Network element registration method, model determination method, device, network element, communication system and 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, belonging to the field of communication technologies. The network element registration method in the embodiments of the present application includes: a first network element sends a registration request to a second network element, and the registration request is used to request to register the federated learning capability information of the first network element to the second network element. The federated learning capability information of the first network element includes 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.

Description

Technical Field

[0001] This application belongs to the field of communication technology, specifically relating 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. Federated learning allows multiple clients to collaborate under the coordination of a central server to obtain a complete machine learning model. However, current technologies do not provide additional enhancements or differentiated treatment for this type of artificial intelligence algorithm. For example, by default, any participant capable of training, such as network elements, base stations, and user equipment (UE), can participate in federated learning. This unfiltered selection of participants can lead to problems such as lost participants and reduced learning efficiency due to various reasons (e.g., poor network signal at the UE). Therefore, enhancing the network element registration process in federated learning is an urgent problem to be solved. Summary of the Invention

[0003] This application provides 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 process in federated learning.

[0004] Firstly, a network element registration method is provided, the method comprising: a first network element sending a registration request to a second network element, the registration request being used to request the registration of the federated learning capability information of the first network element to 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.

[0005] Secondly, a network element registration device is provided, applied to a first network element. The network element registration device includes a sending module. The sending module is used to send a registration request to a second network element. The registration request requests the registration of the federated learning capability information of the first network element to the second network element. The federated learning capability information of the first network element includes at least one of the following: information on the type of federated learning training supported by the first network element, information on the time period during which the first network element supports federated learning training, and metadata information possessed by the first network element.

[0006] Thirdly, a network element registration method is provided, the method comprising: a second network element receiving a registration request sent by a first network element, the registration request being used to request the registration of the federated learning capability information of the first network element to 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 federated learning training supported by the first network element, and metadata information possessed by the first network element.

[0007] Fourthly, a network element registration device is provided, applied to a second network element. The network element registration device includes a receiving module. The receiving module is used to receive a registration request sent by a first network element. The registration request requests the registration of the federated learning capability information of the first network element to the second network element. The federated learning capability information of the first network element includes at least one of the following: information on the type of federated learning training supported by the first network element, information on the time period during which the first network element supports federated learning training, and metadata information possessed by the first network element.

[0008] Fifthly, a model determination method is provided, comprising: a third network element sending a search request to a second network element, the search request being used to request the search for network elements capable of federated learning training, the search request including first information, the first information including at least one of the following: data analysis identification information corresponding to the target task, type information of federated learning training corresponding to the target task, time information of federated learning training corresponding to the target task, metadata information of the network elements of federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information, the network element type requirement information being used to indicate the network element type corresponding to the network element capable of federated learning training to be searched, the network element type including federated learning server network element type and / or federated learning client network element type, the network element quantity requirement information being used to indicate the quantity requirement of the network elements capable of federated learning training to be searched, and the data volume requirement information being used to indicate the data volume requirement of the network elements capable of federated learning training to be searched.

[0009] Sixthly, a model determination apparatus is provided, comprising: a sending module. The sending module is configured to send a search request to a second network element. The search request requests the search for network elements capable of performing federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification 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 of the network element performing the federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information indicates the network element type corresponding to the network element to be searched that can perform federated learning training, and the network element type includes a federated learning server network element type and / or a federated learning client network element type. The network element quantity requirement information indicates the quantity requirement of the network element to be searched that can perform federated learning training, and the data volume requirement information indicates the required amount of data possessed by the network element to be searched that can perform federated learning training.

[0010] In a seventh aspect, a network element is provided, the network element including a processor and a memory, the memory storing a program or instructions executable on the processor, the program or instructions, when executed by the processor, implementing the steps of the network element registration method as described in the first aspect.

[0011] Eighthly, a network element is provided, including a processor and a communication interface, wherein the communication interface is used to send a registration request to a second network element, the registration request being used to request the registration of the federated learning capability information of a first network element to 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 federated learning training supported by the first network element, and metadata information possessed by the first network element.

[0012] In a ninth aspect, a network element is provided, the network element including a processor and a memory, the memory storing a program or instructions executable on the processor, the program or instructions, when executed by the processor, implementing the steps of the network element registration method as described in the third aspect.

[0013] In a tenth aspect, a network element is provided, including a processor and a communication interface, wherein the communication interface is used to receive a registration request sent by a first network element, the registration request being used to request the registration of the federated learning capability information of the first network element to a 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.

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

[0015] In a twelfth aspect, a network element is provided, including a processor and a communication interface, wherein the communication interface is used to send a search request to a second network element. The search request is used to search for a network element capable of performing federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification 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 of the network element for the federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element capable of performing federated learning training to be searched. The network element type includes a federated learning server network element type and / or a federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element capable of performing federated learning training to be searched. The data volume requirement information is used to indicate the data volume requirement of the network element capable of performing federated learning training to be searched.

[0016] In a thirteenth aspect, a communication system is provided, comprising: a network element registration device as described in the second aspect, a network element registration device as described in the fourth aspect, and a model determination device as described in the sixth aspect; or, comprising: a network element as described in the seventh, ninth, and eleventh aspects; or, comprising: a network element as described in the eighth, tenth, and twelfth aspects. The network element may be used to perform the steps of the methods described in the first, third, and fifth aspects.

[0017] In a fourteenth aspect, a readable storage medium is provided, on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect, or the steps of the method described in the third aspect, or the steps of the method described in the fifth aspect.

[0018] In a fifteenth aspect, a chip is provided, the chip including a processor and a communication interface coupled to the processor, the processor being configured to run a program or instructions to implement the method as described in the first aspect, or the method as described in the third aspect, or the method as described in the fifth aspect.

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

[0020] In this embodiment of the application, the first network element can register its federated learning capability information to the second network element through a registration request, thereby enabling other network elements to find the first network element that can perform federated learning training through the second network element. This solves the problem of how the first network element can register its federated learning capability information to the second network element and be found by other network elements.

[0021] The second network element can receive a registration request from the first network element and register the first network element's federated learning capability information into the second network element. This allows other network elements to find the first network element that can perform federated learning training, thus solving the problem of how the first network element can register its federated learning capability information into the second network element and be found by other network elements.

[0022] The third network element can use a lookup request to search for network elements that can be trained through federated learning from the second network element. Since the lookup request includes the first information, the third network element can find target network elements that match the first information. Then, it can perform federated learning training with the found target network elements that can be trained through federated learning. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the architecture of a wireless communication system provided in an embodiment of this application;

[0024] Figure 2 This is a schematic diagram illustrating the principle of horizontal federated learning provided in an embodiment of this application;

[0025] Figure 3 This is a schematic diagram of a neural network provided in an embodiment of this application;

[0026] Figure 4 This is a schematic diagram of a neuron provided in an embodiment of this application;

[0027] Figure 5 This is one of the flowcharts of a network element registration method provided in the embodiments of this application;

[0028] Figure 6 This is a second flowchart of a network element registration method provided in the embodiments of this application;

[0029] Figure 7 This is the third flowchart of a network element registration method provided in the embodiments of this application;

[0030] Figure 8 This is the fourth flowchart of a network element registration method provided in the embodiments of this application;

[0031] Figure 9 This is one of the flowcharts of a model determination method provided in the embodiments of this application;

[0032] Figure 10 This is a second flowchart of a model determination method provided in an embodiment of this application;

[0033] Figure 11 This is the third flowchart of a model determination method provided in the embodiments of this application;

[0034] Figure 12 This is the fourth flowchart of a model determination method provided in the embodiments of this application;

[0035] Figure 13 This is a flowchart of a network element registration method and a model determination method provided in an embodiment of this application;

[0036] Figure 14 This is one of the structural schematic diagrams of a network element registration device provided in the embodiments of this application;

[0037] Figure 15 This is a second schematic diagram of the structure of a network element registration device provided in the embodiments of this application;

[0038] Figure 16 This is a schematic diagram of the structure of a model determining device provided in an embodiment of this application;

[0039] Figure 17 This is a schematic diagram of the hardware structure of a communication device provided in an embodiment of this application;

[0040] Figure 18 This is a schematic diagram of the hardware structure of a network element provided in an embodiment of this application. Detailed Implementation

[0041] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0042] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0043] It is worth noting that the technologies described in this application are not limited to Long Term Evolution (LTE) / LTE-Advanced (LTE-A) systems, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency Division Multiple Access (SC-FDMA), and other systems. The terms "system" and "network" in this application are often used interchangeably, and the described technologies can be used with the systems and radio technologies mentioned above, as well as with other systems and radio technologies. The following description describes New Radio (NR) systems for illustrative purposes, and the term NR is used in most of the following description; however, these technologies can also be applied to applications other than NR systems, such as 6th generation (6G) radio systems. th Generation 6G communication system.

[0044] Figure 1This diagram illustrates a block diagram of a wireless communication system applicable to embodiments of this application. The wireless communication system includes a terminal 11 and a network-side device 12. Terminal 11 can be a mobile phone, tablet computer, laptop computer, personal digital assistant (PDA), handheld computer, netbook, ultra-mobile personal computer (UMPC), mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, vehicle-mounted device (VUE), pedestrian terminal (PUE), smart home (home devices with wireless communication capabilities, such as refrigerators, televisions, washing machines, or furniture), game console, personal computer (PC), ATM, or self-service machine, etc. Wearable devices include: smartwatches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart chains, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. It should be noted that the specific type of terminal 11 is not limited in this embodiment. Network-side equipment 12 may include access network equipment or core network equipment. Access network equipment 12 may also be referred to as radio access network equipment, radio access network (RAN), radio access network function, or radio access network unit. Access network equipment 12 may include base stations, WLAN access points, or WiFi nodes, etc. Base stations may be referred to as Node B, evolved Node B (eNB), access point, base transceiver station (BTS), radio base station, radio transceiver, basic service set (BSS), extended service set (ESS), home B node, home evolved B node, transmitting and receiving point (TRP), or any other suitable term in the field, as long as the same technical effect is achieved. The base station is not limited to specific technical terms. It should be noted that in this application embodiment, only a base station in an NR system is used as an example for description, and the specific type of base station is not limited.Core network equipment may include, but is not limited to, at least one of the following: core network node, core network function, Mobility Management Entity (MME), Access and Mobility Management Function (AMF), Session Management Function (SMF), User Plane Function (UPF), Policy Control Function (PCF), Policy and Charging Rules Function (PCRF), Edge Application Server Discovery Function (EASDF), Unified Data Management (UDM), Unified Data Repository (UDR), Home Subscriber Server (HSS), Centralized network configuration (CNC), Network Repository Function (NRF), Network Exposure Function (NEF), Local NEF (or L-NEF), Binding Support Function (BSF), and Application Function. Function (AF), etc. It should be noted that the embodiments of this application only use the core network equipment in the NR system as an example for introduction, and do not limit the specific type of core network equipment.

[0045] The following explains some concepts and / or terms involved in the network element registration method, model determination method, apparatus, network element, communication system, and storage medium provided in the embodiments of this application.

[0046] 1. Horizontal Federated Learning:

[0047] Applicable Scenarios: The essence of horizontal federated learning is the jointing of samples. It is suitable for scenarios where participants have the same business model but different customers, i.e., when there is a lot of feature overlap but little user overlap, such as the 5th generation (5... thIn Generation 5G communication systems, different users (e.g., each UE, i.e., different samples) in different cities can use the same service (e.g., video service, voice service, and over-the-top (OTT) services from other internet companies). By combining the same data features of different samples from the participating parties, horizontal federation increases the number of training samples, thus obtaining a better model.

[0048] like Figure 2 As shown, Server A is the coordinator, responsible for sending information such as the federated learning task and model to other clients and members, collecting information such as the model, parameters, gradients, and rate of change fed back by clients, updating the initial model, and then distributing the updated model. Figure 2 This is a schematic diagram of the principle of horizontal federated learning.

[0049] Wherein, ① represents sending encrypted gradients;

[0050] ② This is a secure aggregation;

[0051] ③ This refers to sending back model updates;

[0052] ④ Updating models.

[0053] 2. Artificial Intelligence (AI) and AI Models:

[0054] Artificial intelligence has been widely applied in various fields. AI models can be implemented using various algorithms, such as neural networks, decision trees, support vector machines, and Bayesian classifiers. This application uses neural networks as an example for illustration, but does not limit the specific type of AI module.

[0055] like Figure 3 The diagram shows a schematic of a neural network. X1, X2…Xn, etc., are input values, Y is the output result, and each circle represents a neuron, which is also where the computation is performed. The result is then passed to the next layer. These numerous neurons, forming an input layer, hidden layers, and an output layer, constitute a neural network. The number of hidden layers and the number of neurons in each layer define the "network structure" of the neural network.

[0056] Neural networks are composed of neurons, such as... Figure 4The diagram shows a schematic of a neuron. Here, a1, a2, ..., aK (i.e., X1...Xn from above) are the inputs, w is the weight (multiplicative coefficient), b is the bias (additive coefficient), σ(z) is the activation function, and z is the output value. Common activation functions include Sigmoid, tanh, and ReLU (Rectified Linear Unit). The parameter information of each neuron, combined with the algorithm used, constitutes the "parameter information" of the entire network, which is a crucial part of the AI ​​model file.

[0057] In practical use, an AI model refers to a file containing elements such as network structure and parameter information. The trained AI model can be directly reused by its framework platform without repeated construction or learning, and can directly perform intelligent functions such as judgment and recognition.

[0058] 3. The Network Data Analytics Function (NWDAF) element can be decomposed into two parts: NWDAF (AnLF) and NWDAF (MTLF). The former is the network element responsible for the inference function (Analytics Logical Function, AnLF), and the latter is the network element responsible for the training function (Model Training Logical Function, MTLF).

[0059] The network element registration method and model determination method provided in this application will be described in detail below with reference to the accompanying drawings and through some embodiments and application scenarios.

[0060] There are multiple NWDAF-MTLFs in a 5G network, such as NWDAF-MTLF (coordinator), NWDAF-MTLF1, and NWDAF-MTLF2, among which:

[0061] 1) NWDAF-MTLF (server A, coordinator) is responsible for data analysis of the entire Jiangsu Province, but it cannot have the data for the entire Jiangsu Province (because the data is distributed and stored in various cities in Jiangsu Province);

[0062] 2) NWDAF-MTLF1 (Database B1, participant 1) can collect data in the Suzhou area and train the AI ​​model locally in Suzhou;

[0063] 3) NWDAF-MTLF2 (Database B2, participant 2) can collect data in Xuzhou and train AI models in Suzhou.

[0064] Therefore, in order to obtain an AI model trained on data from the entire Jiangsu province, it is necessary to use lateral federated learning, which means that NWDAF-MTLF0 is not required to obtain data from the entire province.

[0065] Therefore, the process and details of how to perform federated inference based on the model obtained from federated learning in a communication network are still unclear. For example, it is unclear which network elements participate in federated inference, how they interact with each other during the process, and what necessary information they interact with.

[0066] Existing network element registration and discovery technologies only require information on network element identification, services, and some capabilities, but the registration and discovery of its federated learning capabilities are not currently within the scope of the standard.

[0067] In this embodiment, devices in different domains such as NWDAF network elements, base stations, and UEs that can participate in federated learning need to initiate a registration process with network elements responsible for capability storage, such as NRF, indicating that they possess the capability information to participate in federated learning. When a federated learning initiator (such as an MTLF network element) initiates federated learning training, it requests information from network elements such as NRF to find a suitable federated learning client (such as an NWDAF network element). After the federated learning training process is completed, the federated learning initiator MTLF network element sends the model and federated learning results to the AnLF network element, and can also indicate the information of the participants in the federated learning.

[0068] Specifically, one approach is that network elements such as NWDAF can register their federated learning capability information with NRF network elements through a registration request. This allows other network elements to find network elements capable of federated learning training through the NRF network element, thus solving the problem of how network elements such as NWDAF can register their federated learning capability information with the NRF network element and be found by other network elements.

[0069] Another approach is that NRF network elements can receive registration requests from network elements such as NWDAF and register the federated learning capability information of NWDAF and other network elements into the NRF network element. This allows other network elements to find network elements capable of federated learning training through the NRF network element, thus solving the problem of how network elements such as NWDAF can register their federated learning capability information into the NRF network element and be found by other network elements.

[0070] Another approach involves the MTLF network element using a lookup request to search for target network elements suitable for federated learning training from the NRF network element. Since the lookup request includes first information, the MTLF network element can find target network elements that match the first information. It can then perform federated learning training with the found target network elements that are suitable for federated learning training.

[0071] This application provides a method for network element registration. Figure 5 A flowchart of a network element registration method provided in an embodiment of this application is shown. Figure 5 As shown, the network element registration method provided in this application embodiment may include the following steps 201 to 202.

[0072] Step 201: The first network element sends a registration request to the second network element.

[0073] In this embodiment of the application, the above registration request is used to request the registration of the federated learning capability information of the first network element to the second network element.

[0074] Step 202: The second network element receives the registration request sent by the first network element.

[0075] In this embodiment of the application, the federated learning capability information of the first network element includes at least one of the following: the type information of the federated learning training supported by the first network element, the time information of the federated learning training supported by the first network element, and the metadata information of the first network element.

[0076] Optionally, in the embodiments of this application, the first network element can 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 can be an NRF network element, a UDM network element, or a Data Collection Application Function (DCAF) network element.

[0077] Optionally, in this embodiment of the application, the first network element may send an Nnrf_NFManagement_NFRegister Register request to the second network element to request that the first network element be registered in the second network element.

[0078] It should be noted that the type information of federated learning training supported by the first network element is used to indicate the type of AI model training algorithm supported by the first network element.

[0079] Optionally, in this embodiment of the application, the type information of federated learning training supported by the first network element includes at least one of the following: indication information on whether the first network element supports federated learning training, horizontal federated learning training type, vertical federated learning training type, federated learning server capability, and federated learning client capability.

[0080] It should be noted that the aforementioned federated learning server capability is used to indicate whether the first network element has federated learning server capability, or to indicate whether the first network element supports acting as a federated learning server.

[0081] Optionally, in this embodiment of the application, the first network element can serve as a federated learning server, which has the ability to aggregate local model training information provided by various federated learning clients to generate a global model, or to coordinate the federated learning process.

[0082] It should be noted that the aforementioned federated learning client capability is used to indicate whether the first network element has federated learning client capability, or to indicate whether the first network element supports acting as a federated learning client.

[0083] Optionally, in this embodiment of the application, the first network element can serve as a federated learning client, which has the ability to participate in federated learning, train a local model, and provide local model training information.

[0084] It should be noted that the aforementioned time information for the first network element to support federated learning training is used to indicate the time during which the federated learning training supported by the first network element can be carried out. Federated learning training can achieve better performance within this time period, such as from 10 pm to 6 am the next day.

[0085] It should be noted that the metadata information of the first network element is used to indicate the data information that the first network element can cover, obtain, and provide.

[0086] Optionally, in this embodiment of the application, the metadata information of the first network element includes at least one of the following: input data type, output data type, data volume, and data range.

[0087] It should be noted that the above-mentioned input data type refers to the input data type of federated learning training, and the above-mentioned output data type refers to the output data type of federated learning training.

[0088] It should be noted that the above-mentioned input data type is used to indicate the data type that the first network element can collect for use as input data in the model training process, or the data type that the first network element can acquire.

[0089] Optionally, in this embodiment of the application, the above-mentioned data type refers to whether the data in the area has obvious characteristics, such as being concentrated in the morning and dispersed in the evening.

[0090] Optionally, in this embodiment of the application, the above-mentioned data type can be a data category, such as: UE location information, UE time information, network element load information, network status information, network element resource information, etc.

[0091] Optionally, in this embodiment of the application, the above-mentioned data range refers to the service range of the first network element.

[0092] Optionally, in this embodiment of the application, the above-mentioned data range includes at least one of the following: the service area of ​​the first network element, the area where the first network element can collect data, and the object where the first network element can collect data.

[0093] Optionally, in this embodiment of the application, the data that the first network element can collect includes metadata information and training data used for model training.

[0094] Optionally, in this embodiment of the application, the area where the first network element can collect data can be any of the following: the service area of ​​the first network element, the sub-area range under the service area of ​​the first network element, or the object of more granular data collection (e.g., a UE list).

[0095] Optionally, the amount of data includes the amount of data that can be used for model training.

[0096] Optionally, the data volume refers to the amount of data that the first network element has already collected;

[0097] Optionally, the data volume refers to the amount of data that the first network element is capable of acquiring, regardless of whether the data collection has been completed. Any amount of data that the first network element is capable of acquiring and has the authority to acquire can be considered as the data volume.

[0098] Optionally, in this embodiment of the application, the object on which the first network element can collect data may include one or more specific network elements, one or more UEs, or any UE.

[0099] Optionally, in this embodiment of the application, the registration request may further include at least one of the following: identification information of the first network element, identification information of data analysis supported by the first network element, information of model filter supported by the first network element, and information of the network element type of the first network element, wherein the network element type includes federated learning server network element type and / or federated learning client network element type.

[0100] Optionally, in this embodiment of the application, the data analysis identification information supported by the first network element can be user mobility trajectory (UE mobility) information.

[0101] Optionally, in this embodiment of the application, the above-mentioned data analysis identification information corresponds to the type information of the federated learning training.

[0102] Optionally, in this embodiment of the application, the above-mentioned data analysis identification information corresponds to the federated learning capability information of the first network element.

[0103] Optionally, in this embodiment of the application, the federated learning capability information of the first network element corresponds to the data analysis identification information. For example, the federated learning capability information of the first network element may correspond to a certain data analysis identification, that is, the first network element has federated learning capability for the task of the data analysis identification, or even the federated learning capability is a certain type of federated learning capability.

[0104] Optionally, in this embodiment of the application, there is a mapping relationship between the data analysis identification information and the federated learning capability information of the first network element. The mapping relationship can be one-to-one, many-to-one, or one-to-many. When the mapping relationship is one-to-one, the federated learning capability information of the first network element is different for different data analysis identification information.

[0105] For example, the first network element supports 10 data analysis identifiers. Data analysis identifiers 1-3 correspond to support for horizontal federated learning training, data analysis identifiers 4-6 correspond to support for vertical federated learning training, and data analysis identifiers 7-10 correspond to support for non-federated learning training.

[0106] For example, the first network element supports 5 data analysis identifiers. For data analysis identifiers 1-2, the first network element supports federated learning server capabilities; for data analysis identifiers 3-5, the first network element supports federated learning client capabilities.

[0107] Optionally, in this embodiment of the application, the identification information of the first network element can be any of the following: Fully Qualified Domain Name (FQDN) information or IP address information.

[0108] Optionally, in this embodiment of the application, the fully qualified domain name information is used to indicate the location of the first network element and the connection to the first network element.

[0109] It should be noted that the above-mentioned model filter information can also be called model effective range information. This model effective range information is used to indicate the effective range of the model generated by the training of the first network element.

[0110] Optionally, in this embodiment of the application, the above-mentioned effective range information of the model includes the slice to which the model is applicable, the external data network (DNN) to which the model is applicable, the region to which the model is applicable, and the time to which the model is applicable.

[0111] Optionally, in the embodiments of this application, the time to which the above model applies can be a large time range, such as within a year; or it can be a periodic time, such as 9:00-10:00 every day.

[0112] Optionally, in this embodiment of the application, the model filter information supported by the first network element includes the model filter information corresponding to the model generated by the first network element through federated learning.

[0113] Optionally, in this embodiment of the application, the federated learning capability information of the first network element further includes at least one of the following:

[0114] The algorithmic information upon which the federated learning training of the first network element is based;

[0115] Information on model accuracy achievable through federated learning training of the first network element;

[0116] The speed information of the first network element's federated learning and training;

[0117] Information on the model description methods supported by the first network element;

[0118] The first network element supports models that can share information;

[0119] First Network Element's vendor information;

[0120] Information on the number of federated learning models supported by the first network element.

[0121] It should be noted that the algorithm information on which the federated learning training of the first network element is based refers to the specific algorithm used in training the model, such as linear regression or neural networks.

[0122] It should be noted that the model accuracy information achievable through federated learning training of the first network element is used to indicate the accuracy of the model output.

[0123] It should be noted that the speed information of the federated learning training of the first network element mentioned above includes the time required for the model trained by the first network element based on federated learning to reach the target accuracy of the model.

[0124] 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 the model representation method based on the model description method information.

[0125] Optionally, in the embodiments of this application, the above-mentioned model description method information can also be referred to as model description method requirement information or model description method expectation information.

[0126] Optionally, in the embodiments of this application, the model description method information may be a model expression language represented by Open Neural Network Exchange (ONNX), or a model framework represented by TensorFlow, PyTorch, or other public or private model expression languages.

[0127] It should be noted that the shared model information supported by the first network element is used to indicate that the first network element supports a shared model. Here, "shareable" means that it can be interoperable, or "shareable" means that it can be understood by each other, and "shareable" means that it can be operated.

[0128] Optionally, in the embodiments of this application, the above-mentioned model-shareable information may also be referred to as model-shareable requirement information or model-shareable expectation information.

[0129] Specifically, when the model-shareable information supported by the first network element includes one or more vendor information, other network elements belonging to that vendor information can interact with the first network element, such as obtaining the model trained by the first network element and model information.

[0130] Specifically, when the model-shareable information supported by the first network element includes one or more model information, other network elements can interact with the first network element to obtain the model and / or model information corresponding to the one or more model information, such as obtaining the model and model information trained by the first network element. Conversely, other network elements cannot obtain models that are not included in the shareable information.

[0131] Optionally, when the model-shareable information supported by the first network element includes information from one or more network element objects, it indicates that the one or more network element objects are allowed to obtain the model from the first network element. The information of the network element objects includes the identification information, address information, FQDN information, etc. of the network element objects.

[0132] It should be noted that the information that can be shared in the above models can also be based on the granularity of data analysis tasks. That is, for each type of data analysis task, there is different information that can be shared (different vendor information, network element objects, model information, etc.).

[0133] It should be noted that the vendor information of the first network element mentioned above is used to indicate the vendor corresponding to the first network element.

[0134] It should be noted that the vendor information of the first network element is used by the second network element to determine whether the model generated by the first network element based on federated learning can be shared with other network elements.

[0135] For example, if the vendor information of other network elements is the same as that of the first network element, then the other network elements can share the model generated by the first network element based on federated learning.

[0136] Optionally, in this embodiment of the application, the vendor information of the first network element can be used by the second network element to select network elements that can participate in federated learning training.

[0137] It should be noted that the number of federated learning models supported by the first network element is used to indicate the number of federated learning-based models supported by the first network element for each data analysis identifier, such as the maximum and minimum number.

[0138] Optionally, in this embodiment of the application, the registration request may further include at least one of the following: type information of the first network element and service name supported by the first network element.

[0139] For example, the first network element is an NWDAF (AnLF) network element, and the service name supported by the NWDAF network element is Nnwdaf_AnalyticsInfo_Request. As another example, the first network element is an NWDAF (MTLF) network element, and the service name supported by the NWDAF network element is Nnwdaf_MLModelInfo. Optionally, in this embodiment, the type information of the first network element is used to indicate what type of network element is being registered. For example, if the type information of the first network element is NWDAF network element type, it indicates that the network element being registered is an NWDAF network element.

[0140] Optionally, in this embodiment of the application, after step 202, steps 203 and 204 are further included.

[0141] Step 203: The second network element sends a registration response to the first network element.

[0142] In this embodiment of the application, the above registration response is used to indicate that the first device capability registration was successful.

[0143] Optionally, in this embodiment of the application, the second network element can send an Nnrf_NFManagement_NFRegister response to the first network element to notify the first network element that the registration was successful.

[0144] Step 204: The first network element receives the registration response sent by the second network element.

[0145] This application provides a network element registration method. A first network element can register its federated learning capability information to a second network element through a registration request. This allows the first network element to be registered to the second network element, enabling other network elements to find the first network element capable of federated learning training through the second network element. This solves the problem of how the first network element can register its federated learning capability information to the second network element and be found by other network elements.

[0146] Optionally, in the embodiments of this application, combined with Figure 5 ,like Figure 6 As shown, after step 202 above, the network element registration method provided in this application embodiment further includes steps 205 and 206 as described below.

[0147] Step 205: The third network element sends a search request to the second network element.

[0148] In this embodiment of the application, the search request is used to request the search for network elements that can perform federated learning training.

[0149] Step 206: The second network element receives the search request sent by the third network element.

[0150] In this embodiment of the application, the search request includes first information, which includes at least one of the following: data analysis identification information corresponding to the target task, type information of federated learning training corresponding to the target task, time information of federated learning training corresponding to the target task, metadata information of the network elements of the federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information indicates the network element type corresponding to the network element that needs to be searched for that can perform federated learning training. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information indicates the quantity requirement of the network elements that need to be searched for that can perform federated learning training. The data volume requirement information indicates the data volume requirement of the network elements that need to be searched for that can perform federated learning training.

[0151] In this embodiment of the application, the third network element is the Model Training Logical Function (MTLF) network element, that is, the model training network element, which can be understood as the network element that initiates federated learning.

[0152] Optionally, in this embodiment of the application, the third network element may send an Nnrf_NFDiscovery_Request to the second network element to request the search for a network element capable of performing federated learning training.

[0153] Optionally, in this embodiment of the application, if the third network element does not support / does not perform independent training to generate model information corresponding to the target task, the third network element sends a search request to the second network element.

[0154] Optionally, in the embodiments of this application, the situation where the third network element does not support / does not perform independent training to generate model information corresponding to the target task may be due to insufficient data of the third network element or the accuracy of the model trained by the third network element not meeting the standard.

[0155] Optionally, in this embodiment of the application, the aforementioned third network element may be a network element with federated learning server capabilities, which can search for network elements with federated learning client capabilities to perform federated learning training.

[0156] Optionally, in this embodiment of the application, the third network element may be a network element with federated learning client capability. This third network element may seek network elements with federated learning server capability to request network elements with federated learning server capability to organize federated learning training.

[0157] It should be noted that the type information of federated learning training corresponding to the above-mentioned target task is used to instruct the network element information of the network element that supports the type of federated learning training corresponding to the target task to be obtained. For example, if the third network element determines that federated learning training is required, it instructs the network element information of the network element that supports federated learning training to be obtained. This network element information is used to instruct the network element to support federated learning training. Alternatively, if the third network element determines that horizontal federated learning training is required, it instructs the network element information of the network element that supports horizontal federated learning training to be obtained. This network element information is used to instruct the network element to support horizontal federated learning training.

[0158] It should be noted that the time information for federated learning training corresponding to the above-mentioned target task is used to indicate the time information for the third network element plan to conduct federated learning training, so as to find the network elements that support federated learning training within this time range. This time information is a future time period, such as 10 pm to 6 am the next day.

[0159] It should be noted that the metadata information of the network elements trained by federated learning for the aforementioned target tasks is used to indicate the data information that the target network elements can cover, acquire, and provide.

[0160] Optionally, in this embodiment of the application, the metadata information of the network element for federated learning training corresponding to the above-mentioned target task includes at least one of the following: input data type, output data type, data volume, and data range.

[0161] Optionally, in this embodiment of the application, the first information mentioned above also includes second information, which is used to request the acquisition of multiple network elements that can be trained through federated learning.

[0162] Optionally, in this embodiment of the application, the above-mentioned network element quantity requirement information may include at least one of the following: minimum quantity requirement information, suggested quantity information, and maximum quantity requirement information.

[0163] Optionally, in this embodiment of the application, the first information mentioned above further includes at least one of the following:

[0164] The algorithm information on which the federated learning training for the target task is based;

[0165] Information on the model accuracy achievable through federated learning training for the target task;

[0166] Speed ​​information of federated learning training corresponding to the target task;

[0167] Information on the model description methods supported by the network elements trained in federated learning for the target task;

[0168] The network elements trained by federated learning for the target task support models that can share information.

[0169] The vendor information of the network elements used in the federated learning training corresponding to the target task;

[0170] Information on the number of federated learning models supported by the network elements trained for the target task.

[0171] It should be noted that the algorithm information on which the federated learning training for the aforementioned target task is based is used to indicate the specific algorithm used in training the model, such as linear regression or neural networks.

[0172] It should be noted that the model accuracy information achievable through federated learning training for the aforementioned target tasks is used to indicate the required level of accuracy in the model's output.

[0173] It should be noted that the speed information of federated learning training corresponding to the above target task is used to indicate the time required for the network element to reach the target accuracy of the model based on federated learning training.

[0174] It should be noted that the model description information supported by the network elements trained by federated learning for the above-mentioned target tasks is used to indicate that the network elements need to support the model representation based on the model description information.

[0175] It should be noted that the model-shared information supported by the network elements trained by federated learning for the above-mentioned target tasks is used to indicate that the network elements need to support shared models.

[0176] It should be noted that the vendor information of the network elements trained by the federated learning for the aforementioned target task is used to indicate the vendor of the network element.

[0177] It should be noted that the number of federated learning models supported by the network elements trained by federated learning for the above-mentioned target tasks is used to indicate the number of federated learning-based models that the network element needs to support for each data analysis identifier.

[0178] Optionally, in the embodiments of this application, the first information mentioned above further includes at least one of the following: service information of the network element trained by federated learning corresponding to the target task, type information of the network element trained by federated learning corresponding to the target task, and region of interest information.

[0179] Optionally, in this embodiment of the application, if the target service is a request for data analysis identification information, the service information of the network element trained by the federated learning corresponding to the target task can be Nnwdaf_AnalyticsInfo_Request.

[0180] Optionally, in this embodiment of the application, if the network element of the target service is an NWDAF network element, then the type information of the network element trained by the federated learning corresponding to the target task can be NWDAF network element type information.

[0181] Optionally, in the embodiments of this application, the aforementioned region of interest information may be TA(s), cell(s), or other representations, used to indicate the request to discover network elements participating in federated learning training within the region of interest.

[0182] Optionally, in the embodiments of this application, combined with Figure 6 ,like Figure 7 As shown, after step 206 above, the network element registration method provided in this application embodiment further includes step 207 below.

[0183] Step 207: The second network element determines the target network element based on the search request.

[0184] In this embodiment of the application, the federated learning capability information of the target network element is matched with the first information.

[0185] It is understandable that the second network element identifies the network element (one or more network elements) that matches the federated learning capability information with the first information as the target network element.

[0186] Optionally, in the embodiments of this application, step 207 can be specifically implemented by steps 207a to 207f as described below.

[0187] Step 207a: The second network element determines the target network element based on the data analysis and identification information corresponding to the target task.

[0188] In this embodiment of the application, the target network element supports federated learning training corresponding to the data analysis identification information.

[0189] It is understandable that the second network element will identify the network element that supports federated learning training corresponding to the data analysis identification information as the target network element.

[0190] Step 207b: The second network element determines the target network element based on the type of federated learning training corresponding to the target task.

[0191] In this embodiment of the application, the target network element supports the type of federated learning training corresponding to the target task.

[0192] It is understandable that the second network element will identify the network element that supports the type of federated learning training corresponding to the target task as the target network element.

[0193] Optionally, in this embodiment of the application, the second network element can be determined as the target network element by identifying the network element that supports federated learning training corresponding to the target task.

[0194] Optionally, in this embodiment of the application, the second network element can be determined as the target network element if it supports the capabilities of the federated learning server corresponding to the target task.

[0195] Optionally, in this embodiment of the application, the second network element can be determined as the target network element if it supports the capabilities of the federated learning client corresponding to the target task.

[0196] Step 207c: The second network element determines the target network element based on the time information of the federated learning training corresponding to the target task.

[0197] In this embodiment of the application, the target network element supports federated learning training at the time corresponding to the target time information, where the target time information is the time information of federated learning training corresponding to the target task.

[0198] It is understandable that the second network element will identify the network element that supports federated learning training at the time corresponding to the target time information as the target network element.

[0199] Step 207d: The second network element determines the target network element based on the metadata information of the network element trained by federated learning corresponding to the target task.

[0200] In this embodiment of the application, the target network element supports the metadata information of the network element that supports federated learning training corresponding to the target task.

[0201] It is understandable that the second network element will identify the network element that supports the federated learning training corresponding to the target task as the target network element.

[0202] Step 207e: The second network element determines the target network element based on the network element quantity requirement information.

[0203] In this embodiment of the application, the number of the target network elements meets the network element quantity requirements.

[0204] It is understandable that when the number of network elements found by the second network element that can be used for federated learning training meets the number of network elements indicated by the network element quantity requirement information, the network elements that meet the quantity requirement can be identified as target network elements.

[0205] Step 207f: The second network element determines the target network element based on the network element type requirements information.

[0206] In this embodiment of the application, the network element type of the target network element meets the network element type requirements.

[0207] It is understandable that the second network element determines the network element that meets the network element type requirements as the target network element.

[0208] Optionally, in this embodiment of the application, the second network element determines the target network element based on the algorithm information on which the federated learning training corresponding to the target task is based.

[0209] It is understandable that the second network element will identify the network element that supports the algorithm on which the federated learning training corresponding to the target task is based as the target network element.

[0210] Optionally, in this embodiment of the application, the second network element determines the target network element based on the model accuracy information that can be achieved by federated learning training corresponding to the target task.

[0211] It is understandable that the second network element is the network element that can achieve the accuracy of the model trained by federated learning corresponding to the target task, and is identified as the target network element.

[0212] Optionally, in this embodiment of the application, the second network element determines the target network element based on the speed information of the federated learning training corresponding to the target task.

[0213] It is understandable that the second network element will identify the network element that supports the speed of federated learning training corresponding to the target task as the target network element;

[0214] Optionally, in this embodiment, the second network element determines the target network element based on the model description information supported by the network element trained by federated learning corresponding to the target task.

[0215] It is understandable that the second network element will identify the network element that supports the model description information (i.e., the model description information supported by the network element trained by federated learning for the target task) as the target network element.

[0216] Optionally, in this embodiment of the application, the second network element determines the target network element based on the model-shared information supported by the network element trained by the federated learning corresponding to the target task.

[0217] It is understandable that the second network element will identify the network element that supports the shared model as the target network element.

[0218] Optionally, in this embodiment of the application, the second network element determines the target network element based on the vendor information of the network element trained by the federated learning corresponding to the target task.

[0219] It is understandable that the second network element will identify the network element corresponding to the vendor information as the target network element.

[0220] Optionally, in this embodiment of the application, the second network element determines the target network element based on the number of federated learning models supported by the network element trained by the federated learning corresponding to the target task.

[0221] It is understandable that the second network element will identify the network element that supports the above-mentioned federated learning model quantity information (i.e., the quantity information of the federated learning models supported by the network elements trained in the federated learning corresponding to the target task) as the target network element.

[0222] Optionally, in this embodiment of the application, when the second network element determines the target network element based on the network element quantity requirement information, the network element registration method provided in this embodiment of the application further includes the following step A1.

[0223] Step A1: If the number of network elements that the second network element finds that can be used for federated learning training is less than the number of network elements indicated by the network element quantity requirement information, the second network element shall identify the network elements that can be used for federated learning training as the target network element.

[0224] It is understood that the network elements found above that can undergo federated learning training meet other requirements of the third network element (e.g., network element type, federated learning training time, etc.), but do not meet the network element quantity requirements.

[0225] Optionally, in this embodiment of the application, if the number of network elements that the second network element finds that can be trained through federated learning is equal to the number of network elements indicated by the network element quantity requirement information, the second network element determines the found network elements that can be trained through federated learning as target network elements.

[0226] It can be understood that the network elements found above that are capable of federated learning training are network elements that meet the requirements of the third network element.

[0227] Optionally, in this embodiment of the application, if the number of network elements that the second network element finds that can be used for federated learning training is greater than the number of network elements indicated by the network element quantity requirement information, the second network element will determine some of the network elements that can be used for federated learning training as target network elements.

[0228] Optionally, in this embodiment of the application, the second network element can sort all the network elements that can be trained by federated learning according to a preset internal logic, and then select the network element that meets the network element quantity requirement information from the sorting results as the target network element.

[0229] Optionally, in this embodiment of the application, if the amount of data possessed by the network element capable of federated learning training found by the second network element is greater than the amount of data possessed by the network element indicated by the data volume requirement information, the second network element will determine some of the network elements capable of federated learning training found as target network elements.

[0230] Optionally, in the embodiments of this application, combined with Figure 7 ,like Figure 8 As shown, after step 207 above, the network element registration method provided in this application embodiment further includes steps 208 and 209 as described below.

[0231] Step 208: The second network element sends a search response to the third network element.

[0232] In this embodiment of the application, the above-mentioned search response includes the identification information or address information of the target network element.

[0233] Step 209: The third network element receives the search response sent by the second network element.

[0234] Optionally, in this embodiment of the application, the identification information of the target network element can be any of the following: FQDN information, IP address information.

[0235] Optionally, in this embodiment of the application, the second network element can send an Nnrf_NFDiscovery_Request to the third network element to find the response, so as to send network element information of the network element that can perform federated learning training.

[0236] Optionally, in this embodiment of the application, the above-mentioned lookup response further includes second information, which includes at least one of the following: type information of federated learning training supported by the target network element, time information of the target network element supporting federated learning training, and metadata information of the target network element.

[0237] In this embodiment, the type information of federated learning training supported by the target network element is used to indicate the type of AI model training algorithm supported by the target network element.

[0238] In this embodiment of the application, the time information of the target network element supporting federated learning training is used to indicate the time during which the target network element supports federated learning training. Federated learning training can perform better within this time period, for example, from 10 pm to 6 am the next day.

[0239] In this embodiment of the application, the metadata information of the target network element is used to indicate the data information that the target network element can cover, obtain, and provide.

[0240] Optionally, in this embodiment of the application, the metadata information of the target network element includes at least one of the following: input data type, output data type, data volume, and data range.

[0241] Optionally, in this embodiment of the application, the above-mentioned search response may also include the type information of the target network element.

[0242] Optionally, in the embodiments of this application, the type information of the target network element is used to indicate what kind of network element the target network element is. For example, if the type information of the target network element is NWDAF network element type, it indicates that the target network element is an NWDAF network element.

[0243] Optionally, in this embodiment of the application, if the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, the search response further includes third information, which includes at least one of the following:

[0244] The first indication information is used to indicate that the number of target network elements is less than the number of network elements indicated by the network element quantity requirement information;

[0245] Information on the number of target network elements;

[0246] The fourth information, which is used to suggest that the third network element postpone the search for network elements that can be trained through federated learning, includes: time information for postponing the search for network elements that can be trained through federated learning.

[0247] Optionally, in the embodiments of this application, the aforementioned time information can be a specific time.

[0248] For example, the fourth piece of information mentioned above may include the time information "half an hour", so that when the third network element receives the third piece of information, it can use the time information to search for network elements that can perform federated learning training again after half an hour.

[0249] It is understandable that when the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, on the one hand, the second network element can identify the found network elements capable of federated learning training as target network elements and carry the identification information or address information of the target network elements in the search response; on the other hand, the second network element can carry third information in the search response to inform the third network element of at least one of the following: the number of target network elements does not meet the number indicated by the network element quantity requirement information, the number information of target network elements, or a suggestion that the third network element can postpone searching for network elements capable of federated learning training.

[0250] Optionally, in this embodiment of the application, when the number of network elements that the second network element finds that can be trained through federated learning is equal to the number of network elements indicated by the network element quantity requirement information, the search response may further include fifth information, which is used to indicate that the number of target network elements meets the number of network elements indicated by the network element quantity requirement information.

[0251] It is understandable that when the number of network elements capable of federated learning training found by the second network element is equal to the number of network elements indicated by the network element quantity requirement information, on the one hand, the second network element can identify the found network elements capable of federated learning training as target network elements and carry the identification information or address information of the target network elements in the search response; on the other hand, the second network element can carry the fifth information in the search response to inform the third network element that the number of target network elements meets the number indicated by the network element quantity requirement information.

[0252] Optionally, in this embodiment of the application, if the number of network elements capable of federated learning training found by the second network element is greater than the number of network elements indicated by the network element quantity requirement information, the search response may further include sixth information, which includes at least one of the following:

[0253] The third instruction information is used to indicate that the number of network elements found that can be trained through federated learning is greater than the number of network elements indicated by the network element number requirement information.

[0254] Information on network elements that can be trained in federated learning was found;

[0255] Information on the number of network elements found that can undergo federated learning training.

[0256] It is understandable that when the number of network elements capable of federated learning training found by the second network element exceeds the number of network elements indicated by the network element quantity requirement information, on the one hand, the second network element can select some network elements that meet the network element quantity requirement information as target network elements from the network elements capable of federated learning training, and carry the identification information or address information of the target network elements in the search response; on the other hand, the second network element can carry a sixth piece of information in the search response to inform the third network element of at least one of the following: relevant information of all network elements capable of federated learning training found by the second network element, quantity information of all network elements capable of federated learning training found by the second network element, and the number of network elements capable of federated learning training found by the second network element exceeds the number indicated by the network element quantity requirement information.

[0257] Optionally, in this embodiment, after receiving the third information, the third network element can obtain the model information corresponding to the target task by performing federated learning training with the target network element; it can also stop performing federated learning training by not performing federated learning training with the target network element determined by the second network element; or it can postpone and re-initiate the search request based on the fourth information or internal logic to search for network elements that can perform federated learning training and perform federated learning.

[0258] Optionally, in this embodiment, after receiving the sixth information, the third network element can select all or some network elements from the target network element for federated learning training to obtain the model information corresponding to the target task; or it can select some network elements from all network elements that can be federated learning trained by the second network element based on internal logic to obtain the model information corresponding to the target task.

[0259] Optionally, in this embodiment of the application, after step 209 above, the network element registration method provided in this embodiment of the application further includes the following step 209a.

[0260] Step 209a: The third network element selects all or some network elements from the target network element for federated learning training.

[0261] Optionally, in the embodiments of this application, all or some of the above-mentioned network elements can serve as clients or servers for federated learning training; specifically, the third network element can select a client or server for federated learning training from the target network element according to the network element type and network element type requirements information.

[0262] Optionally, in this embodiment of the application, the aforementioned third network element may select all or some network elements from the target network element for federated learning training based on the target network element's federated learning capability, the target network element's registration information, or the third network element's internal logic.

[0263] This application provides a network element registration method. A second network element can receive a search request sent by a third network element to determine the target network element that can be trained by federated learning. Then, it can send a search response to the third network element so that the third network element can determine the target network element that can be trained by federated learning. The second network element can then perform federated learning training with the target network element that can be trained by federated learning to obtain the model information corresponding to the target task.

[0264] It should be noted that the network element registration method provided in this application embodiment can also be executed by a network element registration device, or by a control module in the network element registration device used to execute the network element registration method.

[0265] This application provides a model determination method. Figure 9 A flowchart of a model determination method provided in an embodiment of this application is shown. Figure 9 As shown, the model determination method provided in this application embodiment may include the following steps 301 and 302.

[0266] Step 301: The third network element sends a search request to the second network element.

[0267] In this embodiment of the application, the search request is used to request the search for network elements that can perform federated learning training.

[0268] Step 302: The second network element receives the search request sent by the third network element.

[0269] Optionally, in this embodiment of the application, if the third network element does not support / does not perform independent training to generate model information corresponding to the target task, the third network element sends a search request to the second network element.

[0270] Optionally, in the embodiments of this application, the situation where the third network element does not support / does not perform independent training to generate model information corresponding to the target task may be due to insufficient data of the third network element or the accuracy of the model trained by the third network element not meeting the standard.

[0271] In this embodiment of the application, the third network element is the Model Training Logical Function (MTLF) network element, that is, the model training network element, which can be understood as the network element that initiates federated learning.

[0272] Optionally, in this embodiment of the application, the third network element may send an Nnrf_NFDiscovery_Request to the second network element to request the search for a network element capable of performing federated learning training.

[0273] In this embodiment of the application, the search request includes first information, which includes at least one of the following: data analysis identification information corresponding to the target task, type information of federated learning training corresponding to the target task, time information of federated learning training corresponding to the target task, metadata information of the network element of the federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element that needs to be searched for that can perform federated learning training. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element that needs to be searched for that can perform federated learning training. The data volume requirement information is used to indicate the data volume requirement of the network element that needs to be searched for that can perform federated learning training.

[0274] Optionally, in this embodiment of the application, the type information of federated learning training corresponding to the target task is used to indicate the acquisition of network element information of network elements that support the type of federated learning training corresponding to the target task. For example, if the third network element determines that federated learning training is required, it indicates the acquisition of network element information of network elements that support federated learning training. This network element information is used to indicate that the network element supports federated learning training. Alternatively, if the third network element determines that horizontal federated learning training is required, it indicates the acquisition of network element information of network elements that support horizontal federated learning training. This network element information is used to indicate that the network element supports horizontal federated learning training.

[0275] Optionally, in this embodiment of the application, the time information of the federated learning training corresponding to the above-mentioned target task is used to indicate the time information of the third network element plan to carry out federated learning training, so as to find the network element that supports federated learning training within this time range. The time information is a future time period, such as 10 pm to 6 am the next day.

[0276] Optionally, in this embodiment of the application, the metadata information of the network element trained by the federated learning corresponding to the above-mentioned target task is used to indicate the data information that the target network element can cover, obtain, and provide.

[0277] Optionally, in this embodiment of the application, the metadata information of the network element for federated learning training corresponding to the above-mentioned target task includes at least one of the following: input data type, output data type, data volume, and data range.

[0278] The amount of data includes the amount of data that can be used for model training.

[0279] Optionally, the data volume refers to the amount of data that the first network element has already collected;

[0280] Optionally, the data volume refers to the amount of data that the first network element is capable of acquiring, regardless of whether the data collection has been completed. Any amount of data that the first network element is capable of acquiring and has the authority to acquire can be considered as the data volume.

[0281] Optionally, in the embodiments of this application, the above-mentioned input data type refers to the input data type of federated learning training, and the above-mentioned output data type refers to the output data type of federated learning training.

[0282] Optionally, in the embodiments of this application, the above-mentioned data type refers to whether the input and output data of federated learning training have obvious characteristics, such as gathering in the morning and dispersing in the evening.

[0283] Optionally, in the embodiments of this application, the above-mentioned data range refers to the service range of the target network element.

[0284] Optionally, in the embodiments of this application, the above-mentioned data range includes at least one of the following: the service area of ​​the target network element, the area where the target network element can collect data, and the object on which the target network element can collect data.

[0285] Optionally, in this embodiment of the application, the data that the target network element can collect includes metadata information and training data used for model training.

[0286] Optionally, in this embodiment of the application, the area where the target network element can collect data can be any of the following: the target network element service area, the sub-area range under the target network element service area, or an object with a finer granularity of collectable data (e.g., a UE list).

[0287] Optionally, in this embodiment of the application, the first information mentioned above further includes at least one of the following:

[0288] The algorithm information on which the federated learning training for the target task is based;

[0289] Information on the model accuracy achievable through federated learning training for the target task;

[0290] Speed ​​information of federated learning training corresponding to the target task;

[0291] Information on the model description methods supported by the network elements trained in federated learning for the target task;

[0292] The network elements trained by federated learning for the target task support models that can share information.

[0293] The vendor information of the network elements used in the federated learning training corresponding to the target task;

[0294] Information on the number of federated learning models supported by the network elements trained for the target task.

[0295] Optionally, in the embodiments of this application, the algorithm information on which the federated learning training corresponding to the above-mentioned target task is based is used to indicate the specific algorithm used by the training model, such as linear regression or neural network.

[0296] Optionally, in this embodiment of the application, the model accuracy information that can be achieved by federated learning training for the above-mentioned target task is used to indicate the level of accuracy that the model output results need to achieve.

[0297] Optionally, in this embodiment of the application, the speed information of federated learning training corresponding to the above-mentioned target task is used to indicate the time required for the network element to reach the target accuracy of the model based on federated learning training.

[0298] Optionally, in the embodiments of this application, the first information mentioned above further includes at least one of the following: service information of the network element trained by federated learning corresponding to the target task, type information of the network element trained by federated learning corresponding to the target task, and region of interest information.

[0299] Optionally, in this embodiment of the application, if the target service is a request for data analysis identification information, the service information of the network element trained by the federated learning corresponding to the target task can be Nnwdaf_AnalyticsInfo_Request.

[0300] Optionally, in this embodiment of the application, if the network element of the target service is an NWDAF network element, then the type information of the network element trained by the federated learning corresponding to the target task can be NWDAF network element type information.

[0301] Optionally, in the embodiments of this application, the aforementioned region of interest information may be TA(s), cell(s), or other representations, used to indicate the request to discover network elements participating in federated learning training within the region of interest.

[0302] This application provides a model determination method. A third network element can use a search request to search for network elements capable of federated learning training from a second network element. Since the search request includes first information, the third network element can find target network elements that match the first information. Then, federated learning training is performed with the found target network elements capable of federated learning training.

[0303] Optionally, in the embodiments of this application, combined with Figure 9 ,like Figure 10 As shown, after step 302 above, the model determination method provided in this application embodiment further includes steps 303 to 305 as described below.

[0304] Step 303: The second network element sends a search response to the third network element.

[0305] In this embodiment of the application, the above-mentioned lookup response includes the identification information or address information of the target network element, and the target network element is a network element that supports federated learning training.

[0306] Optionally, in this embodiment of the application, the identification information of the target network element can be any of the following: FQDN information, IP address information.

[0307] Step 304: The third network element receives the search response sent by the second network element.

[0308] Optionally, in this embodiment of the application, the second network element can send an Nnrf_NFDiscovery_Request to the third network element to find the response, so as to send network element information of the network element that can perform federated learning training.

[0309] Optionally, in this embodiment of the application, the above-mentioned lookup response further includes second information, which includes at least one of the following: type information of federated learning training supported by the target network element, time information of the target network element supporting federated learning training, and metadata information of the target network element.

[0310] Optionally, in this embodiment of the application, the type information of federated learning training supported by the target network element is used to indicate the type of AI model training algorithm supported by the target network element.

[0311] Optionally, in this embodiment of the application, the time information of the target network element supporting federated learning training is used to indicate the time during which the target network element supports federated learning training. Federated learning training can perform better within this time period, for example, from 10 pm to 6 am the next day.

[0312] Optionally, in this embodiment of the application, the metadata information of the target network element is used to indicate the data information that the target network element can cover, acquire, and provide.

[0313] Optionally, in this embodiment of the application, the metadata information of the target network element includes at least one of the following: input data type, output data type, data volume, and data range.

[0314] Optionally, in this embodiment of the application, the above-mentioned search response may also include the type information of the target network element.

[0315] Optionally, in the embodiments of this application, the type information of the target network element is used to indicate what kind of network element the target network element is. For example, if the type information of the target network element is NWDAF network element type, it indicates that the target network element is an NWDAF network element.

[0316] Step 305: The third network element obtains the model information corresponding to the target task by performing federated learning training with the target network element.

[0317] Optionally, in this embodiment of the application, the third network element can determine and search for target network elements participating in federated learning training based on the information included in the received search response, so as to perform federated learning with the searched target network elements.

[0318] Optionally, in the embodiments of this application, combined with Figure 10 ,like Figure 11 As shown, prior to step 301 above, the model determination method provided in this application embodiment further includes steps 306 to 307 as described below.

[0319] Step 306: The fourth network element sends a model request to the third network element.

[0320] In this embodiment of the application, the above-mentioned model request is used to request model information corresponding to the target task.

[0321] Step 307: The third network element receives the model request sent by the fourth network element.

[0322] Optionally, in the embodiments of this application, the fourth network element is a model inference network element, which can be an AnLF network element, an NWDAF network element, or an NWDAF containing AnLF network element.

[0323] Optionally, in this embodiment of the application, the fourth network element can send an Nnwdaf_MLMoldelInfo_Request model request to the third network element to request the third network element to provide model information that matches the task.

[0324] Optionally, in the embodiments of this application, the above model request includes at least one of the following: data analysis identification information corresponding to the target task, and limitation information corresponding to the target task.

[0325] Optionally, in this embodiment of the application, the data analysis identification information corresponding to the above-mentioned target task can be user mobility trajectory (UE mobility) information.

[0326] Optionally, in the embodiments of this application, the limiting information corresponding to the target task is used to indicate the specific limiting conditions of the target task, such as limiting the time, location and other information required for the target task.

[0327] Optionally, in the embodiments of this application, the above-mentioned model request may further include at least one of the following: indication information of the target network element and reporting limitation information.

[0328] Optionally, in the embodiments of this application, the above-mentioned target network element indication information is used to indicate the target network element of the target task. For example, in the user mobility trajectory information, the SUPI and other identity identification information of the target network element can be specified.

[0329] Optionally, in this embodiment of the application, the above-mentioned report limitation information is used to indicate the information, format, etc. that the target task needs to report, for example, the sorting method is ascending order.

[0330] Optionally, in the embodiments of this application, combined with Figure 11 ,like Figure 12 As shown, after step 305 above, the model determination method provided in this application embodiment further includes steps 308 to 309 as described below.

[0331] Step 308: The third network element sends a model response to the fourth network element.

[0332] In this embodiment of the application, the above-mentioned model response includes model information corresponding to the target task, and the model information is used by the fourth network element for model inference.

[0333] Step 309: The fourth network element receives the model response sent by the third network element.

[0334] Optionally, in the embodiments of this application, the model information corresponding to the above-mentioned target task includes at least one of the following: target information and federated learning training related information.

[0335] Optionally, in this embodiment of the application, the target information includes at least one of the following: model identification information corresponding to the target task, model description information corresponding to the target task, model file corresponding to the target task, and model storage address information corresponding to the target task.

[0336] Optionally, in this embodiment of the application, the model identification information corresponding to the target task is used to identify the model corresponding to the target task, and the model identification information can be used to find the model description information, address information, etc. of the model.

[0337] Optionally, in this embodiment of the application, the model storage address information corresponding to the target task can be the URL or FQDN information of the model file, from which the model, model file, etc. corresponding to the target task can be obtained.

[0338] Optionally, in this embodiment of the application, the model file corresponding to the target task includes at least one of the following: 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.

[0339] Optionally, in this embodiment of the application, the above-mentioned federated learning training related information includes at least one of the following: second instruction information, identification information of network elements participating in federated learning training, and capability information of network elements participating in federated learning training.

[0340] In this embodiment of the application, the second indication information is used to indicate that the model information corresponding to the target task is model information obtained through federated learning training.

[0341] Optionally, in this embodiment of the application, the identification information of the network element participating in the federated learning training can be the identity and address information of the network element, such as fully qualified domain name information, IP address information, etc.

[0342] Optionally, in this embodiment, the third network element can send the identification information of the network elements participating in the federated learning training to the fourth network element, so that the fourth network element can perform inference in the subsequent federated learning training.

[0343] Optionally, in the embodiments of this application, the model information corresponding to the target task may further include at least one of the following: data analysis identification information corresponding to the target task, and limitation information corresponding to the target task.

[0344] Optionally, in this embodiment, the limiting information corresponding to the target task includes information about the region of interest or information about the target terminal. The model determination method provided in this embodiment may further include the following step 401.

[0345] Step 401: If the third network element cannot obtain the training data of the target task corresponding to the region of interest or the target terminal, the third network element performs federated learning training for the target task.

[0346] It should be noted that the region of interest mentioned above is the region where the model trained by federated learning needs / is to predict information.

[0347] For example, using a model to predict the weather in a certain area tomorrow.

[0348] It should be noted that the target terminal mentioned above is the terminal that requires / is to predict information based on the model trained by federated learning.

[0349] For example: predicting the target terminal's itinerary for tomorrow using a model.

[0350] Optionally, in this embodiment, the limiting information corresponding to the target task includes information about the region of interest or information about the target terminal. The model determination method provided in this embodiment may further include step 402.

[0351] Step 402: The third network element determines the metadata information of the network element corresponding to the federated learning training of the target task based on the information of the region of interest or the information of the target terminal.

[0352] In this embodiment of the application, the aforementioned metadata information includes the data range of network elements trained through federated learning.

[0353] This application provides a model determination method. A third network element can receive a model request sent by a fourth network element, requesting model information corresponding to a target task. Then, the third network element can search for network elements capable of federated learning training from a second network element through a search request. Since the search request includes first information, the third network element can find a target network element that matches the first information. By performing federated learning training with the found target network element, the model information corresponding to the target task is obtained, thereby improving the success rate of model transfer.

[0354] It should be noted that the model determination method provided in this application embodiment can also be executed by a model determination device, or by a control module in the model determination device for executing the model determination method.

[0355] The interaction process of the network element registration and model determination method provided in this application embodiment will be described in detail below through specific implementation methods.

[0356] like Figure 13 As shown, the method provided in this application embodiment includes the following steps 21 to 32.

[0357] Step 21: Intelligent devices in different domains (e.g., NWDAF network elements, base stations, UEs, third-party application servers, etc. on the core network side) send capability registration messages to capability storage devices such as NRF (e.g., NRF network elements, UDM network elements, DCAF network elements, etc.) to register their capabilities.

[0358] In this embodiment of the application, the NWDAF network element can be registered via (Nnrf_NFManagement_NFRegister Register).

[0359] Step 22: The NRF network element stores the information of the NWDAF network element.

[0360] Step 23: The NRF network element sends a registration response message.

[0361] In this embodiment of the application, the NRF network element can notify the NWDAF network element of successful registration by responding with the (Nnrf_NFManagement_NFRegister response) message.

[0362] These steps are similar to existing technologies, but the differences are:

[0363] In step 21, when the NWDAF network element sends a capability registration message to the NRF network element, in addition to its own identification information and supported analytics ID, it also sends "supported training type information", "supported federated learning time" and "metadata information".

[0364] The information required for registration includes:

[0365] 1. NF type, network element type; refers to the type of network element being registered. In this scheme, NF type = NWDAFtype, etc.

[0366] 2. NF instance ID, FQDN, or IP address of NF: Network element instance identification information; refers to the identification information of the network element registered this time, such as its FQDN information (Fully Qualified Domain Name, used to indicate the location and connection of this network element) or IP address information (another type of identification information).

[0367] 3. Names of supported NF services (if applicable): The names of services supported by the network element; for example, the NWDAF network element may have services named Nnwdaf_AnalyticsInfo_Request.

[0368] In addition, the network element registration information must also include at least one of the following:

[0369] 4. Supported training type information; refers to the types of AI model training algorithms supported by this network element, such as "federated learning", "deep learning", etc.

[0370] 5. Supported Federated Learning Time: This refers to the time period during which the network element supports federated learning. Federated learning can perform better within this time period, such as "10 pm to 6 am the next day".

[0371] 6. Meta data: refers to the data information that the network element can cover, obtain, and provide, including data type, data characteristics, data volume, etc.

[0372] 7. Analytics ID; such as "UE mobility" (user mobility trajectory), etc.

[0373] 8. Algorithms used for model training; the specific algorithms used to train the model, such as linear regression, neural networks, etc.

[0374] 9. Model training can achieve model accuracy; the accuracy of model output results.

[0375] 10. Model training speed; used to indicate the time required for a model to reach a specific level of accuracy.

[0376] Step 24: The task consumer sends a data analysis request message to the model inference network element AnLF (or NWDAF, NWDAF containing AnLF).

[0377] In this embodiment of the application, the task consumer can send a data analysis request message via (Nnwdaf_AnalyticsInfo_Request or Nnwdaf_AnalyticsSubscription_Subscribe).

[0378] Step 25: After receiving the task request, AnLF sends a model request to MTLF, requesting MTLF to provide a model that matches the task.

[0379] In this embodiment of the application, AnLF can send a model request via (Nnwdaf_MLMoldelInfo_Request), which includes task description information, such as:

[0380] 1. Analytics ID; such as "UE mobility" (user mobility trajectory), etc.

[0381] 2. Filter info: Task constraint information; indicates the specific constraints of this task, such as the time and location required for the task.

[0382] 3. Target network element / UE information; indicating the target network element or UE of the task. For example, in UE mobility, the identity information of the target UE, such as SUPI, can be specified.

[0383] 4. Reporting info: Specifies the information, format, etc., required for the task to be reported, such as the sorting order in ascending order.

[0384] Step 26: MTLF determines whether a model that meets the requirements can be trained independently based on the task description information in Step 25. If it cannot be trained independently, federated learning can be initiated.

[0385] In this embodiment of the application, the inability to train independently may be due to insufficient dataset quantity, insufficient data feature differentiation, or other situations that prevent the accuracy from reaching the target accuracy.

[0386] Step 27: The network element device MTLF, etc., which initiates federated training, sends a request to the storage capacity network element NRF, etc., to find network elements and other devices that can perform federated learning training.

[0387] In this embodiment of the application, the model training network element MTLF can send a search request via (Nnrf_NFDiscovery_Request).

[0388] These steps are similar to existing technologies, but the differences are:

[0389] MTLF sends a request to NRF to find network elements participating in federated learning, which must include:

[0390] 1. target NF service Name(s): Target network element service information; if the target service is a request for data analysis information, then this information can be Nnwdaf_AnalyticsInfo_Request.

[0391] 2. NF type of the target NF: The type of the target network element; for example, if the target network element is NWDAF, then this information can be NWDAF type.

[0392] 3. Area of ​​interest; can be TA(s), cell(s), or other representations, used to indicate a request to discover network elements participating in federated learning within the area of ​​interest.

[0393] When MTLF sends a request to NRF, it must also include "training type" and "training condition constraint information".

[0394] 4. Training type: This refers to the training type supported by the network element device that you want to obtain. For example, in this embodiment, if the network element wants to perform federated learning, then you want to obtain network element information that supports federated learning training. This information can be information that indicates that the network element supports federated learning, such as "federated learning".

[0395] 5. Training condition constraint information; this refers to the screening information for network elements and devices, further narrowing down the scope of network elements and devices to select more suitable network elements, such as "federated learning time information" and "required data information." It includes at least one of the following:

[0396] Supported Federated Learning Time: This refers to the target time information for planned federated learning, aiming to select network element devices that can perform federated learning within this time frame. This time information will be a future time period, such as "10 PM to 6 AM the next day".

[0397] Required data information; refers to the data information that the target network element is expected to cover, acquire, and provide, including data type, data characteristics, data volume, and other information.

[0398] Step 28: NRF returns the network element and other device information that matches the information retrieved in step 27 to MTLF.

[0399] In this embodiment of the application, the network element such as NRF capability storage network element returns network element device information that conforms to the training condition limitation information in step 26 to the federated learning initiator, and can return a response message through (Nnrf_NFDiscovery_Request).

[0400] This should include:

[0401] 1. Network element identification information; indicating the FQDN, IP address, etc. of the target network element or device.

[0402] 2. Network element device type information; such as NWDAF.

[0403] 3. Network element device capability information; including at least one of the following:

[0404] Training type; refers to the training type supported by the network element device that you want to obtain, such as "federated learning".

[0405] Supported Federated Learning Time: This refers to the time period during which the network element supports federated learning. Federated learning can perform better within this time period, such as "10 pm to 6 am the next day".

[0406] Metadata information refers to the data information that the network element can cover, obtain, and provide, including data type, data characteristics, data volume, and other information.

[0407] Step 29: MTLF determines and finds network elements and other devices to participate in training based on the information returned in Step 28. MTLF then performs federated learning with the found network elements and other devices.

[0408] Step 30: After MTLF training is complete, send the feedback model to AnLF, which includes task-specific descriptive information.

[0409] 1. analytic ID; information indicating the task type.

[0410] 2. Task limitation information; further describing the task requirements, such as time limits, location limits, etc.

[0411] 3. Models, etc.;

[0412] It may also include identification and competency information about federated learning members, including at least one of the following:

[0413] 4. Federated Learning Information; Instructions to complete this training using federated learning.

[0414] 5. Identification information of members participating in federated learning; this can be the member's identity and address information, such as FQDN or IP address. MTLF will send the network element and other device information participating in federated learning to AnLF for AnLF to use in subsequent federated learning inference.

[0415] 6. Information on the capabilities of members participating in federated learning; including information on the time frame for supporting federated learning and data about the devices themselves.

[0416] Step 31: After receiving the feedback model and other information, AnLF performs model inference.

[0417] In this embodiment of the application, AnLF can perform cooperative reasoning based on the network element and other device information participating in federated learning from the obtained federated learning information.

[0418] Step 32: AnLF sends a task report to the consumer based on the reasoning results.

[0419] It should be noted that the relevant explanations and beneficial effects of steps 21 to 32 above can be found in the descriptions in the above embodiments, and will not be repeated here.

[0420] Figure 14 A schematic diagram of a possible structure of the network element registration device involved in an embodiment of this application is shown. For example... Figure 14 As shown, the network element registration device 80 may include: a sending module 81.

[0421] The sending module 81 is used to send a registration request to the second network element. The registration request is used to request the registration of the federated learning capability information of the first network element to the second network element. The federated learning capability information of the first network element includes at least one of the following: the type information of federated learning training supported by the first network element, the time information of the first network element supporting federated learning training, and the metadata information of the first network element.

[0422] This application provides a network element registration device. The network element registration device can register the federated learning capability information of a first network element to a second network element through a registration request, so as to realize the registration of the first network element to the second network element. This allows other network elements to find the first network element that can perform federated learning training through the second network element, thereby solving the problem of how the first network element registers its federated learning capability information to the second network element and can be found by other network elements.

[0423] In one possible implementation, the type information of federated learning training supported by the first network element includes at least one of the following: indication information on whether the first network element supports federated learning training, horizontal federated learning training type, vertical federated learning training type, federated learning server capability, and federated learning client capability.

[0424] In one possible implementation, the metadata information of the first network element mentioned above includes at least one of the following: input data type, output data type, data volume, and data range.

[0425] Optionally, the amount of data includes the amount of data that can be used for model training.

[0426] In one possible implementation, the aforementioned data range includes at least one of the following: the service area of ​​the first network element, the area where the first network element can collect data, and the object on which the first network element can collect data.

[0427] In one possible implementation, the registration request may further include at least one of the following: identification information of the first network element, identification information of data analysis supported by the first network element, information of model filters supported by the first network element, and information of the network element type of the first network element, wherein the network element type includes the federated learning server network element type and / or the federated learning client network element type.

[0428] In one possible implementation, the aforementioned data analysis identification information corresponds to the type information of federated learning training.

[0429] In one possible implementation, the aforementioned data analysis identification information corresponds to the federated learning capability information of the first network element.

[0430] In one possible implementation, the aforementioned model filter information includes the model filter information corresponding to the model generated by the first network element through federated learning.

[0431] In one possible implementation, the federated learning capability information of the first network element further includes at least one of the following: algorithm information on which the federated learning training of the first network element is based; model accuracy information achievable by the federated learning training of the first network element; speed information of the federated learning training of the first network element; model description method information supported by the first network element; model sharing information supported by the first network element; vendor information of the first network element; and the number of federated learning models supported by the first network element.

[0432] In one possible implementation, the speed information of the federated learning training of the first network element includes the time required for the first network element to reach the target accuracy of the model based on the federated learning training.

[0433] The network element registration device provided in this application embodiment can implement all the processes implemented by the first network element in the above method embodiment and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0434] Figure 15 A schematic diagram of a possible structure of the network element registration device involved in an embodiment of this application is shown. For example... Figure 15 As shown, the network element registration device 90 may include a receiving module 91.

[0435] The receiving module 91 is used to receive a registration request sent by the first network element. The registration request is used to request the first network element's federated learning capability information to be registered with the second network element. The first network element's federated learning capability information includes at least one of the following: the type information of federated learning training supported by the first network element, the time information of the first network element supporting federated learning training, and the metadata information of the first network element.

[0436] This application provides a network element registration device. The network element registration device can register the federated learning capability information of the first network element to the second network element by receiving a registration request sent by the first network element. This enables other network elements to find the first network element that can perform federated learning training, thus solving the problem of how the first network element can register its federated learning capability information to the second network element and be found by other network elements.

[0437] Optionally, the metadata information of the first network element includes at least one of the following: input data type, output data type, data volume, and data range.

[0438] Optionally, the amount of data includes the amount of data that can be used for model training.

[0439] Optionally, the data volume refers to the amount of data that the first network element has already collected;

[0440] Optionally, the data volume refers to the amount of data that the first network element is capable of acquiring, regardless of whether the data collection has been completed. Any amount of data that the first network element is capable of acquiring and has the authority to acquire can be considered as the data volume.

[0441] In one possible implementation, the federated learning capability information of the first network element mentioned above also includes at least one of the following:

[0442] Information on the algorithm upon which the federated learning training of the first network element is based; information on the model accuracy achievable through the federated learning training of the first network element; information on the speed of the federated learning training of the first network element; information on the model description methods supported by the first network element; information on the model sharing capabilities supported by the first network element; information on the vendor of the first network element; information on the number of federated learning models supported by the first network element.

[0443] In one possible implementation, the receiving module 91 is further configured to receive a search request sent by a third network element. The search request is used to search for network elements capable of performing federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification information corresponding to the target task, type information of federated learning training corresponding to the target task, time information of federated learning training corresponding to the target task, metadata information of the network element for federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element capable of performing federated learning training to be searched. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element capable of performing federated learning training to be searched. The data volume requirement information is used to indicate the data volume requirement of the network element capable of performing federated learning training to be searched.

[0444] In one possible implementation, the aforementioned first information further includes at least one of the following: algorithm information on which the federated learning training corresponding to the target task is based; 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 the network elements of the federated learning training corresponding to the target task; model sharing information supported by the network elements of the federated learning training corresponding to the target task; vendor 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 the network elements of the federated learning training corresponding to the target task.

[0445] In one possible implementation, the network element registration device 90 provided in this application embodiment further includes a determining module. The determining module is configured to determine a target network element based on the search request after the receiving module 91 receives a search request sent by a third network element, wherein the federated learning capability information of the target network element matches the first information.

[0446] In one possible implementation, the aforementioned determining module is specifically used to: determine target network elements based on data analysis identification information corresponding to the target task, wherein the target network elements support federated learning training corresponding to the data analysis identification information; determine target network elements based on the type of federated learning training corresponding to the target task, wherein the target network elements support the type of federated learning training corresponding to the target task; determine target network elements based on the time information of federated learning training corresponding to the target task, wherein the target network elements support federated learning training at the time corresponding to the target time information, where the target time information is the time information of federated learning training corresponding to the target task; determine target network elements based on the metadata information possessed by the network elements in the federated learning training corresponding to the target task, wherein the target network elements support the metadata information possessed by the network elements in the federated learning training corresponding to the target task; determine target network elements based on network element quantity requirement information, wherein the quantity of target network elements meets the network element quantity requirement information; and determine target network elements based on network element type requirement information, wherein the network element type of the target network elements meets the network element type requirement information.

[0447] In one possible implementation, when the second network element determines the target network element based on the network element quantity requirement information, the aforementioned determining module is used to determine the found network element capable of federated learning training as the target network element if the number of network elements found by the second network element that can be trained through federated learning is less than the number of network elements indicated by the network element quantity requirement information.

[0448] In one possible implementation, the network element registration device 90 provided in this application embodiment further includes a sending module. The sending module is configured to send a search response to a third network element after the determining module determines the target network element based on the search request. The search response includes the identification information or address information of the target network element.

[0449] In one possible implementation, the above lookup response further includes second information, which includes at least one of the following: type information of federated learning training supported by the target network element, time information of the target network element supporting federated learning training, and metadata information of the target network element.

[0450] In one possible implementation, if the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, the search response further includes third information, which includes at least one of the following: first indication information, which indicates that the number of target network elements is less than the number of network elements indicated by the network element quantity requirement information; target network element quantity information; and fourth information, which suggests that the third network element postpone searching for network elements capable of federated learning training, and the fourth information includes: time information for postponing the search for network elements capable of federated learning training.

[0451] The network element registration device provided in this application embodiment can implement all the processes implemented by the second network element in the above method embodiment and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0452] Figure 16 A schematic diagram of a possible structure of the model determination device involved in an embodiment of this application is shown. For example... Figure 16 As shown, the model determination device 100 may include a sending module 101.

[0453] The sending module 101 is used to send a search request to the second network element. The search request is used to search for network elements that can perform federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification information corresponding to the target task, type information of federated learning training corresponding to the target task, time information of federated learning training corresponding to the target task, metadata information of the network element that performs federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element that can perform federated learning training to be searched. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element that can perform federated learning training to be searched. The data volume requirement information is used to indicate the data volume requirement of the network element that can perform federated learning training to be searched.

[0454] This application provides a model determination device. The device can search for network elements capable of federated learning training from a second network element via a search request. Since the search request includes first information, a third network element can find a target network element that matches the first information. Then, federated learning training is performed between the third network element and the found target network element capable of federated learning training.

[0455] In one possible implementation, the metadata information of the network element trained by federated learning for the aforementioned target task includes at least one of the following: input data type, output data type, data volume, and data range.

[0456] Optionally, the amount of data includes the amount of data that can be used for model training.

[0457] Optionally, the data volume refers to the amount of data that the first network element has already collected;

[0458] Optionally, the data volume refers to the amount of data that the first network element is capable of acquiring, regardless of whether the data collection has been completed. Any amount of data that the first network element is capable of acquiring and has the authority to acquire can be considered as the data volume.

[0459] In one possible implementation, the aforementioned first information further includes at least one of the following: algorithm information on which the federated learning training corresponding to the target task is based; 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 the network elements of the federated learning training corresponding to the target task; model sharing information supported by the network elements of the federated learning training corresponding to the target task; vendor 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 the network elements of the federated learning training corresponding to the target task.

[0460] In one possible implementation, the model determination apparatus 100 provided in this application embodiment 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 the identification information or address information of the target network element, and the target network element is a network element that supports federated learning training. The acquiring module is used to obtain model information corresponding to the target task by performing federated learning training with the target network element.

[0461] In one possible implementation, the above lookup response further includes second information, which includes at least one of the following: type information of federated learning training supported by the target network element, time information of the target network element supporting federated learning training, and metadata information of the target network element.

[0462] In one possible implementation, if the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, the search response further includes third information, which includes at least one of the following: first indication information, which indicates that the number of target network elements is less than the number of network elements indicated by the network element quantity requirement information; target network element quantity information; and fourth information, which suggests that the third network element postpone searching for network elements capable of federated learning training, and the fourth information includes: time information for postponing the search for network elements capable of federated learning training.

[0463] In one possible implementation, the model determination device 100 provided in the application embodiment further includes: a selection module; the selection module is used to select all or some network elements from the target network elements for federated learning training after the receiving module receives the search response sent by the second network element.

[0464] In one possible implementation, the receiving module is further configured to receive a model request sent by the fourth network element before the sending module 101 sends a lookup request to the second network element. The model request is used to request model information corresponding to the target task, and the model request includes at least one of the following: data analysis identification information corresponding to the target task and limitation information corresponding to the target task.

[0465] In one possible implementation, the sending module 101 is further configured to send a model response to the fourth network element after the acquisition module obtains the model information corresponding to the target task through federated learning training with the target network element. The model response includes the model information corresponding to the target task, and the model information is used by the fourth network element for model inference.

[0466] In one possible implementation, the model determination device 100 provided in this application embodiment further includes an execution module. The limiting information corresponding to the target task includes information about the region of interest or the target terminal; 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 region of interest or the target terminal.

[0467] In one possible implementation, the model determination device 100 provided in this application embodiment further includes: a determination module. The limiting information corresponding to the target task includes information about the region of interest or information about the target terminal; the determination module is used to determine the metadata information of the network element trained by federated learning corresponding to the target task based on the information about the region of interest or the information about the target terminal, the metadata information including the data range of the network element trained by federated learning.

[0468] In one possible implementation, the model information corresponding to the target task includes at least one of the following: target information and federated learning training-related information. The target information includes at least one of the following: model identification information of the model, model description information corresponding to the target task, model file corresponding to the target task, and model storage address information corresponding to the target task.

[0469] In one possible implementation, the above-mentioned federated learning training related information includes at least one of the following: second instruction information, identification information of network elements participating in federated learning training, and capability information of network elements participating in federated learning training. The second instruction information is used to indicate that the model information corresponding to the target task is the model information obtained through federated learning training.

[0470] The model determination device provided in this application embodiment can implement the various processes implemented by the third network element in the above method embodiment and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0471] Optionally, such as Figure 17As shown, this application embodiment also provides a communication device 5000, including a processor 5001 and a memory 5002. The memory 5002 stores a program or instructions that can run on the processor 5001. For example, when the communication device 5000 is a network element, when the program or instructions are executed by the processor 5001, they implement the various steps of the first network element side method embodiment described above and achieve the same technical effect, or implement the various steps of the second network element side method embodiment described above and achieve the same technical effect, or implement the various steps of the third network element side method embodiment described above and achieve the same technical effect. To avoid repetition, further details are omitted here.

[0472] This application embodiment also provides a network element, including a processor and a communication interface. The communication interface is used to send a registration request to a second network element. The registration request is used to request the registration of the federated learning capability information of a first network element to the second network element. The federated learning capability information of the first network element includes at least one of the following: information on the type of federated learning training supported by the first network element, information on the time of federated learning training supported by the first network element, and metadata information possessed by the first network element. This network element embodiment corresponds to the above-described method embodiment on the first network element side. All implementation processes and methods of the above method embodiments can be applied to this network element embodiment and can achieve the same technical effect.

[0473] This application embodiment also provides a network element, including a processor and a communication interface. The communication interface is used to receive a registration request sent by a first network element. The registration request is used to request the registration of the federated learning capability information of the first network element to a second network element. The federated learning capability information of the first network element includes at least one of the following: information on the type of federated learning training supported by the first network element, information on the time of federated learning training supported by the first network element, and metadata information possessed by the first network element. This network element embodiment corresponds to the above-described method embodiment on the second network element side. All implementation processes and methods of the above method embodiments can be applied to this network element embodiment and can achieve the same technical effect.

[0474] This application embodiment also 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 search for network elements capable of performing federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification information corresponding to the target task, type information of federated learning training corresponding to the target task, time information of federated learning training corresponding to the target task, metadata information of the network element for federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the type of network element corresponding to the network element capable of performing federated learning training to be searched. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element capable of performing federated learning training to be searched. The data volume requirement information is used to indicate the data volume requirement of the network element capable of performing federated learning training to be searched. This network element embodiment corresponds to the above-described third network element side method embodiment. All implementation processes and methods of the above method embodiments can be applied to this network element embodiment and can achieve the same technical effect.

[0475] Specifically, Figure 18 To realize a hardware structure diagram of a network element in an embodiment of this application, 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.

[0476] like Figure 18 As shown, network element 1200 includes: processor 1201, network interface 1202, and memory 1203. The network interface 1202 is, for example, a general-purpose public wireless interface.

[0477] Specifically, the network element 1200 in this application embodiment further includes: instructions or programs stored in memory 1203 and executable on processor 1201. Processor 1201 calls the instructions or programs in memory 1203 to execute the methods executed by the above modules and achieve the same technical effect. To avoid repetition, it will not be described in detail here.

[0478] The network element provided in this application embodiment can implement the various processes implemented by the first network element, the second network element and the third network element in the above method embodiment, and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0479] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0480] The processor is the processor in the communication device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0481] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0482] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0483] This application also provides a computer program / program product, which is stored in a storage medium and executed by at least one processor to implement the various processes of the above method embodiments and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0484] This application embodiment also provides a communication system, including: a first network element, a second network element, and a third network element. The first network element can be used to perform the steps of the network element registration method described above, the second network element can be used to perform the steps of the network element registration method described above, and the third network element can be used to perform the steps of the model determination method described above.

[0485] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0486] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0487] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for registering network elements, characterized in that, The method includes: The first network element sends a registration request to the second network element. The registration request is used to request the first network element's federated learning capability information to be registered to the second network element. The first network element's federated learning capability information includes at least one of the following: the type information of federated learning training supported by the first network element, the time information of the first network element supporting federated learning training, and the metadata information of the first network element.

2. The method according to claim 1, characterized in that, The federated learning training type information supported by the first network element includes at least one of the following: indication information on whether the first network element supports federated learning training, horizontal federated learning training type, vertical federated learning training type, federated learning server capability, and federated learning client capability.

3. The method according to claim 1, characterized in that, The metadata information possessed by the first network element includes at least one of the following: input data type, output data type, data volume, and data range.

4. The method according to claim 3, characterized in that, The data range includes at least one of the following: the service area of ​​the first network element, the area where the first network element can collect data, and the object where the first network element can collect data.

5. The method according to any one of claims 1 to 4, characterized in that, The registration request also includes at least one of the following: the identification information of the first network element, the identification information of the data analysis supported by the first network element, the model filter information supported by the first network element, and the network element type information of the first network element, wherein the network element type includes the federated learning server network element type and / or the federated learning client network element type.

6. The method according to claim 5, characterized in that, The data analysis identification information corresponds to the type information of the federated learning training.

7. The method according to claim 5, characterized in that, The data analysis identification information corresponds to the federated learning capability information of the first network element.

8. The method according to claim 5, characterized in that, The model filter information includes the model filter information corresponding to the model generated by the first network element through federated learning.

9. The method according to claim 1, characterized in that, The federated learning capability information of the first network element also includes at least one of the following: The algorithm information upon which the federated learning training of the first network element is based; Information on the model accuracy achievable through federated learning training of the first network element; The speed information of the federated learning training of the first network element; Information on the model description methods supported by the first network element; The models supported by the first network element can share information; The vendor information of the first network element; Information on the number of federated learning models supported by the first network element.

10. The method according to claim 9, characterized in that, The speed information of the federated learning training of the first network element includes the time required for the model trained by the first network element based on federated learning to reach the target accuracy of the model.

11. The method according to claim 3, characterized in that, The amount of data includes the amount of data that can be used for model training.

12. A method for registering network elements, characterized in that, The method includes: The second network element receives a registration request sent by the first network element. The registration request is used to request the first network element's federated learning capability information to be registered to the second network element. The first network element's federated learning capability information includes at least one of the following: the type information of federated learning training supported by the first network element, the time information of the first network element supporting federated learning training, and the metadata information of the first network element.

13. The method according to claim 12, characterized in that, The federated learning capability information of the first network element also includes at least one of the following: The algorithm information upon which the federated learning training of the first network element is based; Information on the model accuracy achievable through federated learning training of the first network element; The speed information of the federated learning training of the first network element; Information on the model description methods supported by the first network element; The models supported by the first network element can share information; The vendor information of the first network element; Information on the number of federated learning models supported by the first network element.

14. The method according to claim 12, characterized in that, After the second network element receives the registration request sent by the first network element, the method further includes: The second network element receives a search request sent by the third network element. The search request is used to search for network elements capable of performing federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification 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 of the network element for the federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element capable of performing federated learning training to be searched. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element capable of performing federated learning training to be searched. The data volume requirement information is used to indicate the data volume requirement of the network element capable of performing federated learning training to be searched.

15. The method according to claim 14, characterized in that, The first information also includes at least one of the following: The algorithm information on which the federated learning training corresponding to the target task is based; Information on the model accuracy achievable through federated learning training for the target task; The speed information of federated learning training corresponding to the target task; Information on the model description methods supported by the network elements trained by federated learning for the target task; The network elements trained by federated learning for the target task can share information with the model. The vendor information of the network elements used in the federated learning training corresponding to the target task; Information on the number of federated learning models supported by the network elements trained for the target task.

16. The method according to claim 14 or 15, characterized in that, After the second network element receives the search request sent by the third network element, the method further includes: The second network element determines the target network element based on the search request, and the federated learning capability information of the target network element matches the first information.

17. The method according to claim 16, characterized in that, The second network element determines the target network element based on the search request, including at least one of the following: The second network element determines the target network element based on the data analysis identification information corresponding to the target task, wherein the target network element supports federated learning training corresponding to the data analysis identification information; The second network element determines the target network element according to the type of federated learning training corresponding to the target task, wherein the target network element supports the type of federated learning training corresponding to the target task; The second network element determines the target network element based on the time information of the federated learning training corresponding to the target task, wherein the target network element supports federated learning training at the time corresponding to the target time information, and the target time information is the time information of the federated learning training corresponding to the target task; The second network element determines the target network element based on the metadata information of the network element trained by the federated learning corresponding to the target task, wherein the target network element supports the metadata information of the network element trained by the federated learning corresponding to the target task. The second network element determines the target network element based on the network element quantity requirement information, wherein the quantity of the target network element satisfies the network element quantity requirement information; The second network element determines the target network element based on the network element type requirement information, wherein the network element type of the target network element satisfies the network element type requirement information.

18. The method according to claim 17, characterized in that, When the second network element determines the target network element based on the network element quantity requirement information, the method further includes: If the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, the second network element determines the found network elements capable of federated learning training as the target network element.

19. The method according to claim 16, characterized in that, After the second network element determines the target network element according to the search request, the method further includes: The second network element sends a lookup response to the third network element, the lookup response including the identification information or address information of the target network element.

20. The method according to claim 19, characterized in that, The search response also includes second information, which includes at least one of the following: The target network element supports federated learning training type information, the target network element supports federated learning training time information, and the target network element has metadata information.

21. The method according to claim 19, characterized in that, If the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, the search response further includes third information, which includes at least one of the following: The first indication information is used to indicate that the number of target network elements is less than the number of network elements indicated by the network element quantity requirement information; The quantity information of the target network elements; The fourth information is used to suggest that the third network element postpone searching for network elements that can be trained through federated learning. The fourth information includes: time information for postponing the search for network elements that can be trained through federated learning.

22. The method according to claim 12, characterized in that, The metadata information possessed by the first network element includes at least one of the following: input data type, output data type, data volume, and data range.

23. The method according to claim 22, characterized in that, The amount of data includes the amount of data that can be used for model training.

24. A model determination method, characterized in that, The method includes: The third network element sends a search request to the second network element. The search request is used to search for network elements that can perform federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification 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 of the network element that performs federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element that needs to be searched for that can perform federated learning training. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element that needs to be searched for that can perform federated learning training. The data volume requirement information is used to indicate the data volume requirement of the network element that needs to be searched for that can perform federated learning training.

25. The method according to claim 24, characterized in that, The metadata information of the network element trained by federated learning for the target task includes at least one of the following: input data type, output data type, data volume, and data range.

26. The method according to claim 24, characterized in that, The first information also includes at least one of the following: The algorithm information on which the federated learning training corresponding to the target task is based; Information on the model accuracy achievable through federated learning training for the target task; The speed information of federated learning training corresponding to the target task; Information on the model description methods supported by the network elements trained by federated learning for the target task; The network elements trained by federated learning for the target task can share information with the model. The vendor information of the network elements used in the federated learning training corresponding to the target task; Information on the number of federated learning models supported by the network elements trained for the target task.

27. The method according to any one of claims 24 to 26, characterized in that, After the third network element sends a search request to the second network element, the method further includes: The third network element receives a search response sent by the second network element. The search response includes the identification information or address information of the target network element, which is a network element that supports federated learning training. The third network element obtains the model information corresponding to the target task by performing federated learning training with the target network element.

28. The method according to claim 27, characterized in that, The search response also includes second information, which includes at least one of the following: The target network element supports federated learning training type information, the target network element supports federated learning training time information, and the target network element has metadata information.

29. The method according to claim 27, characterized in that, If the number of network elements capable of federated learning training found by the second network element is less than the number of network elements indicated by the network element quantity requirement information, the search response further includes third information, which includes at least one of the following: The first indication information is used to indicate that the number of target network elements is less than the number of network elements indicated by the network element quantity requirement information; The quantity information of the target network elements; The fourth information is used to suggest that the third network element postpone searching for network elements that can be trained through federated learning. The fourth information includes: time information for postponing the search for network elements that can be trained through federated learning.

30. The method according to claim 27, characterized in that, After the third network element receives the search response sent by the second network element, the method further includes: The third network element selects all or some network elements from the target network element for federated learning training.

31. The method according to claim 24, characterized in that, Before the third network element sends a search request to the second network element, the method further includes: The third network element receives a model request sent by the fourth network element. The model request is used to request model information corresponding to the target task. The model request includes at least one of the following: data analysis identification information corresponding to the target task and limitation information corresponding to the target task.

32. The method according to claim 27, characterized in that, After the third network element obtains the model information corresponding to the target task through federated learning training with the target network element, the method further includes: The third network element sends a model response to the fourth network element. The model response includes model information corresponding to the target task. The model information is used by the fourth network element to perform model inference.

33. The method according to claim 31, characterized in that, The limiting information corresponding to the target task includes information about the region of interest or information about the target terminal; the method further includes: If the third network element is unable to obtain training data for the target task corresponding to the region of interest or the target terminal, the third network element performs federated learning training for the target task.

34. The method according to claim 31, characterized in that, The limiting information corresponding to the target task includes information about the region of interest or information about the target terminal; the method further includes: The third network element determines the metadata information of the network element for federated learning training corresponding to the target task based on the information of the region of interest or the information of the target terminal. The metadata information includes the data range of the network element for federated learning training.

35. The method according to claim 31, characterized in that, The model information corresponding to the target task includes at least one of the following: target information and federated learning training-related information. The target information includes at least one of the following: model identification information of the model, model description information corresponding to the target task, model file corresponding to the target task, and model storage address information corresponding to the target task.

36. The method according to claim 35, characterized in that, The federated learning training related information includes at least one of the following: second instruction information, identification information of network elements participating in federated learning training, and capability information of network elements participating in federated learning training. The second instruction information is used to indicate that the model information corresponding to the target task is model information obtained through federated learning training.

37. The method according to claim 25, characterized in that, The amount of data includes the amount of data that can be used for model training.

38. A network element registration device, applied to a first network element, characterized in that, The device includes: a transmitting module; The sending module is used to send a registration request to the second network element. The registration request is used to request the registration of the federated learning capability information of the first network element to the second network element. The federated learning capability information of the first network element includes at least one of the following: the type information of federated learning training supported by the first network element, the time information of the federated learning training supported by the first network element, and the metadata information of the first network element.

39. A network element registration device, applied to a second network element, characterized in that, The device includes: a receiving module; The receiving module is used to receive a registration request sent by the first network element. The registration request is used to request the registration of the federated learning capability information of the first network element to the second network element. The federated learning capability information of the first network element includes at least one of the following: the type information of federated learning training supported by the first network element, the time information of the federated learning training supported by the first network element, and the metadata information of the first network element.

40. A model determining device, characterized in that, The device includes: a transmitting module; The sending module is used to send a search request to the second network element. The search request is used to request the search for network elements that can perform federated learning training. The search request includes first information, which includes at least one of the following: data analysis identification 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 of the network element for the federated learning training corresponding to the target task, network element type requirement information, network element quantity requirement information, and data volume requirement information. The network element type requirement information is used to indicate the network element type corresponding to the network element that needs to be searched for that can perform federated learning training. The network element type includes federated learning server network element type and / or federated learning client network element type. The network element quantity requirement information is used to indicate the quantity requirement of the network element that needs to be searched for that can perform federated learning training. The data volume requirement information is used to indicate the data volume requirement of the network element that needs to be searched for that can perform federated learning training.

41. A network element, characterized in that, It includes a processor and a memory, the memory storing a program or instructions that can run on the processor, the program or instructions being executed by the processor to implement the steps of the network element registration method as described in any one of claims 1 to 11.

42. A network element, characterized in that, It includes a processor and a memory, the memory storing a program or instructions that can run on the processor, the program or instructions being executed by the processor to implement the steps of the network element registration method as described in any one of claims 12 to 23.

43. A network element, characterized in that, It includes a processor and a memory, the memory storing a program or instructions that can run on the processor, the program or instructions being executed by the processor to implement the steps of the model determination method as described in any one of claims 24 to 37.

44. A communication system, characterized in that, The communication system includes the network element registration device as described in claim 38, the network element registration device as described in claim 39, and the model determination device as described in claim 40; or... The communication system includes the network elements as described in claims 41 to 43.

45. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the network element registration method as described in any one of claims 1 to 11, or the steps of the network element registration method as described in any one of claims 12 to 23, or the steps of the model determination method as described in any one of claims 24 to 37.

Images