, where the network element type request information is used to indicate a network element type corresponding to a network element capable of the federated learning training that needs to be searched, the network element type including a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used to indicate a request for the number of network elements capable of the federated learning training that needs to be searched.

[0115] In the embodiment of the present application, the third network element is a Model Training Logical Function (MTLF) network element, i.e., a model training network element, which may be understood as a network element that initiates federated learning.

[0116] Optionally, in the embodiment of the present application, the third network element may send a search request Nnrf_NFDiscovery_Request to the second network element to request to search for network elements capable of federated learning training.

[0117] Optionally, in an embodiment of the present application, when the third network element does not support / does not independently train and generate model information corresponding to the target task, the third network element sends a search request to the second network element.

[0118] Optionally, in the embodiments of the present application, if the third network element does not support / does not independently train and generate model information corresponding to the target task, it may be that the amount of data of the third network element is insufficient, or the accuracy rate of the training model of the third network element does not meet the standard.

[0119] Optionally, in an embodiment of the present application, the third network element may be a network element having federated learning server capability, and the third network element may search for a network element having federated learning client capability to perform federated learning training.

[0120] Optionally, in an embodiment of the present application, the third network element may be a network element having federated learning client capability, and the third network element may search for a network element having federated learning server capability and request the network element having federated learning server capability to organize a federated learning training.

[0121] It should be explained that the type information of the federated learning training corresponding to the target task is used to instruct the acquisition of network element information of a network element that supports the type of federated learning training corresponding to the target task, for example, when the third network element determines that federated learning training needs to be performed, it instructs the acquisition of network element information of a network element that supports federated learning training, and the network element information is used to indicate that the network element supports federated learning training, or when the third network element determines that horizontal federated learning training needs to be performed, it instructs the acquisition of network element information of a network element that supports horizontal federated learning training, and the network element information is used to indicate that the network element supports horizontal federated learning training.

[0122] It should be noted that the time information of the associative learning training corresponding to the target task is used to indicate the time information when the third network element plans to perform the associative learning training, and to search for network elements that support performing the associative learning training within this time range, where the time information is a future time period, for example, from 10 p.m. to 6 a.m. the next morning.

[0123] It should be noted that the metadata information possessed by the network elements of the federated learning training corresponding to the target task is used to indicate the data information that the demanding target network element can cover, obtain and provide.

[0124] Optionally, in an embodiment of the present application, the metadata information possessed by the network elements of the associative learning training corresponding to the target task includes at least one of an input data type, an output data type, a data amount, and a data range.

[0125] Optionally, in an embodiment of the present application, the first information further includes second information, which is used to request acquisition of network elements capable of federated learning training requiring multiple searches.

[0126] Optionally, in an embodiment of the present application, the network element number request information may include at least one of minimum number request information, proposed number information, and maximum number request information.

[0127] Optionally, in an embodiment of the present application, the first information is: Algorithm information that serves as the basis for training associative learning corresponding to the target task; Model accuracy information that can be achieved by associative learning training corresponding to the target task; and speed information of associative learning training corresponding to the target task; Model description method information supported by network elements of associative learning training corresponding to a target task; Model shareable information supported by network elements of federated learning training corresponding to the target task; Manufacturer information of the network elements for the associative learning training corresponding to the target task; and a number of associative learning models supported by the network elements of the associative learning training corresponding to the target task.

[0128] It should be noted that the algorithm information underlying the associative learning training corresponding to the target task is used to indicate the specific algorithm, such as linear regression, neural network, to be used in the training model.

[0129] It should be noted that the model accuracy information achievable by the associative learning training corresponding to the target task is used to indicate the degree of accuracy that the model output result needs to achieve.

[0130] It should be noted that the speed information of the associative learning training corresponding to the target task is used to indicate the time length information that the network element takes to reach the target accuracy of the model based on the associative learning training model.

[0131] It should be noted that the model description method information supported by the network element of the federated learning training corresponding to the above target task is used to indicate that the network element needs to support representing a model based on the model description method corresponding to this model description method information.

[0132] It should be explained that the model shareable information supported by the network element of the federated learning training corresponding to the above target task is used to indicate that the network element needs to support the sharing model.

[0133] It should be noted that the manufacturer information of the network element in the associative learning training corresponding to the target task is used to indicate the manufacturer corresponding to the network element.

[0134] It should be noted that the information on the number of federated learning models supported by the network elements of the federated learning training corresponding to the above target task is used to indicate the number of models based on federated learning training that need to be supported for each data analysis identifier and network element, respectively.

[0135] Optionally, in an embodiment of the present application, the first information further includes at least one of service information of a network element of the federated learning training corresponding to the target task, type information of a network element of the federated learning training corresponding to the target task, and area of ​​interest information.

[0136] Optionally, in the embodiment of the present application, when the target service is a request for data analysis identifier information, the service information of the network element of the federated learning training corresponding to the target task may be Nnwdaf_AnalyticsInfo_Request.

[0137] Optionally, in an embodiment of the present application, when the network element of the target service is an NWDAF network element, the type information of the network element of the federated learning training corresponding to the target task may be NWDAF network element type information.

[0138] Optionally, in an embodiment of the present application, the above-mentioned area of ​​interest information may be expressed as TA(s), cell(s) or other formats and is used to indicate a request for discovery of network elements within the area of ​​interest that will participate in the federated learning training.

[0139] Optionally, in an embodiment of the present application, as shown in FIG. 7 in conjunction with FIG. 6, after the above step 206, the network element registration method according to the embodiment of the present application further includes the following step 207:

[0140] Step 207: the second network element determines a target network element based on the search request.

[0141] In an embodiment of the present application, the associative learning capability information of the target network element matches the first information.

[0142] As can be seen, the second network element determines the network element whose associative learning capability information matches the first information as the target network element (one or more network elements).

[0143] Alternatively, in the embodiment of the present application, the above step 207 may be specifically realized by the following steps 207a to 207f.

[0144] In step 207a, the second network element determines a target network element based on the data analysis identifier information corresponding to the target task.

[0145] In an embodiment of the present application, the target network element supports federated learning training corresponding to data analysis identifier information.

[0146] As can be seen, the second network element determines the network element that supports the federated learning training corresponding to the data analysis identifier information as the target network element.

[0147] In step 207b, the second network element determines a target network element based on a type of associative learning training corresponding to the target task.

[0148] In an embodiment of the present application, the target network element supports a type of associative learning training corresponding to the target task.

[0149] As can be seen, the second network element determines as the target network element the network element that supports the type of associative learning training corresponding to the target task.

[0150] Optionally, in an embodiment of the present application, the second network element may determine a network element that supports federated learning training corresponding to the target task as a target network element.

[0151] Optionally, in an embodiment of the present application, the second network element may determine a network element that supports a federated learning server capability corresponding to the target task as the target network element.

[0152] Optionally, in an embodiment of the present application, the second network element may determine as the target network element a network element that supports a federated learning client capability corresponding to the target task.

[0153] Step 207c: the second network element determines a target network element based on the time information of the associative learning training corresponding to the target task.

[0154] In an embodiment of the present application, the target network element supports performing associative learning training at a time corresponding to the target time information, and the target time information is time information of the associative learning training corresponding to the target task.

[0155] As can be seen, the second network element determines, as the target network element, a network element that supports performing associative learning training at a time corresponding to the target time information.

[0156] In step 207d, the second network element determines a target network element based on the metadata information possessed by the network element of the federated learning training corresponding to the target task.

[0157] In an embodiment of the present application, the target network element supports metadata information possessed by the network element of the federated learning training corresponding to the target task.

[0158] As can be seen, the second network element determines a network element that supports associative learning training corresponding to the target task as a target network element.

[0159] Step 207e, the second network element determines a target network element based on the network element count request information.

[0160] In an embodiment of the present application, the number of the target network elements satisfies the network element number request information.

[0161] As can be understood, if the number of network elements capable of federated learning training searched by the second network element meets the number of network elements indicated by the network element number request information, the network element that meets the number request may be determined as the target network element.

[0162] Step 207f, the second network element determines a target network element based on the network element type request information.

[0163] In an embodiment of the present application, the network element type of the target network element satisfies the network element type request information.

[0164] As can be seen, the second network element determines the network element that satisfies the network element type request information as the target network element.

[0165] Optionally, in an embodiment of the present application, the second network element determines the target network element based on algorithm information underlying the associative learning training corresponding to the target task.

[0166] As can be seen, the second network element determines as the target network element the network element that supports the algorithm underlying the associative learning training corresponding to the target task.

[0167] Optionally, in an embodiment of the present application, the second network element determines the target network element based on model accuracy information achievable by the federated learning training corresponding to the target task.

[0168] As can be understood, the second network element determines the network element that can achieve the model accuracy of the associative learning training corresponding to the target task as the target network element.

[0169] Optionally, in an embodiment of the present application, the second network element determines the target network element based on speed information of the associative learning training corresponding to the target task.

[0170] As can be seen, the second network element determines the network element that supports the speed of associative learning training corresponding to the target task as the target network element.

[0171] Optionally, in an embodiment of the present application, the second network element determines the target network element based on model description method information supported by the network element of the federated learning training corresponding to the target task.

[0172] As can be understood, the second network element determines a network element that supports a model description method corresponding to the above model description method information (i.e., the model description method information supported by the network element of the associative learning training corresponding to the target task) as a target network element.

[0173] Optionally, in an embodiment of the present application, the second network element determines the target network element based on model shareable information supported by the network element of the federated learning training corresponding to the target task.

[0174] As can be seen, the second network element determines the network element that supports the sharing model as the target network element.

[0175] Optionally, in an embodiment of the present application, the second network element determines the target network element based on manufacturer information of the network element of the associative learning training corresponding to the target task.

[0176] As can be seen, the second network element determines the network element corresponding to the manufacturer information as the target network element.

[0177] Optionally, in an embodiment of the present application, the second network element determines the target network element based on the number information of the associative learning models supported by the network element of the associative learning training corresponding to the target task.

[0178] As can be understood, the second network element determines a network element that supports the above number information of the associative learning model (i.e., the number information of the associative learning model supported by the network element of the associative learning training corresponding to the target task) as a target network element.

[0179] Optionally, in the embodiment of the present application, when the second network element determines the target network element based on the network element count request information, the network element registration method according to the embodiment of the present application further includes the following step A1.

[0180] Step A1: When the number of network elements capable of associative learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the second network element determines the searched network element capable of associative learning training as a target network element.

[0181] As can be understood, the network elements capable of associative learning training searched above are network elements that meet other request information of the third network element (e.g., network element type, time of associative learning training, etc.) but do not meet the network element number request information.

[0182] Optionally, in an embodiment of the present application, when the number of network elements capable of federated learning training searched by the second network element is equal to the number of network elements indicated by the network element number request information, the second network element determines the searched network elements capable of federated learning training as target network elements.

[0183] As can be understood, the searched network elements capable of federated learning training are network elements that satisfy the request information of the third network element.

[0184] Optionally, in an embodiment of the present application, when the number of network elements capable of federated learning training searched by the second network element is greater than the number of network elements indicated by the network element number request information, the second network element determines some of the network elements capable of federated learning training searched as target network elements.

[0185] Optionally, in an embodiment of the present application, the second network element may sort all the searched network elements capable of associative learning training according to a preset internal logic, and select a network element that satisfies the network element number request information from among them as a target network element according to the sorting result.

[0186] Optionally, in an embodiment of the present application, as shown in FIG. 8 in combination with FIG. 7, after the above step 207, the network element registration method according to the embodiment of the present application further includes the following steps 208 and 209.

[0187] Step 208, the second network element sends the search response to the third network element.

[0188] In an embodiment of the present application, the search response includes identifier information or address information of the target network element.

[0189] Step 209, the third network element receives the search response sent by the second network element.

[0190] Optionally, in the embodiment of the present application, the identifier information of the target network element may be one of FQDN information and IP address information.

[0191] Optionally, in the embodiment of the present application, the second network element may send a Nnrf_NFDiscovery_Request search response to the third network element to send the network element information of the network elements capable of federated learning training.

[0192] Optionally, in an embodiment of the present application, the search response further includes second information, which includes at least one of type information of federated learning training supported by the target network element, time information when the target network element supports federated learning training, and metadata information possessed by the target network element.

[0193] In an embodiment of the present application, the type information of the federated learning training supported by the target network element is used to indicate the AI ​​model training algorithm type supported by the target network element.

[0194] In an embodiment of the present application, the time information in which the target network element supports federated learning training is used to indicate the time at which federated learning training supported by the target network element is performed, and performing federated learning training within this time period, for example, from 10 p.m. to 6 a.m. the following morning, can obtain relatively good performance.

[0195] In an embodiment of the present application, the metadata information possessed by the target network element is used to indicate the data information that the target network element can cover, obtain, and provide.

[0196] Optionally, in an embodiment of the present application, the metadata information possessed by the target network element includes at least one of an input data type, an output data type, a data amount, and a data range.

[0197] Optionally, in an embodiment of the present application, the search response further includes type information of the target network element.

[0198] Optionally, in an embodiment of the present application, the type information of the target network element is used to indicate what type of network element the target network element is, for example, if the type information of the target network element is an NWDAF network element type, it is used to indicate that the target network element is an NWDAF network element.

[0199] Optionally, in the embodiment of the present application, when the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the search response further includes third information, and the third information is: first indication information for indicating that the number of the target network elements is smaller than the number of the network elements indicated by the network element number request information; Information on the number of target network elements; and fourth information including time information for delaying the search for network elements capable of associative learning training, for proposing to the third network element to delay the search for network elements capable of associative learning training.

[0200] Optionally, in the embodiment of the present application, the time information may be a specific time.

[0201] For example, the fourth information may include time information "half an hour" so that when the third network element receives the third information, it can re-search for network elements capable of federated learning training half an hour later based on this time information.

[0202] As can be understood, if the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the second network element may determine the searched network elements capable of federated learning training as target network elements and carry identifier information or address information of the target network element in the search response, while the second network element may carry third information in the search response to inform the third network element at least one of: that the number of target network elements does not meet the number indicated by the network element number request information, the number information of the target network elements, and suggesting to the third network element that the search for network elements capable of federated learning training can be delayed.

[0203] Optionally, in an embodiment of the present application, when the number of network elements capable of federated learning training searched by the second network element is equal to the number of network elements indicated by the network element number request information, the search response may further include fifth information, which is used to indicate that the number of the target network elements satisfies the number of network elements indicated by the network element number request information.

[0204] As can be understood, if the number of network elements capable of associative learning training searched by the second network element is equal to the number of network elements indicated by the network element number request information, the second network element may determine the searched network elements capable of associative learning training as target network elements and carry identifier information or address information of the target network elements in the search response, while the second network element may carry fifth information in the search response to inform the third network element that the number of the target network elements meets the number indicated by the network element number request information.

[0205] Optionally, in the embodiment of the present application, when the number of network elements capable of federated learning training searched by the second network element is greater than the number of network elements indicated by the network element number request information, the search response may further include sixth information, and the sixth information may include: third indication information for indicating that the number of the searched network elements capable of associative learning training is greater than the number of the network elements indicated by the network element number request information; Related information of the retrieved network elements capable of federated learning training; and information on the number of network elements capable of the retrieved associative learning training.

[0206] As can be understood, if the number of network elements capable of federated learning training searched by the second network element is greater than the number of network elements indicated by the network element number request information, the second network element may select some network elements that satisfy the network element number request information from the network elements capable of federated learning training as target network elements and carry identifier information or address information of the target network element in the search response, while the second network element may carry sixth information in the search response to inform the third network element at least one of related information of all network elements capable of federated learning training searched by the second network element, number information of all network elements capable of federated learning training searched by the second network element, and that the number of network elements capable of federated learning training searched by the second network element exceeds the number indicated by the network element number request information.

[0207] Optionally, in an embodiment of the present application, after receiving the third information, the third network element may obtain model information corresponding to the target task by performing associative learning training with the target network element, or may stop performing associative learning training without performing associative learning training with the target network element determined by the second network element, and may delay re-initiating the search request based on the fourth information or internal logic to re-search for network elements capable of associative learning training and perform associative learning.

[0208] Optionally, in an embodiment of the present application, after receiving the sixth information, the third network element may select all or some of the network elements from the target network elements to perform federated learning training to obtain model information corresponding to the target task, and may select some of the network elements from all the network elements capable of federated learning training searched by the second network element based on internal logic to perform federated learning training to obtain model information corresponding to the target task.

[0209] Optionally, in an embodiment of the present application, after the above step 209, the network element registration method according to the embodiment of the present application further includes the following step 209a.

[0210] In step 209a, the third network element selects all or a part of the network elements from the target network elements to perform federated learning training.

[0211] Optionally, in the embodiments of the present application, all or some of the above network elements may be federated learning training clients or federated learning training servers, specifically, the third network element may select a federated learning training client or a federated learning training server from the target network element based on the network element type and the network element type request information.

[0212] Alternatively, in an embodiment of the present application, the third network element may select all or some of the network elements from the target network elements based on the federated learning capability of the target network element, or the registration information of the target network element, or the internal logic of the third network element, etc., to perform federated learning training.

[0213] An embodiment of the present application provides a network element registration method, in which a second network element can determine a target network element capable of federated learning training based on the search request by receiving a search request sent by a third network element, and send a search response to the third network element to allow the third network element to determine a target network element capable of federated learning training, thereby performing federated learning training with the target network element capable of federated learning training to obtain model information corresponding to a target task.

[0214] It should be mentioned that in the network element registration method according to the embodiment of the present application, the execution body may be a network element registration device, or a control module for executing the network element registration method in the network element registration device.

[0215] An embodiment of the present application provides a model determination method, and FIG. 9 shows a flowchart of the model determination method according to an embodiment of the present application. As shown in FIG. 9, the model determination method according to an embodiment of the present application may include the following steps 301 and 302:

[0216] Step 301: a third network element sends a search request to a second network element.

[0217] In the embodiment of the present application, the search request is used to request a search for network elements capable of federated learning training.

[0218] Step 302, a second network element receives a search request sent by a third network element.

[0219] Optionally, in an embodiment of the present application, when the third network element does not support / does not independently train and generate model information corresponding to the target task, the third network element sends a search request to the second network element.

[0220] Optionally, in the embodiments of the present application, if the third network element does not support / does not independently train and generate model information corresponding to the target task, it may be that the amount of data of the third network element is insufficient, or the accuracy rate of the training model of the third network element does not meet the standard.

[0221] In the embodiment of the present application, the third network element is a Model Training Logical Function (MTLF) network element, i.e., a model training network element, which may be understood as a network element that initiates federated learning.

[0222] Optionally, in the embodiment of the present application, the third network element may send a search request Nnrf_NFDiscovery_Request to the second network element to request to search for network elements capable of federated learning training.

[0223] In an embodiment of the present application, the search request includes first information, which includes at least one of data analysis identifier information corresponding to the target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, where the network element type request information is used to indicate a network element type corresponding to a network element capable of the federated learning training that needs to be searched for, the network element type including a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used to indicate a request for the number of network elements capable of the federated learning training that needs to be searched for.

[0224] Optionally, in an embodiment of the present application, the type information of the federated learning training corresponding to the target task is used to instruct the acquisition of network element information of a network element that supports the type of federated learning training corresponding to the target task, for example, when the third network element determines that it is necessary to perform federated learning training, it instructs the acquisition of network element information of a network element that supports federated learning training, and this network element information is used to indicate that the network element supports federated learning training, or when the third network element determines that it is necessary to perform horizontal federated learning training, it instructs the acquisition of network element information of a network element that supports horizontal federated learning training, and this network element information is used to indicate that the network element supports horizontal federated learning training.

[0225] Optionally, in an embodiment of the present application, the time information of the associative learning training corresponding to the target task is used to indicate time information for which the third network element plans to perform the associative learning training, and to search for network elements that support performing the associative learning training within this time range, where the time information is a future time period, for example, from 10 p.m. to 6 a.m. the next morning.

[0226] Optionally, in an embodiment of the present application, metadata information possessed by the network elements of the federated learning training corresponding to the target task is used to indicate the data information that the demanding target network element can cover, obtain, and provide.

[0227] Optionally, in an embodiment of the present application, the metadata information possessed by the network elements of the associative learning training corresponding to the target task includes at least one of an input data type, an output data type, a data amount, and a data range.

[0228] Optionally, in an embodiment of the present application, the input data type is an input data type of associative learning training, and the output data type is an output data type of associative learning training.

[0229] Optionally, in the embodiment of the present application, the data type refers to the presence or absence of obvious characteristics in the input data and output data of the associative learning training, for example, gathering in the morning and scattering in the evening.

[0230] Optionally, in the embodiment of the present application, the data scope is a service scope of the target network element.

[0231] Optionally, in an embodiment of the present application, the data range includes at least one of: a service area of ​​the target network element, an area from which the target network element can collect data, and an object from which the target network element can collect data.

[0232] Optionally, in an embodiment of the present application, the collectable data of the target network elements includes metadata information and training data for model training.

[0233] Optionally, in an embodiment of the present application, the area in which the target network element can collect data may be any one of the following: the target network element service area, a sub-area range under the target network element service area, and a target of finer collectable data granularity (e.g., a single UE list).

[0234] Optionally, in an embodiment of the present application, the first information is: Algorithm information that serves as the basis for training associative learning corresponding to the target task; Model accuracy information that can be achieved by associative learning training corresponding to the target task; and speed information of associative learning training corresponding to the target task; Model description method information supported by network elements of associative learning training corresponding to a target task; Model shareable information supported by network elements of federated learning training corresponding to the target task; Manufacturer information of the network elements for the associative learning training corresponding to the target task; and a number of associative learning models supported by the network elements of the associative learning training corresponding to the target task.

[0235] Optionally, in an embodiment of the present application, the algorithm information underlying the associative learning training corresponding to the target task is used to indicate the specific algorithm, e.g., linear regression, neural network, to be used in the training model.

[0236] Optionally, in an embodiment of the present application, model accuracy information achievable by the associative learning training corresponding to the target task is used to indicate the degree of accuracy that the model output result needs to achieve.

[0237] Optionally, in an embodiment of the present application, the speed information of the associative learning training corresponding to the target task is used to indicate the length of time information required for the network element to reach the target accuracy of the model based on the associative learning training model.

[0238] Optionally, in an embodiment of the present application, the first information further includes at least one of service information of a network element of the federated learning training corresponding to the target task, type information of a network element of the federated learning training corresponding to the target task, and area of ​​interest information.

[0239] Optionally, in the embodiment of the present application, when the target service is a request for data analysis identifier information, the service information of the network element of the federated learning training corresponding to the target task may be Nnwdaf_AnalyticsInfo_Request.

[0240] Optionally, in an embodiment of the present application, when the network element of the target service is an NWDAF network element, the type information of the network element of the federated learning training corresponding to the target task may be NWDAF network element type information.

[0241] Optionally, in an embodiment of the present application, the above-mentioned area of ​​interest information may be expressed as TA(s), cell(s) or other formats and is used to indicate a request for discovery of network elements within the area of ​​interest that will participate in the federated learning training.

[0242] An embodiment of the present application provides a model determination method, in which a third network element can search for a network element capable of associative learning training in a second network element through a search request, and the search request includes first information, so that the third network element can search for a target network element matching the first information, and perform associative learning training with the searched target network element capable of associative learning training.

[0243] Optionally, in an embodiment of the present application, as shown in FIG. 10 in combination with FIG. 9, after the above step 302, the model determination method according to the embodiment of the present application further includes the following steps 303 to 305.

[0244] Step 303, the second network element sends a search response to the third network element.

[0245] In an embodiment of the present application, the search response includes identifier information or address information of a target network element, where the target network element is a network element that supports federated learning training.

[0246] Optionally, in the embodiment of the present application, the identifier information of the target network element may be one of FQDN information and IP address information.

[0247] Step 304, the third network element receives the search response sent by the second network element.

[0248] Optionally, in the embodiment of the present application, the second network element may send a Nnrf_NFDiscovery_Request search response to the third network element to send the network element information of the network elements capable of federated learning training.

[0249] Optionally, in an embodiment of the present application, the search response further includes second information, which includes at least one of type information of federated learning training supported by the target network element, time information when the target network element supports federated learning training, and metadata information possessed by the target network element.

[0250] Optionally, in an embodiment of the present application, the type information of the federated learning training supported by the target network element is used to indicate the AI ​​model training algorithm type supported by the target network element.

[0251] Optionally, in an embodiment of the present application, the time information in which the target network element supports federated learning training is used to indicate the time to perform the federated learning training supported by the target network element, and performing the federated learning training within this time period, for example, from 10 p.m. to 6 a.m. the following morning, will result in relatively good performance.

[0252] Optionally, in an embodiment of the present application, metadata information possessed by the target network element is used to indicate data information that the target network element can cover, obtain, and provide.

[0253] Optionally, in an embodiment of the present application, the metadata information possessed by the target network element includes at least one of an input data type, an output data type, a data amount, and a data range.

[0254] Optionally, in an embodiment of the present application, the search response further includes type information of the target network element.

[0255] Optionally, in an embodiment of the present application, the type information of the target network element is used to indicate what type of network element the target network element is, for example, if the type information of the target network element is an NWDAF network element type, it indicates that the target network element is an NWDAF network element.

[0256] In step 305, the third network element performs associative learning training with the target network element to obtain model information corresponding to the target task.

[0257] Optionally, in an embodiment of the present application, the third network element may determine and search for target network elements to participate in the federated learning training based on information contained in the received search response, and perform federated learning with the searched target network elements.

[0258] Optionally, in an embodiment of the present application, as shown in FIG. 11 in combination with FIG. 10, before the above step 301, the model determination method according to the embodiment of the present application further includes the following steps 306 and 307.

[0259] Step 306: the fourth network element sends a model request to the third network element.

[0260] In an embodiment of the present application, the model request is used to request model information corresponding to a target task.

[0261] Step 307, the third network element receives the model request sent by the fourth network element.

[0262] Optionally, in the embodiments of the present application, the fourth network element is a model inference network element, which may be an AnLF network element, an NWDAF network element, or an NWDAF containing AnLF network element.

[0263] Optionally, in the embodiment of the present application, the fourth network element may send a model request Nnwdaf_MLMoldelInfo_Request to the third network element to request the third network element to feedback model information suitable for the task.

[0264] Optionally, in an embodiment of the present application, the model request includes at least one of: data analysis identifier information corresponding to the target task; and limitation information corresponding to the target task.

[0265] Optionally, in an embodiment of the present application, the data analysis identifier information corresponding to the target task may be user mobility trajectory (UE mobility) information.

[0266] Optionally, in an embodiment of the present application, the limitation information corresponding to the target task is used to indicate specific limiting conditions of the target task, such as information such as time, location, etc. required to limit the target task.

[0267] Optionally, in an embodiment of the present application, the model request further includes at least one of: indication information of a target network element, and report limiting information.

[0268] Optionally, in an embodiment of the present application, the target network element indication information is used to indicate the target network element of the target task, for example, the user mobility trajectory information may specify identity identifier information such as the SUPI of the target network element.

[0269] Optionally, in the embodiment of the present application, the report limiting information is used to indicate the information, format, etc. that needs to be reported to the target task, for example, the sorting manner is ascending order.

[0270] Optionally, in an embodiment of the present application, as shown in FIG. 12 in combination with FIG. 11, after the above step 305, the model determining method according to the embodiment of the present application further includes the following steps 308 and 309.

[0271] Step 308, the third network element sends the model response to the fourth network element.

[0272] In an embodiment of the present application, the model response includes model information corresponding to the target task, and the model information is used by the fourth network element to perform model inference.

[0273] Step 309, the fourth network element receives the model response sent by the third network element.

[0274] Optionally, in an embodiment of the present application, the model information corresponding to the target task includes at least one of target information and associative learning training related information.

[0275] Optionally, in an embodiment of the present application, the target information includes at least one of: model description information corresponding to the target task, a model file corresponding to the target task, and model storage address information corresponding to the target task.

[0276] Optionally, in an embodiment of the present application, the model file corresponding to the target task includes at least one of a complete network structure for generating model information corresponding to the target task, and parameter information for generating model information corresponding to the target task.

[0277] Optionally, in an embodiment of the present application, the above-mentioned federated learning training related information includes at least one of second instruction information, identifier information of the network elements participating in the federated learning training, and capability information of the network elements participating in the federated learning training.

[0278] In the embodiment of the present application, the second instruction information is used to indicate that the model information corresponding to the target task is model information obtained by associative learning training.

[0279] Optionally, in the embodiment of the present application, the identifier information of the network elements participating in the federated learning training may be the identity or address information of the network elements, such as full qualified domain name information, IP address information, and so on.

[0280] Optionally, in an embodiment of the present application, the third network element may send to the fourth network element identifier information of the network elements participating in the federated learning training, which may be subsequently used by the fourth network element to perform inference of the federated learning training.

[0281] Optionally, in an embodiment of the present application, the model information corresponding to the target task further includes at least one of: data analysis identifier information corresponding to the target task; and restriction information corresponding to the target task.

[0282] Optionally, in an embodiment of the present application, the limited information corresponding to the target task includes information of an area of ​​interest or information of a target terminal. The model determination method according to an embodiment of the present application may further include the following step 401:

[0283] Step 401: when a third network element cannot obtain training data of a target task corresponding to a region of interest or a target terminal, the third network element performs associative learning training for the target task.

[0284] It should be explained that the above-mentioned regions of interest are regions based on the information required / to be predicted by the model of associative learning training.

[0285] For example, a model predicts tomorrow's weather in a certain area.

[0286] It should be noted that the target terminal is a terminal based on information required / to be predicted for the model of federated learning training.

[0287] For example, the model predicts tomorrow's schedule for the target device.

[0288] Optionally, in an embodiment of the present application, the limited information corresponding to the target task includes information of an area of ​​interest or information of a target terminal. The model determination method according to an embodiment of the present application may further include the following step 402:

[0289] Step 402: The third network element determines metadata information held by the network element of the federated learning training corresponding to the target task according to the information of the region of interest or the information of the target terminal.

[0290] In an embodiment of the present application, the metadata information includes data ranges of network elements of federated learning training.

[0291] An embodiment of the present application provides a model determination method, in which a third network element can receive a model request sent by a fourth network element, the model request is used to request model information corresponding to a target task, and the third network element can search for a network element capable of associative learning training in the second network element through a search request, and the search request includes first information, so that the third network element can search for a target network element matching the first information. By performing associative learning training with the searched target network element, model information corresponding to the target task is obtained, thereby improving the success rate of model transmission.

[0292] It should be explained that in the model determining method according to the embodiment of the present application, the execution body may be a model determining device, or a control module for executing the model determining method in the model determining device.

[0293] The following describes in detail the interaction process of the network element registration and model determination method according to the embodiment of the present application through a specific embodiment.

[0294] As shown in FIG. 13, the method according to the embodiment of the present application includes the following steps 21 to 32.

[0295] Step 21: Smart devices in different domains (e.g., NWDAF network elements on the core network side, base stations, UEs, third-party application servers, etc.) send capability registration messages to capability storage devices such as NRFs (e.g., NRF network elements, UDM network elements, DCAF network elements, etc.) to perform capability registration.

[0296] In an embodiment of the present application, an NWDAF network element may be registered via (Nnrf_NFManagement_NFRegister Register).

[0297] Step 22, the NRF network element stores the information of this NWDAF network element.

[0298] Step 23, the NRF network element sends a registration response message.

[0299] In an embodiment of the present application, the NRF network element may acknowledge successful registration of the NWDAF network element via a (Nnrf_NFManagement_NFRegister response) response message.

[0300] These steps are similar to those of the conventional technology, with the following differences:

[0301] In step 21, when the NWDAF network element sends a capability registration message to the NRF network element, in addition to information such as its own identifier information and supported analytic ID, it also sends information such as “supported training type information”, “time to support federated learning”, and “metadata information”.

[0302] The information required for registration is as follows:

[0303] 1. NF type: network element type, which refers to the type of network element that is registered this time. In this case, NF type = NWDAF type, etc.

[0304] 2. NF instance ID, FQDN or IP address of NF, network element instance identifier information, which refers to the currently registered network element identifier information, such as its FQDN information (Fully Qualified Domain Name, a fully qualified domain name for indicating the location and connection of this network element) or IP address information (another identifier information).

[0305] 3. Names of supported NF services (if applicable): The names of services supported by the network element. For example, if it is an NWDAF network element, the service name is Nnwdaf_AnalyticsInfo_Request.

[0306] In addition, the network element registration information further includes at least one of the following:

[0307] 4. Supported training type information, which refers to the AI ​​model training algorithm type supported by this network element, such as “federated learning”, “deep learning”, etc.

[0308] 5. Time period supporting federated learning refers to the time period during which federated learning supported by this network element is performed. When federated learning is performed within this time period, for example, from 10 p.m. to 6 a.m. the following morning, relatively good performance can be obtained.

[0309] 6. Metadata information Meta data refers to the data information that the network element can cover, obtain and provide, including information such as data type, data characteristics, data amount, etc.

[0310] 7. Analytics ID: data analysis ID, such as “UE mobility” (user mobility trajectory).

[0311] 8. The algorithm used for model training, including the specific algorithm used for training the model, such as linear regression, neural network, etc.

[0312] 9. The model accuracy that can be achieved through model training is the degree of accuracy of the model output results.

[0313] 10. The speed of model training, which indicates the length of time it takes for model training to reach a particular accuracy.

[0314] Step 24: The task consumer sends a data analysis request message to the model inference network element AnLF (or NWDAF, NWDAF containing AnLF).

[0315] In an embodiment of the present application, a task consumer may send a data analysis request message via (Nnwdaf_AnalyticsInfo_Request or Nnwdaf_AnalyticsSubscription_Subscribe).

[0316] Step 25: After receiving the task request, the AnLF sends a model request to the MTLF, requesting the MTLF to feed back a model that fits the task.

[0317] In an embodiment of the present application, the AnLF may send a model request via (Nnwdaf_MLMoldelInfo_Request), which includes task description information, for example:

[0318] 1. analytic ID: data analysis ID, such as “UE mobility” (user mobility trajectory).

[0319] 2. filter info: This is task limiting information, which indicates the specific limiting conditions of this task, such as the time, location, etc. required to limit the task.

[0320] 3. Information such as target network element / UE, which indicates the target network element or UE of the task, for example, UE mobility may specify identity information such as the SUPI of the target UE.

[0321] 4. reporting info: This is the report-specific information, which specifies the information that needs to be reported to the task, the format, etc., for example, the sorting method is ascending order.

[0322] In step 26, the MTLF determines whether it can independently train and generate a model that matches the request based on the task description information in step 25, and if not, it can begin federated learning.

[0323] In the embodiments of the present application, cases where independent training generation is not possible may be cases where the number of datasets is insufficient, the differentiation of data features is insufficient, and so on, making it impossible to achieve the target accuracy rate.

[0324] Step 27: A network element device that initiates federated training, such as a model training network element MTLF, sends a request to a storage-capable network element, such as an NRF, to find a device, such as a network element, that can perform federated learning training.

[0325] In an embodiment of the present application, the model training network element MTLF may send a discovery request via (Nnrf_NFDiscovery_Request).

[0326] These steps are similar to those of the conventional technology, with the following differences:

[0327] The MTLF sends a request to the NRF to find network elements to participate in federated learning, which must include the following:

[0328] 1. target NF service Name(s): target network element service information. For example, if the target service is a request for data analysis information, the information may be Nnwdaf_AnalyticsInfo_Request.

[0329] 2. NF type of the target NF, the type of the target network element. For example, if the network element of the target service is an NWDAF, this information may be the NWDAF type.

[0330] 3. Area of ​​interest, which may be TA(s), cell(s) or other representations, is used to indicate a request to discover network elements that will participate in federated learning within the area of ​​interest.

[0331] When the MTLF sends a request to the NRF, it must further include “training type”, “training condition limitation information”, etc.

[0332] 4. Training type, which refers to a desire to obtain the training type supported by the network element equipment. For example, in this embodiment, if a network element desires to perform federated learning and the obtained network element information desires to support federated learning training, the information may be information indicating that the network element supports federated learning, such as "federated learning".

[0333] 5. Training condition limiting information refers to filtering information for network element devices, etc., and further narrows the scope of network element devices, etc. to select network elements that are more suitable as needed, such as 'associative learning time information', 'required data information', etc. Including at least one of the following:

[0334] This time period supports federated learning, and refers to the target time information for when federated learning is to be performed. Network element devices that can perform federated learning within this time range are selected as targets. This time information is a future time period, for example, "from 10 p.m. to 6 a.m. the following morning."

[0335] Required data information refers to the data information that the target network element desires to cover, obtain and provide, including information such as data type, data characteristics, data amount, etc.

[0336] In step 28, the NRF returns device information, such as network elements, that match the search information in step 27 to the MTLF.

[0337] In an embodiment of the present application, a capability storage network element such as an NRF may return network element equipment information that matches the training condition limiting information in step 26 to the federated learning initiator, and return a response message via (Nnrf_NFDiscovery_Request).

[0338] Here you need to include the following:

[0339] 1. Network element device identifier information, which indicates the target network element, the device's FQDN, IP address, etc.

[0340] 2. Type information of network element equipment, such as NWDAF.

[0341] 3. Capability information of network element equipment, including at least one of the following:

[0342] This refers to the training type supported by the network element device that wishes to acquire the training type, such as "federated learning."

[0343] This is the time period that supports federated learning, and refers to the time period during which federated learning supported by this network element is performed. If federated learning is performed within this time period, for example, from 10 p.m. to 6 a.m. the following morning, relatively good performance can be obtained.

[0344] Metadata information refers to the data information that the network element can cover, obtain, and provide, including information such as data type, data characteristics, and data amount.

[0345] In step 29, the MTLF determines and searches for devices such as network elements that will participate in the training based on the information returned in step 28, and the MTLF performs federated learning with the devices such as network elements that have been searched.

[0346] Step 30: After the MTLF training is completed, send a feedback model to the AnLF, which now contains descriptive information for the task.

[0347] 1. Analytic ID, task type instruction information.

[0348] 2. Task definition information, which further describes the task demand information, such as the limited time, location, etc.

[0349] 3. Models, etc.

[0350] The federated learning member information may further include identifier information and capability information regarding the federated learning member, including at least one of the following:

[0351] 4. Associative learning information, indicating that this training is completed using associative learning.

[0352] 5. Identification information of members participating in federated learning, which may be member identities, address information, such as FQDN or IP addresses, etc. The MTLF sends this information to the AnLF together with device information such as network elements participating in federated learning, which is used by the AnLF to subsequently perform federated learning inference.

[0353] 6. Capability information of members participating in federated learning, including time information supporting federated learning and data information of the device itself.

[0354] Step 31: After receiving information such as the feedback model, the AnLF performs model inference.

[0355] In an embodiment of the present application, the AnLF may perform cooperative inference based on device information, such as network elements participating in federated learning, from among the obtained federated learning information.

[0356] In step 32, the AnLF feeds back a task report to the consumer based on the inference result.

[0357] It should be noted that, for the relevant explanations and beneficial effects in the above steps 21 to 32, please refer to the descriptions in the above embodiments, and will not be further described here.

[0358] 14 shows a possible structural schematic diagram of a network element registration device according to an embodiment of the present application. As shown in FIG. 14, a network element registration device 80 may include: a sending module 81.

[0359] Here, the sending module 81 is used to send a registration request to the second network element, and the registration request is used to request the federated learning capability information of the first network element to be registered with the second network element, and the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0360] An embodiment of the present application provides a network element registration device, which can register the federated learning capability information of a first network element to a second network element through a registration request, and realize the registration of the first network element to the second network element, so that other network elements can search for the first network element capable of federated learning training by the second network element, and solve the problem of how the first network element can register its federated learning capability information to the second network element and be searched by other network elements.

[0361] In one possible implementation manner, the type information of federated learning training supported by the first network element includes indication information of whether the first network element supports federated learning training, and at least one of a horizontal federated learning training type, a vertical federated learning training type, a federated learning server capability, and a federated learning client capability.

[0362] In one possible implementation, the metadata information possessed by the first network element includes at least one of an input data type, an output data type, a data amount, and a data range.

[0363] In one possible implementation, the data range includes at least one of the service area of ​​the first network element, the area from which the first network element can collect data, and the object from which the first network element can collect data.

[0364] In one possible implementation manner, the registration request further includes at least one of identifier information of the first network element, data analysis identifier information supported by the first network element, model filter information supported by the first network element, and network element type information of the first network element, where the network element type includes a federated learning server network element type and / or a federated learning client network element type.

[0365] In one possible implementation, the data analysis identifier information corresponds to type information of the associative learning training.

[0366] In one possible implementation, the data analysis identifier information corresponds to the federated learning capability information of the first network element.

[0367] In one possible implementation manner, the model filter information includes model filter information corresponding to a model generated by the first network element through federated learning.

[0368] In one possible implementation manner, the federated learning capability information of the first network element further includes at least one of algorithm information underlying the federated learning training of the first network element, model accuracy information achievable by the federated learning training of the first network element, federated learning training speed information of the first network element, model description method information supported by the first network element, model shareability information supported by the first network element, manufacturer information of the first network element, and number information of federated learning models supported by the first network element.

[0369] In one possible implementation manner, the speed information of the associative learning training of the first network element includes information on the length of time it takes for the first network element to reach a target accuracy of the model based on the associative learning training model.

[0370] The network element registration device according to the embodiment of the present application can realize each process realized by the first network element in the above method embodiment and achieve the same technical effects, and will not be described further here in order to avoid repetition of description.

[0371] 15 shows a possible structural schematic diagram of a network element registration device according to an embodiment of the present application. As shown in FIG. 15, the network element registration device 90 may include: a receiving module 91.

[0372] Here, the receiving module 91 is used for receiving a registration request sent by a first network element, and the registration request is used for requesting that the federated learning capability information of the first network element be registered with the second network element, and the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

[0373] An embodiment of the present application provides a network element registration device, which can register federated learning capability information of the first network element to a second network element by receiving a registration request sent by a first network element, and realize the registration of the first network element to the second network element, so that other network elements can search for a first network element capable of federated learning training, and solve the problem of how the first network element can register federated learning capability information to a second network element and be searched by other network elements.

[0374] In one possible implementation, the federated learning capability information of the first network element is: The network element further includes at least one of algorithm information underlying the federated learning training of the first network element, model accuracy information achievable by the federated learning training of the first network element, federated learning training speed information of the first network element, model description method information supported by the first network element, model shareability information supported by the first network element, manufacturer information of the first network element, and number information of federated learning models supported by the first network element.

[0375] In one possible implementation manner, the receiving module 91 is further used for receiving a search request sent by a third network element, the search request is used to request a search for network elements capable of federated learning training, the search request includes first information, the first information including at least one of data analysis identifier information corresponding to the target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information is used to indicate a network element type corresponding to a network element capable of federated learning training that needs to be searched, the network element type includes a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used to indicate a request for the number of network elements capable of federated learning training that needs to be searched.

[0376] In one possible implementation manner, the first information further includes at least one of the following: algorithm information underlying the federated learning training corresponding to the target task; model accuracy information achievable by the federated learning training corresponding to the target task; speed information of the federated learning training corresponding to the target task; model description method information supported by network elements of the federated learning training corresponding to the target task; model shareability information supported by network elements of the federated learning training corresponding to the target task; manufacturer information of the network elements of the federated learning training corresponding to the target task; and information on the number of federated learning models supported by network elements of the federated learning training corresponding to the target task.

[0377] In one possible implementation manner, the network element registration device 90 according to the embodiment of the present application further includes a determining module, which is used after the receiving module 91 receives the search request sent by the third network element, to determine a target network element according to the search request, and the federated learning capability information of the target network element matches the first information.

[0378] In one possible implementation manner, the determination module specifically includes: determining a target network element based on data analysis identifier information corresponding to the target task, where the target network element supports federated learning training corresponding to the data analysis identifier information; determining a target network element based on a type of federated learning training corresponding to the target task, where the target network element supports the type of federated learning training corresponding to the target task; determining a target network element based on time information of the federated learning training corresponding to the target task, where the target network element supports performing federated learning training at a time corresponding to the target time information, the target time information being time information of the federated learning training corresponding to the target task; determining a target network element based on metadata information held by the network element of the federated learning training corresponding to the target task, where the target network element supports the metadata information held by the network element of the federated learning training corresponding to the target task; determining a target network element based on network element number request information, where the number of the target network elements meets the network element number request information; and determining a target network element based on network element type request information, where the network element type of the target network element meets the network element type request information.

[0379] In one possible implementation manner, when the second network element determines a target network element based on the network element number request information, the determination module is used to determine the searched network element capable of federated learning training as the target network element when the number of network elements searched by the second network element is smaller than the number of network elements indicated by the network element number request information.

[0380] In one possible implementation manner, the network element registration device 90 according to the embodiment of the present application further includes a sending module, which is used for sending a search response to the third network element after the determining module determines the target network element according to the search request, where the search response includes the identifier information or address information of the target network element.

[0381] In one possible implementation manner, the search response further includes second information, which includes at least one of type information of federated learning training supported by the target network element, time information when the target network element supports federated learning training, and metadata information possessed by the target network element.

[0382] In one possible implementation manner, when the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the search response further includes third information, which includes at least one of first instruction information for indicating that the number of target network elements is smaller than the number of network elements indicated by the network element number request information, number information of the target network elements, and fourth information including time information for delaying the search for network elements capable of federated learning training, for suggesting to the third network element to delay the search for network elements capable of federated learning training.

[0383] The network element registration device according to the embodiment of the present application can realize each process realized by the second network element in the above method embodiment and achieve the same technical effects, and will not be described further here in order to avoid repetition of description.

[0384] 16 shows a possible structural schematic diagram of a model determining apparatus according to an embodiment of the present application. As shown in FIG. 16, the model determining apparatus 100 may include: a sending module 101.

[0385] Here, the sending module 101 is used to send a search request to the second network element, and the search request is used to request a search for network elements capable of federated learning training, and the search request includes first information, and the first information includes at least one of data analysis identifier information corresponding to the target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, and the network element type request information is used to indicate a network element type corresponding to the network element capable of federated learning training that needs to be searched, and the network element type includes a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used to indicate a request for the number of network elements capable of federated learning training that needs to be searched.

[0386] An embodiment of the present application provides a model determination device, which can search for a network element capable of associative learning training in a second network element through a search request, and the search request includes first information, so that the third network element can search for a target network element matching the first information, and perform associative learning training with the searched target network element capable of associative learning training.

[0387] In one possible implementation manner, the metadata information possessed by the network elements of the associative learning training corresponding to the target task includes at least one of the following: input data type, output data type, data amount, and data range.

[0388] In one possible implementation manner, the first information further includes at least one of the following: algorithm information underlying the federated learning training corresponding to the target task; model accuracy information achievable by the federated learning training corresponding to the target task; speed information of the federated learning training corresponding to the target task; model description method information supported by network elements of the federated learning training corresponding to the target task; model shareability information supported by network elements of the federated learning training corresponding to the target task; manufacturer information of the network elements of the federated learning training corresponding to the target task; and information on the number of federated learning models supported by network elements of the federated learning training corresponding to the target task.

[0389] In one possible implementation manner, the model determining apparatus 100 according to the embodiment of the present application further includes a receiving module and an acquiring module. The receiving module is used to receive a search response sent by the second network element after the sending module 101 sends a search request to the second network element, the search response includes identifier information or address information of a target network element, the target network element is a network element that supports federated learning training, and the acquiring module is used to obtain model information corresponding to a target task by performing federated learning training with the target network element.

[0390] In one possible implementation manner, the search response further includes second information, which includes at least one of type information of federated learning training supported by the target network element, time information when the target network element supports federated learning training, and metadata information possessed by the target network element.

[0391] In one possible implementation manner, when the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the search response further includes third information, which includes at least one of first instruction information for indicating that the number of target network elements is smaller than the number of network elements indicated by the network element number request information, number information of the target network elements, and fourth information including time information for delaying the search for network elements capable of federated learning training, for suggesting to the third network element to delay the search for network elements capable of federated learning training.

[0392] In one possible implementation manner, the model determination device 100 according to an embodiment of the present application further includes a selection module, which is used to select all or some network elements from the target network elements to perform federated learning training after the receiving module receives the search response sent by the second network element.

[0393] In one possible implementation manner, the receiving module is further used for receiving a model request sent by a fourth network element before the sending module 101 sends a search request to the second network element, and the model request is used to request model information corresponding to the target task, and the model request includes at least one of data analysis identifier information corresponding to the target task and limitation information corresponding to the target task.

[0394] In one possible implementation manner, the sending module 101 is further used to send a model response to a fourth network element after the acquisition module performs associative learning training with the target network element to obtain model information corresponding to the target task, where the model response includes model information corresponding to the target task, and the model information is used by the fourth network element to perform model inference.

[0395] In one possible implementation manner, the model determining device 100 according to the embodiment of the present application further includes an execution module, wherein the limited information corresponding to the target task includes information of an interest area or information of a target terminal, and the execution module is used to perform federated learning training for the target task when the third network element cannot obtain the training data of the target task corresponding to the interest area or the target terminal.

[0396] In one possible implementation manner, the model determining device 100 according to the embodiment of the present application further includes a determining module, the limited information corresponding to the target task includes information of an interest area or information of a target terminal, and the determining module is used to determine metadata information of the network element of the federated learning training corresponding to the target task according to the information of the interest area or the information of the target terminal, and the metadata information includes a data range of the network element of the federated learning training.

[0397] In one possible implementation manner, the model information corresponding to the target task includes at least one of target information and associative learning training related information, and the target information includes at least one of model description information corresponding to the target task, a model file corresponding to the target task, and model storage address information corresponding to the target task.

[0398] In one possible implementation manner, the above-mentioned associative learning training related information includes at least one of second instruction information, identifier information of network elements participating in the associative learning training, and capability information of network elements participating in the associative learning training, and the second instruction information is used to indicate that the model information corresponding to the target task is model information obtained by associative learning training.

[0399] The model determination device according to the embodiment of the present application can realize each process realized by the third network element in the embodiment of the above method and achieve the same technical effect, and will not be described further here in order to avoid repetition of description.

[0400] Optionally, as shown in FIG. 17, an embodiment of the present application further provides a communication device 5000, which includes a processor 5001 and a memory 5002, and the memory 5002 stores a program or instruction that can run on the processor 5001. For example, when the communication device 5000 is a network element, when the program or instruction is executed by the processor 5001, it can realize each step of the method embodiment on the first network element side and achieve the same technical effect, or it can realize each step of the method embodiment on the second network element side and achieve the same technical effect, or it can realize each step of the method embodiment on the third network element side and achieve the same technical effect, and in order to avoid repetition, it will not be described further here.

[0401] An embodiment of the present application further provides a network element, including a processor and a communication interface, the communication interface being used to send a registration request to a second network element, the registration request being used to request the federated learning capability information of the first network element to be registered with the second network element, the federated learning capability information of the first network element including at least one of: type information of federated learning training supported by the first network element; time information of the first network element supporting federated learning training; and metadata information possessed by the first network element. This embodiment of the network element corresponds to the embodiment of the method on the first network element side, and each implementation process and realization manner of the embodiment of the method are all applicable to this embodiment of the network element, and can achieve the same technical effects.

[0402] An embodiment of the present application further provides a network element, including a processor and a communication interface, the communication interface being used to receive a registration request sent by a first network element, the registration request being used to request the federated learning capability information of the first network element to be registered with the second network element, the federated learning capability information of the first network element including at least one of the following: type information of federated learning training supported by the first network element, time information of the first network element supporting federated learning training, and metadata information possessed by the first network element. This embodiment of the network element corresponds to the embodiment of the method on the second network element side, and each implementation process and realization manner of the embodiment of the method are all applicable to this embodiment of the network element, and can achieve the same technical effects.

[0403] An embodiment of the present application further provides a network element, including a processor and a communication interface, the communication interface is used to send a search request to a second network element, the search request is used to request a search for a network element capable of federated learning training, the search request includes first information, the first information includes at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information held by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information is used to indicate a network element type corresponding to a network element capable of federated learning training that needs to be searched, the network element type includes a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used to indicate a request for the number of network elements capable of federated learning training that needs to be searched. This embodiment of the network element corresponds to the embodiment of the method on the third network element side, and each implementation process and realization manner of the embodiment of the method are all applicable to this embodiment of the network element, and can achieve the same technical effects.

[0404] Specifically, FIG. 18 is a hardware structural schematic diagram for implementing a network element of an embodiment of the present application, where the network element is a first network element, or the network element is a second network element, or the network element is a third network element.

[0405] 18, the network element 1200 includes a processor 1201, a network interface 1202, and a memory 1203. Here, the network interface 1202 is, for example, a common public wireless interface.

[0406] Specifically, the network element 1200 of the embodiment of the present application further includes instructions or programs stored in the memory 1203 and capable of running on the processor 1201, and the processor 1201 can call the instructions or programs in the memory 1203, execute the methods performed by each of the above modules, and achieve the same technical effects, which will not be described further here in order to avoid repetition.

[0407] The network element according to the embodiments of the present application can realize each process realized by the first network element, the second network element and the third network element in the embodiments of the above method, and can achieve the same technical effects, and will not be described further here in order to avoid repetition.

[0408] An embodiment of the present application further provides a readable storage medium, in which a program or instruction is stored, which, when executed by a processor, can realize each process of the above method embodiment and achieve the same technical effect, and will not be described further here in order to avoid repetition of description.

[0409] Here, the processor is the processor in the communication device according to the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0410] An embodiment of the present application further provides a chip, the chip including a processor and a communication interface, the communication interface is coupled to the processor, the processor is used to run a program or instruction and realize each process of the above method embodiment, and can achieve the same technical effect, and will not be described further here in order to avoid repetition of description.

[0411] It should be understood that the chips referred to in the embodiments of the present application may be referred to as system level chips, system chips, chip systems, or systems on chips.

[0412] The embodiments of the present application further provide a computer program / program product, which is stored in a storage medium, and which can be executed by at least one processor to realize each process of the above method embodiments and achieve the same technical effects, which will not be described further here in order to avoid repetition of description.

[0413] An embodiment of the present application further provides a communication system, which includes a first network element, a second network element, and a third network element, where the first network element may be used to perform steps of the network element registration method described above, the second network element may be used to perform steps of the network element registration method described above, and the third network element may be used to perform steps of the model determination method described above.

[0414] It should be explained that in this specification, the terms "comprise", "include", or any other variants thereof are intended to cover the non-exclusive "comprise", whereby a process, method, article, or apparatus that includes a set of elements includes not only those elements, but also other elements not expressly listed or inherent to such process, method, article, or apparatus. In the absence of further limitations, an element limited by the phrase "comprises a" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes this element. It should be pointed out that the scope of the method and apparatus in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may include performing functions in an essentially simultaneous manner or in reverse order based on the functions involved, for example, the described method can be performed in a different order than described, and various steps can be added, omitted, or combined. Also, features described with reference to some examples can be combined in other examples.

[0415] From the above description of the embodiments, it is clear to those skilled in the art that the methods of the above embodiments can be realized in the form of software and a necessary general-purpose hardware platform. Of course, they can also be realized in hardware, but in many cases the former is a more preferred embodiment. In light of this understanding, the technical proposal of the present application may be substantially or the part that contributes to the prior art may be embodied in the form of a computer software product, which is stored in a storage medium (e.g., ROM / RAM, magnetic disk, optical disk) and includes some instructions for causing a terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0416] The above describes the embodiments of the present application in conjunction with the drawings, but the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not limiting. Those skilled in the art can take the suggestions of this application and make many forms without departing from the spirit and scope of the claims of this application, all of which belong to the protection scope of this application.

[0417] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to Chinese Patent Application No. 202210317250.2 filed in China on March 28, 2022, and claims priority to Chinese Patent Application No. 202211468901.4 filed in China on November 22, 2022, the entire contents of which are incorporated herein by reference.

Claims

1. A network element registration method, comprising: A network element registration method, comprising: a first network element sending a registration request to a second network element, the registration request being used to request that federated learning capability information of the first network element be registered with the second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

2. 2. The method of claim 1, wherein the type information of federated learning training supported by the first network element includes at least one of indication information of whether the first network element supports federated learning training, a horizontal federated learning training type, a vertical federated learning training type, a federated learning server capability, and a federated learning client capability.

3. The method of claim 1 , wherein the metadata information possessed by the first network element includes at least one of an input data type, an output data type, a data amount, and a data range.

4. 4. The method of claim 3, wherein the data range includes at least one of a service area of ​​the first network element, an area for which the first network element can collect data, and an object for which the first network element can collect data.

5. The method according to any one of claims 1 to 4, wherein the registration request further includes at least one of identifier information of the first network element, data analysis identifier information supported by the first network element, model filter information supported by the first network element, and network element type information of the first network element, wherein the network element type includes a federated learning server network element type and / or a federated learning client network element type.

6. The method of claim 5 , wherein the data analysis identifier information corresponds to type information of the federated learning training.

7. The method of claim 5 , wherein the data analysis identifier information corresponds to federated learning capability information of the first network element.

8. The method of claim 5 , wherein the model filter information includes model filter information corresponding to a model generated by the first network element through federated learning.

9. The first network element's federated learning capability information is: Algorithm information underlying the federated learning training of the first network element; model accuracy information that can be achieved by the federated learning training of the first network element; Associative learning training speed information of the first network element; Model description method information supported by the first network element; Model shareable information supported by the first network element; Manufacturer information of the first network element; and a number of federated learning models supported by the first network element.

10. 10. The method of claim 9, wherein the speed information of the associative learning training of the first network element includes information about the length of time it takes for the first network element to reach a target accuracy of the model based on the model of associative learning training.

11. A network element registration method, comprising: A network element registration method, comprising: a second network element receiving a registration request sent by a first network element, the registration request being used to request that federated learning capability information of the first network element be registered with the second network element, the federated learning capability information of the first network element including at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

12. The first network element's federated learning capability information is: Algorithm information underlying the federated learning training of the first network element; model accuracy information that can be achieved by the federated learning training of the first network element; Associative learning training speed information of the first network element; Model description method information supported by the first network element; Model shareable information supported by the first network element; Manufacturer information of the first network element; and a number of federated learning models supported by the first network element.

13. After the second network element receives a registration request sent by the first network element, the method includes:

12. The method of claim 11, further comprising: the second network element receiving a search request sent by a third network element, the search request being used to request a search for network elements capable of federated learning training, the search request including first information, the first information including at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information being used to indicate a network element type corresponding to a network element capable of federated learning training that needs to be searched for, the network element type including a federated learning server network element type and / or a federated learning client network element type, and the network element number request information being used to indicate a request for the number of network elements capable of federated learning training that needs to be searched for.

14. The first information is Algorithm information underlying associative learning training corresponding to the target task; Model accuracy information that can be achieved by associative learning training corresponding to the target task; speed information of associative learning training corresponding to the target task; Model description method information supported by network elements of associative learning training corresponding to the target task; model shareable information supported by network elements of a federated learning training corresponding to the target task; Manufacturer information of network elements of the associative learning training corresponding to the target task; and a number of associative learning models supported by network elements of associative learning training corresponding to the target task.

15. After the second network element receives a search request sent by a third network element, the method includes: The method according to claim 13 or 14, further comprising: the second network element determining a target network element based on the search request, wherein federated learning capability information of the target network element matches the first information.

16. determining a target network element based on the search request by the second network element; The second network element determines the target network element based on data analysis identifier information corresponding to the target task, and the target network element supports federated learning training corresponding to the data analysis identifier information; The second network element determines the target network element based on a type of associative learning training corresponding to the target task, where the target network element supports the type of associative learning training corresponding to the target task; The second network element determines the target network element according to time information of the associative learning training corresponding to the target task, the target network element supports performing associative learning training at a time corresponding to the target time information, and the target time information is time information of the associative learning training corresponding to the target task; The second network element determines the target network element based on metadata information held by the network element of the federated learning training corresponding to the target task, and the target network element supports the metadata information held by the network element of the federated learning training corresponding to the target task; The second network element determines the target network element based on the network element number request information, where the number of the target network element satisfies the network element number request information; 16. The method of claim 15, further comprising at least one of: the second network element determining the target network element based on the network element type request information, wherein a network element type of the target network element satisfies the network element type request information.

17. When the second network element determines the target network element based on the network element count request information, the method includes:

17. The method of claim 16, further comprising: when the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the second network element determines the searched network element capable of federated learning training as the target network element.

18. After the second network element determines a target network element based on the search request, the method includes:

16. The method of claim 15, further comprising the second network element sending a search response to the third network element, the search response including identifier information or address information of the target network element.

19. The search response further includes second information, the second information being:

20. The method of claim 18, further comprising at least one of type information of federated learning training supported by the target network element, time information when the target network element supports federated learning training, and metadata information possessed by the target network element.

20. When the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the search response further includes third information, and the third information is: first indication information for indicating that the number of the target network elements is less than the number of network elements indicated by the network element number request information; The number information of the target network elements; and fourth information including time information for delaying a search for a network element capable of federated learning training, for proposing to the third network element to delay a search for a network element capable of federated learning training.

21. A model determination method, comprising: A model determination method comprising: a third network element sending a search request to a second network element, the search request being used to request a search for network elements capable of federated learning training, the search request including first information, the first information including at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information being used to indicate a network element type corresponding to a network element capable of federated learning training that needs to be searched for, the network element type including a federated learning server network element type and / or a federated learning client network element type, and the network element number request information being used to indicate a request for the number of network elements capable of federated learning training that needs to be searched for.

22. 22. The method of claim 21, wherein the metadata information possessed by the network elements of the associative learning training corresponding to the target task includes at least one of an input data type, an output data type, a data amount, and a data range.

23. The first information is Algorithm information underlying associative learning training corresponding to the target task; Model accuracy information that can be achieved by associative learning training corresponding to the target task; speed information of associative learning training corresponding to the target task; Model description method information supported by network elements of associative learning training corresponding to the target task; model shareable information supported by network elements of a federated learning training corresponding to the target task; Manufacturer information of network elements of the associative learning training corresponding to the target task; and a number of associative learning models supported by network elements of associative learning training corresponding to the target task.

24. After the third network element sends a search request to the second network element, the method includes: The method according to any one of claims 21 to 23, further comprising: the third network element receiving a search response sent by the second network element, the search response including identifier information or address information of a target network element, the target network element being a network element that supports federated learning training; and the third network element obtaining model information corresponding to the target task by performing federated learning training with the target network element.

25. The search response further includes second information, the second information being:

25. The method of claim 24, further comprising at least one of type information of federated learning training supported by the target network element, time information when the target network element supports federated learning training, and metadata information possessed by the target network element.

26. When the number of network elements capable of federated learning training searched by the second network element is smaller than the number of network elements indicated by the network element number request information, the search response further includes third information, and the third information is: first indication information for indicating that the number of the target network elements is less than the number of network elements indicated by the network element number request information; The number information of the target network elements; and fourth information including time information for delaying a search for a network element capable of federated learning training, for proposing to the third network element to delay a search for a network element capable of federated learning training.

27. After the third network element receives a search response sent by the second network element, the method includes: The method of claim 24, further comprising the third network element selecting all or a portion of the network elements from the target network elements to perform federated learning training.

28. Before the third network element sends a search request to the second network element, the method further comprises:

22. The method of claim 21, further comprising: the third network element receiving a model request sent by a fourth network element, the model request being used to request model information corresponding to the target task, the model request including at least one of data analysis identifier information corresponding to the target task and limitation information corresponding to the target task.

29. After the third network element and the target network element perform associative learning training to obtain model information corresponding to the target task, the method includes:

25. The method of claim 24, further comprising the third network element sending a model response to the fourth network element, the model response including model information corresponding to the target task, the model information being used by the fourth network element to perform model inference.

30. The limited information corresponding to the target task includes information of an area of ​​interest or information of a target terminal, and the method includes:

30. The method of claim 28, further comprising: when the third network element cannot obtain training data for the target task corresponding to the region of interest or the target terminal, the third network element performing federated learning training for the target task.

31. The limited information corresponding to the target task includes information of an area of ​​interest or information of a target terminal, and the method includes: The method of claim 28, further comprising: the third network element determining metadata information held by a network element of a federated learning training corresponding to the target task based on information of the area of ​​interest or information of the target terminal, the metadata information including a data range of the network element of the federated learning training.

32. 29. The method of claim 28, wherein the model information corresponding to the target task includes at least one of target information and associative learning training related information, and the target information includes at least one of model description information corresponding to the target task, a model file corresponding to the target task, and model storage address information corresponding to the target task.

33. 33. The method of claim 32, wherein the federated learning training related information includes at least one of second instruction information, identifier information of network elements participating in the federated learning training, and capability information of network elements participating in the federated learning training, and the second instruction information is used to indicate that the model information corresponding to the target task is model information obtained by federated learning training.

34. A network element registration device for use in a first network element, comprising: a sending module; A network element registration device, wherein the sending module is used for sending a registration request to a second network element, and the registration request is used for requesting that federated learning capability information of the first network element be registered with the second network element, and the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

35. A network element registration device for use in a second network element, comprising: a receiving module; A network element registration device, wherein the receiving module is used for receiving a registration request sent by a first network element, and the registration request is used for requesting that federated learning capability information of the first network element be registered with the second network element, and the federated learning capability information of the first network element includes at least one of type information of federated learning training supported by the first network element, time information when the first network element supports federated learning training, and metadata information possessed by the first network element.

36. A model determiner including a transmission module; A model determination device, wherein the sending module is used for sending a search request to a second network element, the search request is used for requesting a search of network elements capable of federated learning training, the search request includes first information, the first information including at least one of data analysis identifier information corresponding to a target task, type information of the federated learning training corresponding to the target task, time information of the federated learning training corresponding to the target task, metadata information possessed by the network element of the federated learning training corresponding to the target task, network element type request information, and network element number request information, the network element type request information is used for indicating a network element type corresponding to a network element capable of federated learning training that needs to be searched, the network element type includes a federated learning server network element type and / or a federated learning client network element type, and the network element number request information is used for indicating a request for the number of network elements capable of federated learning training that needs to be searched.

37. A network element comprising a processor and a memory, the memory storing a program or instructions operable on the processor, the program or instructions, when executed by the processor, implementing the steps of the network element registration method according to any one of claims 1 to 10.

38. A network element comprising a processor and a memory, the memory storing a program or instructions operable on the processor, the program or instructions, when executed by the processor, implementing the steps of the network element registration method of any one of claims 11 to 20.

39. A network element comprising a processor and a memory, the memory storing a program or instructions operable on the processor, the program or instructions, when executed by the processor, implementing the steps of the model determination method of any one of claims 21 to 33.

40. A communication system, comprising: a network element registration device according to claim 34; a network element registration device according to claim 35; and a model determination device according to claim 36; or The communication system includes a network element according to claims 37 to 39.

41. A readable storage medium, the readable storage medium storing a program or instructions, which, when executed by a processor, realizes steps of a network element registration method according to any one of claims 1 to 10, or realizes steps of a network element registration method according to any one of claims 11 to 20, or realizes steps of a model determination method according to any one of claims 21 to 33.

42. A chip comprising a processor and a communication interface, the communication interface being coupled to the processor, the processor running a program or instruction to implement a network element registration method according to any one of claims 1 to 10, or to implement a network element registration method according to any one of claims 11 to 20, or to implement a model determination method according to any one of claims 21 to 33.

43. A computer program product, the program product being executed by at least one processor to implement the network element registration method according to any one of claims 1 to 10, or to implement the network element registration method according to any one of claims 11 to 20, or to implement the model determination method according to any one of claims 21 to 33.

44. A network element, the network element being configured to perform a network element registration method according to any one of claims 1 to 10.

45. A network element, the network element being configured to perform a network element registration method according to any one of claims 11 to 20.

46. A network element, the network element being configured to perform a model determination method according to any one of claims 21 to 33.

Images