Model acquisition method and related device

Abstract

A model acquisition method and related device, the method comprising: a first network element receiving a model acquisition request from a second network element, the model acquisition request being used for requesting to acquire a model of the first network element, the model acquisition request comprising first indication information; wherein the first indication information is used for indicating the second network element to request to acquire a model generated by the first network element in a transfer learning mode, or indicating the second network element to request to acquire a model meeting a transfer learning requirement in the first network element; and the first network element sends a model acquisition response to the second network element. Therefore, a needed model can be obtained by using a transfer learning mode, and the model training efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of communication technologies, and in particular, to a method for obtaining a model and related devices. Background Art

[0002] For an intelligent network architecture based on a network data analytics function (NWDAF), the purpose is to collect a large amount of information in the network and utilize this large amount of information using existing big data and artificial intelligence technologies to output some valuable information to assist operators in formulating strategies and adjusting network resources, so as to improve user experience, reduce network load, etc. The NWDAF needs to collect data to provide corresponding analysis results. For example, in order to provide a service experience analysis result, the NWDAF needs to collect service-related information such as service identifiers and service experiences from an application function (AF) / user equipment (UE), and collect information such as signal reception power and signal reception quality from operations, administration and management (OAM). The NWDAF trains an artificial intelligence (AI) model based on the collected information, and then obtains an inferred analysis result based on the AI model. For example, it obtains a predicted service experience for a certain future time period.

[0003] NWDAFs can request models from each other. For example, when a consumer NWDAF requests a model from a producer NWDAF, the producer NWDAF may not have the model that the consumer NWDAF wants. In this case, the producer NWDAF still needs to retrain a model according to the requirements of the consumer NWDAF, resulting in low model training efficiency. Summary of the Invention

[0004] This application can provide a method for obtaining a model and related devices, which is beneficial to improving the model training efficiency.

[0005] In a first aspect, the present application provides a method for obtaining a model. This method can be applied to a first network element, or a chip or chip module in the first network element, or a module or unit that can implement all or part of the functions of the first network element, etc. Hereinafter, the first network element will be taken as an example for description. In this method, the first network element receives a model acquisition request from a second network element. The model acquisition request is used to request the acquisition of the model of the first network element, and the first indication information is included in the model acquisition request; wherein, the first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using the transfer learning method, or is used to indicate that the second network element requests to acquire the model in the first network element that meets the requirements of transfer learning; the first network element sends a model acquisition response to the second network element.

[0006] It can be seen that in the case where the first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using the transfer learning method, the first network element can use the transfer learning method to generate the model requested by the second network element, avoiding the problem of low model training efficiency caused by retraining the corresponding model. In the case where the first indication information is used to indicate that the second network element requests to acquire the model in the first network element that meets the requirements of transfer learning, the first network element can provide the model that meets the requirements of transfer learning for the second network element, which is beneficial for the second network element to obtain the required model using the transfer learning method based on this model, thereby improving the model training efficiency.

[0007] Optionally, the first network element may be an NWDAF network element including a model training logical function (MTLF), or an NWDAF network element including MTLF and an analytics logical function (AnLF). Optionally, the second network element is an NWDAF network element service consumer. The second network element may be an NWDAF network element including MTLF, or an NWDAF network element including AnLF, or an NWDAF network element including MTLF and AnLF.

[0008] Optionally, the first indication information is not included in the model acquisition request. The first network element determines to generate the model that meets the model acquisition request using the transfer learning method according to the local configuration, and then returns the model generated using the transfer learning method to the second network element through the model acquisition response.

[0009] Optionally, the model acquisition request includes first indication information, and the first indication information is used to indicate that a second network element requests to acquire a model generated by a first network element using transfer learning. In this case, the first network element can only generate a model using transfer learning. If the first network element is unable to train a model using transfer learning, it will reject the model acquisition request of the second network element, that is, the model acquisition response is used to inform the second network element that the model acquisition request has failed. For example, the situation where the first network element is unable to train a model using transfer learning can be that the first network element finds that the similarity between the training dataset information of all the trained models and the training dataset information in the model acquisition request of the second network element is lower than a threshold.

[0010] Optionally, the model acquisition request includes first indication information, and the first indication information is used to indicate that a second network element requests to acquire a model generated by a first network element using transfer learning. In this case, the first network element first considers generating a model using transfer learning. If the first network element is unable to train a model using transfer learning, it can retrain a model that meets the model acquisition request.

[0011] In an optional implementation, the model acquisition request further includes first training dataset information, and the first training dataset information is the training dataset information that the requested model needs to use. That is to say, the first training dataset information is the training dataset information expected to be used by the second network element. Optionally, the first training dataset information includes the training dataset and / or the description information of the training dataset. Optionally, the first training dataset information is used for the first network element to fine-tune the first model in the first network element, or to calculate the similarity between the training dataset information used by the first model in the first network element. Optionally, if the model acquisition request does not include the first training dataset information, the first network element can further request and obtain the first training dataset information from the second network element.

[0012] In an alternative embodiment, the model acquisition request further includes first threshold information, where the first threshold information is used to indicate the range that the similarity between the training dataset information used by the first model in the first network element and the first training dataset information needs to satisfy; where the first model is a model in the first network element that can use transfer learning to generate the model expected to be acquired by the model acquisition request, or a model in the first network element that meets the transfer learning requirements expected to be acquired by the second network element. That is to say, only a model in the first network element whose similarity between the training dataset information and the first training dataset information is higher than the threshold indicated by the first threshold information or within the threshold range indicated by the first threshold information uses transfer learning to generate the model requested by the model acquisition request. Optionally, if the model acquisition request does not include the first indication information, the first network element can also determine whether to use transfer learning to generate the model requested by the model acquisition request according to the local configuration. For example, a threshold is locally configured, and the first network element uses transfer learning to generate the model requested by the model acquisition request for a model whose similarity between the training dataset information and the first training dataset information is higher than the threshold.

[0013] In an alternative embodiment, the model acquisition request further includes first similarity calculation method information, where the first similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the first model and the first training dataset information. Optionally, if the model acquisition request does not include the first similarity calculation method information, the first network element can determine the similarity calculation method used by itself.

[0014] In an alternative embodiment, the model acquisition request further includes first transfer learning method information, where the first transfer learning method information is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element. For the case where the first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using transfer learning, the model acquisition request may include the first transfer learning method information, such as the transfer learning method based on samples or the transfer learning method based on models, etc.

[0015] In an alternative embodiment, the model acquisition response includes result indication information and a first reason, where the result indication information is used to indicate that the model acquisition request of the second network element fails, and the first reason is used to indicate the reason for the failure of the model acquisition request. Optionally, if the model acquisition response does not include the result indication information and the first reason, it means that the first network element can successfully respond to the model acquisition request of the second network element, that is, the first network element can send the model requested by the second network element to the second network element.

[0016] In another alternative implementation, the model acquisition response includes result indication information, which is used to indicate whether the model acquisition request of the second network element is successful. Optionally, if the result indication information is not included in the model acquisition response, it means that the first network element can successfully respond to the model acquisition request of the second network element, that is, the first network element can send the model requested by the second network element to the second network element.

[0017] In yet another alternative implementation, the model acquisition response includes a first reason, which is used to indicate the reason for the failure of the model acquisition request. Optionally, if the first reason is not included in the model acquisition response, it means that the first network element can successfully respond to the model acquisition request of the second network element, that is, the first network element can send the model requested by the second network element to the second network element.

[0018] In an alternative implementation, the model acquisition response includes second indication information, which is used to indicate that the model fed back by the first network element through the model acquisition response is generated by transfer learning, or is used to indicate that the model fed back by the first network element through the model acquisition response is a model that meets the requirements of transfer learning.

[0019] In an alternative implementation, the model acquisition response further includes second similarity calculation value information, which is used to indicate the calculation result value of the similarity between the training data set information used by the first model and the first training data set information. Among them, the first model is a model in the first network element that can generate the model expected to be obtained by the model acquisition request using the transfer learning method, or is a model in the first network element that meets the requirements of transfer learning expected to be obtained by the second network element.

[0020] In an alternative implementation, the model acquisition response further includes second similarity calculation method information, which is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the first model and the first training data set information.

[0021] In an alternative implementation, the model acquisition response further includes second transfer learning method information, which is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

[0022] In an alternative embodiment, after the first network element receives a model acquisition request from the second network element, the method further includes: the first network element determines a first model according to the model acquisition request; the first network element sends a model retrieval request to the analysis data storage function network element, where the model retrieval request is used to request the analysis data storage function network element to retrieve the first model; the model retrieval request includes third indication information, and the third indication information is used to indicate the training data set information used for the requested first model; the first network element receives a model retrieval response from the analysis data storage function network element, where the model retrieval response is used to feedback the retrieved first model and the training data set information used by the first model; wherein, the first model is a model in the first network element that can generate the model expected to be acquired by the model acquisition request using transfer learning, or a model in the first network element that meets the transfer learning requirements expected to be acquired by the second network element.

[0023] In an alternative embodiment, before the first network element sends a model retrieval request to the analysis data storage function network element, the method further includes: the first network element sends a model or data storage request to the analysis data storage function network element, where the model or data storage request includes the model identifier of the model to be stored, the model file address of the model, and the training data set information used by the model; the first network element receives a model or data storage response from the analysis data storage function network element, where the model or data storage response is used to indicate whether the information requested to be stored by the model or data storage request is successfully stored.

[0024] In an alternative embodiment, the method further includes: the first network element sends a network element registration request to the network storage function network element, where the network element registration request includes fourth indication information, and the fourth indication information is used to indicate whether the first network element supports using the transfer learning method to generate a model; the first network element receives a network element registration response from the network storage function network element, where the network element registration response is used to indicate whether the registration is successful.

[0025] In an alternative embodiment, the network element registration request further includes second training data set information, and the second training data set information is used to indicate the training data set information corresponding to the model supported by the first network element.

[0026] Second aspect, the present application further provides a model acquisition method, which can be applied to a second network element, or a chip or chip module in the second network element, or a module or unit that can implement all or part of the functions of the second network element, etc. Hereinafter, the second network element will be taken as an example for description. In this method, the second network element sends a model acquisition request to the first network element. The model acquisition request is used to request to acquire the model of the first network element, and the model acquisition request includes first indication information; wherein, the first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using the transfer learning method, or is used to indicate that the second network element requests to acquire the model in the first network element that meets the transfer learning requirements; the second network element receives the model acquisition response from the first network element.

[0027] It can be seen that in the case where the first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using the transfer learning method, the first network element can use the transfer learning method to generate the model requested by the second network element, avoiding the problem of low model training efficiency caused by retraining the corresponding model. In the case where the first indication information is used to indicate that the second network element requests to acquire the model in the first network element that meets the transfer learning requirements, the first network element can provide the second network element with a model that meets the transfer learning requirements, which is beneficial for the second network element to obtain the required model using the transfer learning method based on this model, thereby improving the model training efficiency.

[0028] In an optional implementation manner, the model acquisition request further includes first training dataset information, and the first training dataset information is the training dataset information that the requested model is expected to use.

[0029] In an optional implementation manner, the model acquisition request further includes first threshold information, and the first threshold information is used to indicate the range that the similarity between the training dataset information used by the first model in the first network element and the first training dataset information needs to meet; wherein, the first model is the model in the first network element that can use the transfer learning method to generate the model expected to be acquired by the model acquisition request, or is the model in the first network element that meets the transfer learning requirements expected to be acquired by the second network element.

[0030] In an optional implementation manner, the model acquisition request further includes first similarity calculation method information, and the first similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the first model and the first training dataset information.

[0031] In an optional implementation manner, the model acquisition request further includes first transfer learning method information, and the first transfer learning method information is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

[0032] In an optional implementation, the model acquisition response includes result indication information and / or a first reason. The result indication information is used to indicate whether the model acquisition request for the second model training function is successful, and the first reason is used to indicate the reason for the failure of the model acquisition request. Optionally, for the relevant content of the above first aspect, reference may also be made to this implementation, which will not be elaborated here.

[0033] In an optional implementation, the model acquisition response includes second indication information, which is used to indicate that the model fed back by the first network element through the model acquisition response is generated by transfer learning, or is used to indicate that the model fed back by the first network element through the model acquisition response is a model that meets the requirements of transfer learning.

[0034] In an optional implementation, the model acquisition response further includes second similarity calculation value information, which is used to indicate the calculation result value of the similarity between the training data set information used by the first model and the first training data set information.

[0035] In an optional implementation, the model acquisition response further includes second similarity calculation method information, which is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the first model and the first training data set information.

[0036] In an optional implementation, the model acquisition response further includes second transfer learning method information, which is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

[0037] In an optional implementation, before the second network element sends a model acquisition request to the first network element, the method further includes: the second network element sends a network element discovery request to the network storage function network element, and the network element discovery request includes fifth indication information, which is used to indicate a request to discover a model training function network element with transfer learning ability; the second network element receives a network element discovery response from the network storage function network element.

[0038] In an optional implementation, the network element discovery request further includes third training data set information, which is the training data set information that the model expected to be discovered by the model training function network element supports providing.

[0039] In an optional implementation, the network element discovery request further includes third threshold information, which is used to indicate the range that the similarity between the training data set information used by the model in the model training function network element and the third training data set information needs to meet; the model training function network element is the model training function network element that the network element discovery request in the network storage function network element expects to obtain.

[0040] In an alternative embodiment, the network element discovery request further includes third similarity calculation method information, which is used to indicate the similarity calculation method for calculating the similarity between the training data set information used by the model in the model training function network element and the third training data set information.

[0041] In an alternative embodiment, the network element discovery response includes the identifiers of one or more model training function network elements. For each model training function network element identified by an identifier, the network element discovery response may further include one or more of the following information: model identifier, the value of the similarity between the training data set information used by the model and the third training data set information, and the similarity calculation method used to calculate the similarity.

[0042] In a third aspect, the present application further provides a network element discovery method, which can be applied to a network storage function network element, or a chip or chip module in the network storage function network element, or a module or unit that can implement all or part of the functions of the network storage function network element, etc. Hereinafter, the network storage function network element will be taken as an example for description. In this network element discovery method, the network storage function network element receives a network element discovery request from a second network element, where the network element discovery request includes fifth indication information, and the fifth indication information is used to indicate a request to discover a model training function network element with transfer learning ability; the network storage function network element discovers a model training function network element with transfer learning ability, such as the first network element, based on the network element discovery request. The network storage function network element sends a network element discovery response to the second network element, and the network element discovery response includes the identifiers and / or address information of one or more model training function network elements.

[0043] It can be seen that this method enables a consumer network element, such as the second network element being the consumer NWDAF, to discover a model training function network element with or supporting transfer learning ability through the network element storage function network element, which is beneficial for requesting a model obtained through the transfer learning method from the discovered model training function network element, thereby facilitating the improvement of model training efficiency.

[0044] Alternatively, the fifth indication information is used to indicate a request to discover a model training function network element with a model that meets the requirements of transfer learning. In this way, the consumer NWDAF can discover a model that meets the requirements of transfer learning through the network element storage function network element, which is beneficial for using these models to perform fine-tuning through the transfer learning method to obtain the required model, thereby facilitating the improvement of model training efficiency.

[0045] In an alternative embodiment, the network element discovery request further includes third training data set information, which is the training data set information expected to be used by the model supported by the model training function network element to be discovered.

[0046] In an alternative embodiment, the network element discovery request further includes third threshold information, which is used to indicate the range that the similarity between the training data set information used by the model in the model training function network element and the third training data set information needs to satisfy. The model training function network element is the model training function network element that the network element discovery request in the network storage function network element expects to obtain. It can be seen that this embodiment is beneficial for the network storage function network element to return a model training function network element with a relatively high similarity between the training data set information of the model and the third training data set information.

[0047] In an alternative embodiment, the network element discovery request further includes third similarity calculation method information, which is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the model in the model training function network element and the third training data set information.

[0048] In an alternative embodiment, for each model training function network element identified by an identifier and / or an address, the network element discovery response further includes one or more of the following information: the model identifier, the value of the similarity between the training data set information used by the model and the third training data set information, and the similarity calculation method for calculating the similarity. It can be seen that this embodiment is beneficial for the second network element to select a model training function network element as the first network element according to the network element discovery response, so as to send a model acquisition request to the first network element.

[0049] Optionally, the network element discovery method may further include: the network storage function network element receives a network element registration request from the first network element, where the network element registration request includes fourth indication information, which is used to indicate whether the first network element supports using the transfer learning method to generate a model. Further, the network storage function network element sends a network element registration response to the first network element, where the network element registration response is used to indicate whether the registration is successful. Optionally, the network element registration response includes result indication information, which is used to indicate whether the first network element is successfully registered.

[0050] In an alternative embodiment, the network element registration request further includes second training data set information, which is the training data set information of the model that the first network element supports to provide.

[0051] Fourthly, the present application further provides a model retrieval method, which can be applied to an analysis data storage function network element, or a chip or chip module in the analysis data storage function network element, or a module or unit that can implement all or part of the functions of the analysis data storage function network element, etc. Hereinafter, the analysis data storage function network element will be taken as an example for description. In this model retrieval method, the analysis data storage function network element receives a model retrieval request from a first network element, and the model retrieval request is used to request the analysis data storage function network element to retrieve a first model; the first model is a model in the first network element that can generate a model using transfer learning, or a model in the first network element that meets the requirements of transfer learning; the analysis data storage function network element sends a model retrieval response to the first network element, and the model retrieval response is used to feedback the first model.

[0052] Optionally, the model retrieval request further includes third indication information, and the third indication information is used to indicate the training data set information used to request the first model; thus, the model retrieval response further includes the training data set information used by the first model.

[0053] It can be seen that after the first network element deletes the local model and training data set information, this method enables the first network element to obtain the required first model and the training data set information used by the first model through the analysis storage function network element, avoiding the situation that after the first network element deletes the local model and training data set information, it is impossible to distinguish which data are the training data associated with the model, resulting in the inability to determine the source domain of the model (i.e., the training data set information used by the model in the first network element), and the inability to calculate the similarity between the source domain and the target domain of the model (i.e., the requested training data set information, such as the first training data set information, etc.) to determine whether the model meets the requirements of transfer learning.

[0054] Optionally, this method further includes: the analysis data storage function network element receives a model or data storage request from the first network element, and the model or data storage request includes the model identifier of the model to be stored, the model file address of the model, and the training data set information used by the model; the analysis data storage function network element sends a model or data storage response to the first network element, and the model or data storage response is used to indicate whether the information requested to be stored by the model or data storage request is successfully stored.

[0055] Fifth aspect, the present application further provides a model acquisition method, which is described from the perspective of the interaction between the first network element and the second network element. Among them, the model acquisition method includes: the second network element sends a model acquisition request to the first network element, and correspondingly, the first network element receives the model acquisition request. The model acquisition request is used to request to obtain the model of the first network element, and the model acquisition request includes first indication information; the first indication information is used to indicate that the second network element requests to obtain the model generated by the first network element using the transfer learning method, or is used to indicate that the second network element requests to obtain the model in the first network element that meets the transfer learning requirements; the first network element sends a model acquisition response to the second network element.

[0056] It can be seen that in the case where the first indication information is used to indicate that the second network element requests to obtain the model generated by the first network element using the transfer learning method, the first network element can use the transfer learning method to generate the model requested by the second network element, avoiding the problem of low model training efficiency caused by retraining the corresponding model. In the case where the first indication information is used to indicate that the second network element requests to obtain the model in the first network element that meets the transfer learning requirements, the first network element can provide the second network element with a model that meets the transfer learning requirements, which is beneficial for the second network element to obtain the required model using the transfer learning method based on this model, thereby improving the model training efficiency.

[0057] The following elaborates on the optional implementation manners of the parameters that may be included in the model acquisition request and the model acquisition response.

[0058] In an optional implementation manner, the model acquisition request includes first training dataset information, and the first training dataset information is the training dataset information that the requested model is expected to use.

[0059] In an optional implementation manner, the model acquisition request further includes first threshold information, and the first threshold information is used to indicate the range that the similarity between the training dataset information used by the first model in the first network element and the first training dataset information needs to meet; the first model is the model in the first network element that can use the transfer learning method to generate the model expected to be obtained by the model acquisition request, or is the model in the first network element that meets the transfer learning requirements expected to be obtained by the second network element.

[0060] In an optional implementation manner, the model acquisition request further includes first similarity calculation method information, and the first similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the first model and the first training dataset information.

[0061] In an optional implementation manner, the model acquisition request further includes first transfer learning method information, and the first transfer learning method information is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

[0062] In an optional implementation manner, the model acquisition response includes result indication information and a first reason. The result indication information is used to indicate whether the model acquisition request for the second model training function is successful, and the first reason is used to indicate the reason for the failure of the model acquisition request.

[0063] In an optional implementation manner, the model acquisition response includes second indication information, which is used to indicate that the model fed back by the first network element through the model acquisition response is generated by transfer learning, or is used to indicate that the model fed back by the first network element through the model acquisition response is a model that meets the requirements of transfer learning.

[0064] In an optional implementation manner, the model acquisition response further includes second similarity calculation value information, which is used to indicate the calculation result value of the similarity between the training data set information used by the first model and the first training data set information.

[0065] In an optional implementation manner, the model acquisition response further includes second similarity calculation method information, which is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the first model and the first training data set information.

[0066] In an optional implementation manner, the model acquisition response further includes second transfer learning method information, which is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

[0067] Optionally, after receiving the model acquisition request, the first network element determines a first model according to the model acquisition request, where the first model is a model in the first network element that can generate the model expected to be obtained by the model acquisition request in a transfer learning manner, or is a model in the first network element that meets the requirements of transfer learning expected to be obtained by the second network element; if the first network element stores the first model in the analysis data storage function network element and deletes the local model, the first network element may send a model retrieval request to the analysis data storage function network element. Correspondingly, the analysis data storage function network element receives the model retrieval request from the first network element. The model retrieval request is used to request the data storage function network element to retrieve the first model; the model retrieval request includes third indication information, which is used to indicate the training data set information used by the requested first model; the analysis data storage function network element sends a model retrieval response to the first network element. Correspondingly, the first network element receives the model retrieval response, and the model retrieval response is used to feedback the retrieved first model and the training data set information used by the first model. Correspondingly, the first network element may send the first model and the training data set information used by the first model to the second network element through the model acquisition response.

[0068] Optionally, before the first network element sends a model retrieval request to the analysis data storage function network element, the method further includes: the first network element sends a model or data storage request to the analysis data storage function network element, and correspondingly, the analysis data storage function network element receives the model or data storage request. The model or data storage request includes the model identifier of the model to be stored, the model file address of the model, and the training dataset information used by the model. The analysis data storage function network element sends a model or data storage response to the first network element, and correspondingly, the first network element receives the model or data storage response. The model or data storage response is used to indicate whether the information requested to be stored in the model or data storage request is successfully stored.

[0069] Optionally, before the second network element sends a model acquisition request to the first network element, the method further includes: the second network element sends a network element discovery request to the network storage function network element, and correspondingly, the network storage function network element receives the network element discovery request. The network element discovery request includes fifth indication information, and the fifth indication information is used to indicate a request to discover a model training function network element with transfer learning ability. The network storage function network element discovers a model training function network element with transfer learning ability, such as the first network element, based on the network element discovery request. The network storage function network element sends a network element discovery response to the second network element, and correspondingly, the second network element receives the network element discovery response from the network storage function network element. The network element discovery response is returned for the network element discovery request. Optionally, the network element discovery response may include the identifiers of one or more model training function network elements. It can be seen that this embodiment is beneficial for the second network element to select the first network element from one or more model training function network elements, and then send a model acquisition request to the first network element.

[0070] In an optional embodiment, the network element discovery request further includes third training dataset information, and the third training dataset information is the training dataset information that the model training function network element to be discovered is expected to provide for the model.

[0071] In an optional embodiment, the network element discovery request further includes third threshold information, and the third threshold information is used to indicate the range that the similarity between the training dataset information used by the model in the model training function network element and the third training dataset information needs to satisfy. The model training function network element is the model training function network element that the network element discovery request in the network storage function network element expects to obtain. It can be seen that this embodiment is beneficial for the network storage function network element to return a model training function network element with a relatively high similarity between the training dataset information of the model and the third training dataset information.

[0072] In an alternative embodiment, the network element discovery request further includes third similarity calculation method information, which is used to indicate the similarity calculation method for calculating the similarity between the training data set information used by the model in the calculation model training function network element and the third training data set information.

[0073] In an alternative embodiment, the network element discovery response includes the identifiers of one or more model training function network elements. For each model training function network element identified by an identifier, the network element discovery response may further include one or more of the following information: model identifier, the value of the similarity between the training data set information used by the model and the third training data set information, and the similarity calculation method for calculating the similarity. It can be seen that this embodiment is beneficial for the second network element to select a model training function network element as the first network element from the network element discovery response and send a model acquisition request to the first network element.

[0074] Optionally, before the second network element sends a network element discovery request to the network storage function network element, the first network element sends a network element registration request to the network storage function network element. Correspondingly, the network storage function network element receives the network element registration request. The network element registration request includes fourth indication information, which is used to indicate whether the first network element supports using the transfer learning method to generate a model. Further, the network storage function network element sends a network element registration response to the first network element. Correspondingly, the first network element receives the network element registration response from the network storage function network element. The network element registration response is used to indicate whether the registration is successful. Optionally, the network element registration response includes result indication information, which is used to indicate whether the first network element is successfully registered.

[0075] In an alternative embodiment, the network element registration request further includes second training data set information, which is the training data set information of the model supported by the first network element.

[0076] In a sixth aspect, an embodiment of the present application further provides a communication device. The communication device is the first network element, or a device of the first network element, or a device that can be used in matching with the first network element. In a possible implementation, the communication device includes a functional module, which is a hardware circuit, or software, or a combination of a hardware circuit and software.

[0077] In a possible implementation, the communication device includes one or more functional units, such as a communication unit. The communication unit is configured to receive a model acquisition request from a second network element. The model acquisition request is used to request the acquisition of a model of a first network element, and the model acquisition request includes first indication information. The first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using transfer learning, or is used to indicate that the second network element requests to acquire the model in the first network element that meets the requirements of transfer learning. The communication unit is further configured to send a model acquisition response to the second network element.

[0078] Optionally, the communication device may further include a processing unit. The processing unit is configured to generate the model requested by the model acquisition request using transfer learning, or determine the model that meets the requirements of transfer learning, and then the communication unit sends the generated or determined model to the second network element through the model acquisition response.

[0079] Optionally, in this implementation, the relevant content such as the model acquisition request and the model acquisition response, as well as the possible implementations of the communication device, can refer to the relevant descriptions in the first aspect, which will not be elaborated here.

[0080] In a seventh aspect, an embodiment of the present application further provides a communication device. The communication device is a second network element, or a device of the second network element, or a device that can be used in matching with the second network element. In a possible implementation, the communication device includes a functional module, and the functional module is a hardware circuit, or software, or an implementation combining a hardware circuit and software.

[0081] In a possible implementation, the communication device includes one or more functional units, such as a communication unit. The communication unit is configured to send a model acquisition request to a first network element. The model acquisition request is used to request the acquisition of a model of the first network element, and the model acquisition request includes first indication information. The first indication information is used to indicate that the second network element requests to acquire the model generated by the first network element using transfer learning, or is used to indicate that the second network element requests to acquire the model in the first network element that meets the requirements of transfer learning. The communication unit is further configured to receive a model acquisition response from the first network element.

[0082] Optionally, the communication device further includes a processing unit. The processing unit is configured to determine the first network element so that the communication unit can send a model acquisition request to the first network element. Optionally, the processing unit can also generate the required model using transfer learning based on the model returned by the model acquisition response.

[0083] Optionally, in this implementation, the relevant content such as the model acquisition request and the model acquisition response, as well as the possible implementations of the communication device, can refer to the relevant descriptions in the second aspect, which will not be elaborated here.

[0084] In an eighth aspect, an embodiment of the present application further provides a communication device. The communication device is a network storage function network element, or a device of a network storage function network element, or a device that can be used in combination with a network storage function network element. In a possible implementation manner, the communication device includes a functional module, and the functional module is a hardware circuit, or software, or a combination of a hardware circuit and software.

[0085] In a possible implementation manner, the communication device includes one or more functional units, such as a communication unit. The communication unit receives a network element discovery request from a second network element. The network element discovery request includes fifth indication information, and the fifth indication information is used to indicate a request to discover a model training function network element with the ability of transfer learning. The communication unit is further configured to send a network element discovery response to the second network element, and the network element discovery response includes the identifiers of one or more model training function network elements.

[0086] Optionally, the communication device further includes a processing unit, and the processing unit is configured to discover a model training function network element with the ability of transfer learning based on the network element discovery request.

[0087] Optionally, for the possible implementation manners of the communication device, reference may be made to the relevant descriptions in the third aspect, which will not be elaborated here.

[0088] In a ninth aspect, an embodiment of the present application further provides a communication device. The communication device is an analysis data storage function network element, or a device of an analysis data storage function network element, or a device that can be used in combination with an analysis data storage function network element. In a possible implementation manner, the communication device includes a functional module, and the functional module is a hardware circuit, or software, or a combination of a hardware circuit and software.

[0089] In a possible implementation manner, the communication device includes one or more functional units, such as a communication unit. The communication unit is configured to receive a model retrieval request from a first network element. The model retrieval request is used to request the analysis data storage function network element to retrieve a first model. The first model is a model in the first network element that can generate a model using the transfer learning method, or a model in the first network element that meets the requirements of transfer learning. The communication unit is further configured to send a model retrieval response to the first network element, and the model retrieval response is used to feedback the first model.

[0090] Optionally, the communication device further includes a processing unit, and the processing unit is configured to retrieve the first model based on the model retrieval request.

[0091] Optionally, for the possible implementation manners of the communication device, reference may be made to the relevant descriptions in the fourth aspect, which will not be elaborated here.

[0092] For the sixth to ninth aspects, as an example, the processing unit can also be embodied as a processing circuit or a logic circuit; the transceiver unit can be an input / output interface, an interface circuit, an output circuit, an input circuit, a pin, or a related circuit, etc. on the chip or the chip system.

[0093] In the implementation process, the processor can be used for, for example but not limited to, baseband-related processing, and the transceiver or communication interface can be used for, for example but not limited to, radio frequency transceiver. The above-mentioned devices can be respectively arranged on independent chips, or at least partially or entirely arranged on the same chip. For example, the processor can be further divided into an analog baseband processor and a digital baseband processor. Among them, the analog baseband processor can be integrated with the transceiver (or communication interface) on the same chip, and the digital baseband processor can be arranged on an independent chip. With the continuous development of integrated circuit technology, more and more devices can be integrated on the same chip. For example, the digital baseband processor can be integrated with multiple application processors (such as but not limited to a graphics processor, a multimedia processor, etc.) on the same chip. Such a chip can be called a System on a Chip (SoC). Whether to arrange the various devices on different chips independently or to integrate them on one or more chips often depends on the needs of product design. The embodiments of the present application do not limit the implementation forms of the above-mentioned devices.

[0094] In the tenth aspect, the embodiments of the present application further provide a processor for executing the method described in any one of the first to fourth aspects above, or the method described in any possible implementation manner of any one of the first to fourth aspects. In the process of executing these methods, the processes of sending the above-mentioned signal and receiving the above-mentioned signal can be understood as the process of the processor outputting the above-mentioned signal and the process of the above-mentioned signal input by the processor. When outputting the above-mentioned signal, the processor outputs the above-mentioned signal to the transceiver so that the transceiver (or communication interface) can transmit it. After the above-mentioned signal is output by the processor, other processing may be required before it reaches the transceiver (or communication interface). Similarly, when the processor receives the input above-mentioned signal, the transceiver (or communication interface) receives the above-mentioned signal and inputs it to the processor. Further, after the transceiver (or communication interface) receives the above-mentioned signal, the above-mentioned signal may need to be processed further before it is input to the processor.

[0095] For operations such as sending and receiving involved in the processor, if there is no special description, or if it does not conflict with its actual function or internal logic in the relevant description, they can generally be understood as operations such as the processor outputting, receiving, and inputting, rather than the sending and receiving operations directly performed by the radio frequency circuit and the antenna.

[0096] In the implementation process, the above-mentioned processor can be a processor specifically designed to execute these methods, or a processor that executes computer instructions in the memory to execute these methods, such as a general-purpose processor. The above-mentioned memory can be a non-transitory memory, such as a Read Only Memory (ROM), which can be integrated with the processor on the same chip or can be separately provided on different chips. The embodiments of the present application do not limit the type of the memory and the setting manner of the memory and the processor.

[0097] In an eleventh aspect, an embodiment of the present application further provides a communication device, including: a processor, configured to call a computer program stored in a memory, and through a transceiver, enable the communication device to implement the method described in any one of the first aspect to the fourth aspect or any possible implementation manner. Optionally, the communication device further includes a memory, and the processor and the memory are coupled.

[0098] In a twelfth aspect, the present application further provides a communication system, the system includes at least one first network element that executes the method described in the first aspect or any optional implementation manner of the first aspect, and at least one second network element that executes the method described in the second aspect or any possible implementation manner of the second aspect; and / or, the system includes a network storage function network element that executes the third aspect or any implementation manner of the third aspect, and an analysis data storage function network element that executes any implementation manner of the fourth aspect. In another possible design, the system may further include other devices that interact with the first network element and / or the second network element in the solution provided by the embodiment of the present application.

[0099] In a thirteenth aspect, the present application provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and when the computer program is run, enables a computer device to implement the method described in any one of the first aspect to the fourth aspect or any possible implementation manner through a transceiver and a processor.

[0100] In a fourteenth aspect, the present application further provides a computer program product including instructions, the computer program product includes: computer program code, and when the computer program code is run, enables a computer device to implement the method described in any one of the first aspect to the fourth aspect or any possible implementation manner through a transceiver and a processor.

[0101] In a fifteenth aspect, the present application provides a chip system, which includes a processor and an interface. The interface is used to obtain programs or instructions, and the processor is used to call the programs or instructions through the interface to implement the method according to any one of the first aspect to the fourth aspect or any possible implementation manner. In a possible design, the chip system further includes a memory for storing necessary program instructions and data of the terminal. The chip system may be composed of chips or may include chips and other discrete devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 is a schematic diagram of a network architecture of a fifth-generation (5G) communication system based on a service-based interface;

[0103] Figure 2 is a schematic diagram of a traditional machine learning method and a transfer learning method;

[0104] Figure 3 is a schematic flowchart of a model acquisition method provided by an embodiment of the present application;

[0105] Figure 4 is a schematic flowchart of another model acquisition method provided by an embodiment of the present application;

[0106] Figure 5 is a schematic flowchart of a network element registration and discovery method provided by an embodiment of the present application;

[0107] Figure 6 is a schematic flowchart of a model / data storage and retrieval method provided by an embodiment of the present application;

[0108] Figure 7 is a schematic structural diagram of a communication device provided by an embodiment of the present application;

[0109] Figure 8 is a schematic structural diagram of another communication device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0110] Based on the idea of generating a model by means of transfer learning, the present application provides a model acquisition method and related devices, which is beneficial to improving the model training efficiency. Optionally, the present application can be applied to, for example, Figure 1 as shown in a schematic diagram of a network architecture of a fifth-generation (5G) communication system based on a service-based interface, such as Figure 1As shown, the 5G communication system is divided into an access network and a core network. Among them, the access network realizes the functions related to wireless access through radio access network (RAN) devices. The network functions are decomposed modularly, and the decoupled network functions (NFs) can be independently expanded, evolved independently, and deployed on demand. Service-based interfaces are adopted between all NFs on the control plane. The same service can be called by multiple NFs, reducing the coupling degree of the interface definitions between NFs. Finally, the functions of the entire network can be customized on demand, flexibly supporting different service scenarios and requirements. For example Figure 1 In the architecture shown, the network elements in the dotted boxes are service-based NF network elements. The interfaces between NF network elements are service-based interfaces, and the messages exchanged are service-based messages. This architecture can include an access network and a core network. Optionally, it can also include user equipment (UE).

[0111] UE is a device with wireless transceiver functions. It can be deployed on land, including indoor or outdoor, handheld, wearable or vehicle-mounted; it can also be deployed on the water (such as ships, etc.); it can also be deployed in the air (such as airplanes, balloons, satellites, etc.). UE can be a mobile phone, a tablet computer (Pad), a computer with wireless transceiver functions, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal in industrial control, a vehicle-mounted terminal device, a wireless terminal in self-driving, a wireless terminal in remote medical, a wireless terminal in smart grid, a wireless terminal in transportation safety, a wireless terminal in smart city, a wireless terminal in smart home, a wearable terminal device, etc. UE can sometimes also be called a terminal, a terminal device, an access terminal device, a vehicle-mounted terminal, an industrial control terminal, a UE unit, a UE station, a mobile station, a mobile platform, a remote station, a remote terminal device, a mobile device, a UE agent, or a UE device, etc. UE can also be fixed or mobile.

[0112] The access network is used to implement access-related functions, can provide network access functions for authorized users in a specific area, and can determine transmission tunnels of different qualities to transmit user data according to the user level, service requirements, etc. The access network forwards control signals and user data between the terminal device and the core network. The access network may include access network devices. For example, in this application, the information exchanged between the tagging device and the network element of the core network can be forwarded by the access network device.

[0113] The access network device can be a device that provides access for the terminal device and may include RAN devices and AN devices. (R)AN devices are mainly responsible for functions such as radio resource management on the air interface side, quality of service (QoS) management, data compression, and encryption. RAN devices can include various forms of base stations, such as macro base stations, micro base stations (also called small stations), relay stations, access points, balloon stations, etc. In systems using different radio access technologies, the names of the devices with base station functions may be different. For example, in the fifth generation (5G), sixth generation (6G), or even seventh generation (7G) systems, the network device can be called: RAN or next-generation Node basestation (gNB), evolved NodeB (eNB or eNodeB), base station controller (BSC), base transceiver station (BTS), home network device (e.g., home evolved Node B, or home Node B, HNB), baseband unit (BBU), access point (AP) in a wireless fidelity (WIFI) system, wireless relay node, wireless backhaul node, transmission and reception point (TRP), transmission point (TP), etc.; or one or a group of antenna panels of the network device in a 5G system (including multiple antenna panels), or, it can also be a network node that constitutes a gNB or a transmission point, such as a baseband unit (BBU), or a distributed unit (DU), or a road side unit (RSU) in a vehicle to everything (V2X) or intelligent driving scenario.

[0114] In some deployments, a gNB or a transmission point may include a centralized unit (CU) and a DU, etc. A gNB or a transmission point may also include a radio unit (RU). The CU implements some functions of the gNB or the transmission point, and the DU implements some functions of the gNB or the transmission point. For example, the CU implements the functions of the radio resource control (RRC) and the packet data convergence protocol (PDCP) layer, and the DU implements the functions of the radio link control (RLC), the media access control (MAC), and the physical (PHY) layer. Since the information of the RRC layer will ultimately become the information of the physical layer, or is transformed from the information of the physical layer, therefore, in this architecture, high-layer signaling, such as RRC layer signaling or PDCP layer signaling, can also be considered to be sent by the DU, or sent by the DU + RU. It can be understood that the network device can be a CU node, or a DU node, or a device including a CU node and a DU node. Optionally, the network device can also be an auxiliary communication device, such as a satellite.

[0115] The core network is responsible for maintaining the subscription data of the mobile network and providing functions such as session management, mobility management, policy management, and security authentication for the UE. The core network may include the following network elements: a network exposure function (NEF) network element, a network repository function (NRF) network element, a network data analytics function (NWDAF), an analytics data repository function (ADRF), an application function (AF) network element, a policy control function (PCF) network element, etc.

[0116] The NEF network element is mainly used to support the opening of capabilities and events. The NRF network element mainly provides service registration, discovery, and authorization, and maintains information on available network function (NF) instances, enabling on-demand configuration of network functions and services and interconnection between NFs. Among them, service registration means that an NF network element needs to register with the NRF network element before it can provide services. Service discovery means that when an NF network element needs other NF network elements to provide services for it, it needs to first perform service discovery through the NRF network element to discover the desired NF network element that provides services for it. For example, when NF network element 1 needs NF network element 2 to provide services for it, it needs to first perform service discovery through the NRF network element to discover NF network element 2.

[0117] The NWDAF network element has functions such as data collection, model training, data analysis, and model inference. It can be used to collect relevant data from network elements, third-party service servers, terminal devices, or network management systems, perform data analysis or model training based on the relevant data, and provide data analysis results to network elements, third-party service servers, terminal devices, or network management systems, or provide the trained model to other data analysis function network elements. The NWDAF network element can be divided into an analytics logical function (AnLF) and a model training logical function (MTLF) according to its functions. Among them, AnLF is the logical reasoning function in NWDAF, used to perform model inference, derive analysis results (that is, derive statistical or predictive analysis results according to the requests of analysis consumers), and open analysis results; MTLF is the model training function in NWDAF, used to train models and open training services (for example, provide trained models). A NWDAF network element may only contain AnLF or only contain MTLF, or may contain both AnLF and MTLF at the same time.

[0118] The AF network element mainly supports interacting with the 3GPP core network to provide services, such as influencing data routing decisions, policy control functions, or providing some third-party services to the network side. The PCF network element mainly supports providing a unified policy framework to control network behavior, providing policy rules to control layer network functions, and is also responsible for obtaining user subscription information related to policy decisions. The PCF network element can provide policies to the AMF network element and the SMF network element, such as quality of service (QoS) policies, slice selection policies, etc.

[0119] The ADRF network element is used to store the data collected by the NWDAF, the analysis results generated by the NWDAF, and the models trained by the NWDAF. The SMF network element is mainly responsible for session management in the mobile network, such as session establishment, modification, and release. Specific functions include, for example, allocating Internet Protocol (IP) addresses for users and selecting UPF network elements that provide packet forwarding functions. The AMF network element is mainly responsible for mobility management in the mobile network, such as user location updates, user network registration, and user handovers. The UPF network element is mainly responsible for the forwarding and reception of user data. It can receive user data from the data network and transmit it to the UE through the access network device; it can also receive user data from the UE through the access network device and forward it to the data network. Operations, Administration and Management (OAM) is mainly used to monitor and manage the status and performance of network devices.

[0120] In addition, Figure 1 In the communication system network architecture shown, the AN and the AMF can interact through the N2 interface, and the N3 interface supports the selective activation / deactivation of the user plane connection; the SMF network element and the UPF network element interact through the N4 interface. All NFs in the control plane can interact using service-based interfaces. For example, the NEF network element and other network function network elements can interact through the service-based interface Nnef. The NRF network element and other network function network elements can interact through the service-based interface Nnrf. The NWDAF network element and other network function network elements can interact through the service-based interface Nnwdaf. The AF network element and other network function network elements can interact through the service-based interface Naf. The PCF network element and other network function network elements can interact through the service-based interface Npcf. The ADRF network element and other network function network elements can interact through the service-based interface Nadrf. The AMF network element and other network function network elements can interact through the service-based interface Namf. The SMF network element and other network function network elements can interact through the service-based interface Nsmf.

[0121] It should be noted that only the 5G communication system applicable to the communication method provided in the embodiments of the present application is listed above. The communication method provided in the embodiments of the present application can also be applicable to future communication systems such as the sixth generation (6G) and even the seventh generation (7G) systems. The network architecture and service scenarios described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those of ordinary skill in the art can understand that with the evolution of the communication network architecture and the emergence of new service scenarios, the technical solutions provided in the embodiments of the present application are equally applicable to similar technical problems.

[0122] For the convenience of clearly describing the technical solutions of the embodiments of the present application, in the embodiments of the present application, terms such as "first", "second", "third", "fourth", "fifth", "sixth", "seventh", "eighth", etc. are used to distinguish identical or similar items with basically the same functions and roles. Those skilled in the art can understand that terms such as "first" and "second" have no specific technical meanings, do not limit the quantity and execution order, and "first", "second", etc. do not necessarily mean different. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after.

[0123] The intelligent network architecture based on NWDAF aims to collect a vast amount of information in the network and utilize this massive information using existing big data and artificial intelligence technologies to output some valuable information to assist operators in formulating strategies and adjusting network resources, so as to improve the user experience, reduce network load, etc. NWDAF needs to collect data to provide corresponding analysis results. Table 1 is an example of the analysis results provided by NWDAF and the data that needs to be collected. As shown in Table 1, in order to provide business experience analysis results, NWDAF needs to collect service-related information such as service identifiers and service experiences from the application function (AF) / user equipment (UE), and collect information such as signal reception power and signal reception quality from OAM. NWDAF trains an artificial intelligence (AI) model based on the collected information, and then obtains the inferred analysis results based on the AI model. For example, it obtains the predicted business experience analysis for a certain future time period. As shown in Table 1 again, in order to provide network element load analysis results, NWDAF needs to collect relevant information such as network element resource usage from the NRF, and collect information such as UE speed or orientation, such as UE minimization of drivetests (MDT) data from OAM. NWDAF trains an artificial intelligence (AI) model based on the collected information, and then obtains the inferred analysis results based on the AI model. For example, it obtains the predicted network element load analysis for a certain future time period. As shown in Table 1 again, in order to provide mobile edge computing (MEC) business experience analysis results, NWDAF needs to collect relevant information such as UE identifiers, UE locations, application identifiers, application locations, uplink or downlink transmission performance data, etc. from AF or UE. NWDAF trains an artificial intelligence (AI) model based on the collected information, and then obtains the inferred analysis results based on the AI model. For example, it obtains the predicted MEC business experience analysis for a certain future time period.

[0124] Table 1. Example of analysis results provided by NWDAF and data to be collected

[0125]

[0126] In addition, NWDAFs can request models from each other. For example, when the consumer NWDAF requests a model from the producer NWDAF, the producer NWDAF may not have the model that the consumer NWDAF wants. In this way, the producer NWDAF still needs to retrain a model according to the requirements of the consumer NWDAF, resulting in a relatively low model training efficiency.

[0127] Based on the idea of generating a model by means of transfer learning, this application provides a model acquisition method and related devices, which is beneficial to improving the model training efficiency. To facilitate the understanding of this application, some concepts involved in this application are briefly described.

[0128] 1. Definition of transfer learning method

[0129] The transfer learning method is a model generation method different from traditional machine learning methods. In the basic concept of transfer learning, the domain is the main body of learning, which consists of data and the probability distribution that generates these data. For example, a dataset composed of pictures of cats and tigers can be called a domain. The task is the goal of learning. A task usually refers to constructing a prediction function f(·) that generates labels through learning. For example, classifying pictures of cats and tigers can be called a task. Let D S be the source domain, T S be a learning task on the source domain D S , D T be the target domain, T T be a learning task on the target domain D T . Transfer learning means using the knowledge on D S and T S to help the learning of the task T T in D T , where D S ≠D T or T S ≠T T .

[0130] Figure 2 is a schematic diagram of a traditional machine learning method and a transfer learning method. As shown on the left side of Figure 2 for the learning process of traditional machine learning, traditional machine learning can only train and use different machine learning models for different domains / tasks (such as obtaining learning systems respectively as shown in Figure 2 ). However, if transfer learning technology is used, such as the learning process of transfer learning shown on the right side of Figure 2 , the source task in the source domain can be fine-tuned through knowledge to obtain a model (such as a learning system), which can be applied to the target task in the target domain, greatly improving the model training efficiency. At the same time, it can also generate a model with better performance when the target domain is small. For example, task A (source task) is to classify cats and tigers in pictures, and task B (target task) is to distinguish the lengths of cats and tigers in pictures. Traditional machine learning methods can only train two different models for these two completely different tasks respectively, but using transfer learning, the model of task A can be fully utilized and fine-tuned on the basis of the task A model to obtain a model applicable to task B.

[0131] 2. Classification of Transfer Learning Methods

[0132] Transfer learning can be classified into the following four categories according to the learning method:

[0133] (1) Sample-based transfer learning method: Find data similar to the target domain in the source domain, adjust the weights of this data so that the data after weight adjustment matches the data in the target domain, and then perform training and learning to obtain a model applicable to the target domain. This method is simple and easy to implement, but the selection of weights and the measurement of similarity rely on experience, and the data distributions of the source domain and the target domain are often different.

[0134] (2) Feature-based transfer learning method: When the source domain and the target domain contain some common cross features, through feature transformation, the features of the source domain and the target domain can be transformed into the same space, so that the data in the source domain and the data in the target domain in this space have an approximate data distribution, and then traditional machine learning is performed. This method can be adapted to most situations, but it is necessary to screen out good common features, which is difficult to solve and prone to overfitting.

[0135] (3) Model-based transfer learning method: Share some model parameters between the source domain and the target domain, and transfer the model trained with a large amount of data in the source domain to the target domain for prediction. This method can make full use of the similarity between models, but the model parameters are not easy to converge.

[0136] (4) Relationship-based transfer learning method: When two domains are similar, they will share a certain similarity relationship, and apply the logical network relationship learned in the source domain to the target domain for transfer. For example, transfer the law of biological virus transmission to the law of computer virus transmission.

[0137] Please refer to Figure 3 , Figure 3 which is a schematic flowchart of a model acquisition method provided by an embodiment of the present application. Figure 3 In the described model acquisition method, the first network element and the second network element are respectively described by taking the first NWDAF and the second NWDAF as examples. As Figure 3 shown, the model acquisition method may include but is not limited to the following steps:

[0138] S101. The second NWDAF sends a model acquisition request to the first NWDAF. Correspondingly, the first NWDAF receives the model acquisition request.

[0139] Optionally, the model acquisition request may be a model subscription request, such as NWDAF_MLModelProvision_Subscribe, that is, the second NWDAF may subscribe to the model provided by the first NWDAF through the model subscription request, or request the first NWDAF to train a machine learning (ML) model for the second NWDAF.

[0140] In an optional implementation manner, the model acquisition request may include one or more of the following parameters:

[0141] (1) One or more Analytics ID(s) for identifying the type of analysis for which the requested model is used. For example, if the Analytics ID in the model acquisition request is an identifier for the type of analysis for Service Experience, then it means that the ML model requested by the second NWDAF is used for business experience analysis.

[0142] (2) Machine learning model filter information, such as single network slice selection assistance information (S-NSSAI), Area of Interest, indicating that the requested model is for a specific slice or a specific area.

[0143] (3) Target of ML Model Reporting, such as specific UEs, a group of UE(s) or any UE (i.e., all UEs), indicating that the model requested by the second NWDAF is for a specific UE, or for a group of UEs, or for all UEs within the scope identified by the model filter information (such as within a slice or within an area).

[0144] (4) The first indication information is used to indicate that the second NWDAF requests to obtain the model generated by the first NWDAF using transfer learning, such as the second NWDAF requests the first NWDAF to generate the model it requests using transfer learning. Optionally, the first indication information can be the Transfer Learning indication. Optionally, if the model acquisition request does not include this first indication information, the first NWDAF can also determine to use transfer learning to generate a model that meets the requirements of the model acquisition request of the second NWDAF according to local configuration; if this first indication information is included, it is used to indicate that the first NWDAF can only use transfer learning to generate the model, which means that if the first NWDAF cannot train the model using transfer learning (for example, the first NWDAF finds that the similarity between the training dataset information of all trained models and the dataset of the consumer is lower than the threshold), it will reject the request of the consumer; if this parameter is included, it can also be used to indicate that the first NWDAF can use transfer learning to generate the model, which means that after receiving this indication, the first NWDAF first considers using transfer learning to generate the model. If the first NWDAF finds that it cannot train the model using transfer learning (for example, the first NWDAF finds that the similarity between the training dataset information of all trained models and the training dataset information requested by the second NWDAF is lower than the threshold), it can also retrain a model that meets the requirements of the second NWDAF.

[0145] (5) The first training dataset information is used to indicate the training dataset information that the model requested by the second NWDAF expects to use. The first training dataset information may contain the training dataset, or contain the training dataset description information, or may also contain both the training dataset and the training dataset description information. In a possible implementation manner, the training dataset information is for each model of each analysis type, that is, at the granularity of each model ID per Analytics ID. This parameter can be used to calculate the similarity between the source domain and the target domain, and can also be used for the first NWDAF to fine-tune the model. If the second NWDAF does not include this parameter, the first NWDAF needs to further request and obtain this information from the second NWDAF.

[0146] (6) The first threshold information is used to indicate the range that the similarity between the training dataset information used by the first model in the first NWDAF and the first training dataset information needs to satisfy; wherein, the first model is the model in the first NWDAF that can use transfer learning to generate the model expected to be obtained by the model acquisition request. Among them, the range indicated by the first threshold information can specify a lower limit value, or specify an upper limit value, or specify both a lower limit value and an upper limit value. The model acquisition request includes the first threshold information, which can indicate that the value of the similarity between the training dataset information in the first NWDAF and the first training dataset information needs to satisfy the range indicated by the first threshold information to use transfer learning to generate the model requested by the model acquisition request. Optionally, if the first threshold information is not included in the model acquisition request, the first NWDAF can determine whether it can use transfer learning to generate the model requested by the model acquisition request according to the local configuration. For example, the first NWDAF can configure a range locally, and the value of the similarity between the training dataset information in the first NWDAF and the first training dataset information needs to satisfy this range to use transfer learning to generate the model requested by the model acquisition request.

[0147] (7) The first similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the first model and the first training dataset information. Optionally, if the first similarity calculation method information is not included in the model acquisition request, the first NWDAF determines the similarity calculation method to be used by itself.

[0148] (8) The first transfer learning method information is used to indicate the transfer learning method used by the first NWDAF to generate the model requested by the second NWDAF. Optionally, if the first transfer learning method information is not included in the model acquisition request, the first NWDAF determines the transfer learning method to be used by itself.

[0149] In another alternative embodiment, the model acquisition request may include one or more of the following parameters: Analytics ID(s), ML model filter information, Target of ML Model Reporting, model information (such as ML Model ID(s), ML model address(es), etc.), first indication information, first training dataset information, and first transfer learning method information. Among them, the definitions of Analytics ID(s), ML model filter information, and Target of ML Model Reporting can refer to the relevant descriptions in the above embodiments and will not be elaborated here. Among them, the model identified by Model ID(s) is provided by the second NWDAF to the first NWDAF, and is used to instruct the first NWDAF to perform transfer learning using this model to generate a model that meets the requirements of the second NWDAF. The first indication information is used to instruct the first NWDAF to generate a model required by the second NWDAF using the transfer learning method. The first training dataset information is the training dataset information expected to be used by the model identified by Model ID(s). The first transfer learning method is used to instruct the transfer learning method used by the first NWDAF. Optionally, in this embodiment, Analytics ID(s) and model information are mandatory parameters, and the remaining parameters can be optional parameters.

[0150] Among them, the training dataset information may include the training dataset and / or the description information of the training dataset as described above. Among them, the description information of the training dataset may include, but is not limited to, the following parameters:

[0151] 1) ML Model Filter Information, such as S-NSSAI(s), Area(s) of Interest, which is used to describe from which slices (S-NSSAI(s)) or which areas (Area(s) of Interest) the training dataset is collected;

[0152] 2) Target of ML Model Reporting, such as specific UEs, a group of UE(s) or any UE, which is used to describe from which UE, which group of UEs the training dataset is collected, or the training dataset is collected from all UEs within the specified range of the filter information;

[0153] 3) Data sources, such as a list of NF instance (or NF set) IDs and corresponding NF Type, which are used to describe from which NF instances the training dataset is collected, and the NF Type corresponding to the NF instance can also be specified. For example, this parameter can specify that the training dataset is collected from AMF1, or this parameter can specify that the training dataset is collected from SMF1, or the parameter can specify that the training dataset is collected from AMF2, SMF2, and AF1;

[0154] 4) Event ID(s) per NF instance ID, which are used to describe through which Event ID(s) the training dataset is collected from the corresponding NF instance. In a possible implementation, the Event ID(s) are at the NF instance ID granularity, that is, for the data source specified for each NF instance ID, the information on which Event ID(s) the data is collected from is given. For example, <AMF1, Event ID = Location Report> indicates that the data (i.e., location data) is collected from AMF1 through the Location Report event, and another example <SMF2, Event ID = QFI allocation> indicates that the data (i.e., information such as QFI, 5QI, DNN, S-NSSAI, etc.) is collected from SMF2 through the QFI allocation event;

[0155] 5) Other filter information, such as App ID, which is used to describe the data related to which application the dataset is collected for. This filter information may be at the Event ID granularity, such as collecting data related to a specific application through a specific Event ID;

[0156] 6) The timestamp (a specific moment) when the data is collected, or the time period for collecting the data (per Event ID), and this parameter may also be at the Event ID granularity.

[0157] 7) Data metrics, such as a sampling ratio, the maximum number of input values for collecting data, and / or the maximum time interval between samples of this input data, the data range including maximum and minimum values, data distribution information such as mean, standard deviation, and data distribution, etc.

[0158] S102. The first NWDAF determines a model acquisition response according to the model acquisition request.

[0159] Optionally, the model acquisition response can be a model subscription notification, such as Nnwdaf_MLModelProvision_Notify.

[0160] S103. The first NWDAF sends the model acquisition response to the second NWDAF. Correspondingly, the second NWDAF receives the model acquisition response.

[0161] In an alternative implementation, in the case where the model acquisition request does not include model identifier(s) (Model ID(s)), the first NWDAF determines a model acquisition response based on the model acquisition request, including: the first NWDAF determines, based on information such as MLModel Filter Information and Target of ML Model Reporting in the model acquisition request, whether there is a trained model that can meet the requirements of the second NWDAF (i.e., whether there is a trained model that meets the requirements of the model requested by the model acquisition request); if so, the corresponding model is directly returned through the model acquisition response, where the model acquisition response includes one or more ML model information, and each ML model information includes an ML model identifier (ML Model identifier) and an ML model file address information (such as the uniform resource locator (URL) or fully qualified domain name (FQDN) of the model file), or each ML model information includes an ML model identifier (ML Model identifier) and an ADRF(Set)ID information. If the ML model information contains the ML model file address information, the second NWDAF can download the model based on the ML model file address information; if the ML model information contains the ADRF(Set)ID information, the second NWDAF obtains the model from the ADRF identified by the ADRF(Set)ID based on the ML model identifier.If there is no pre-trained model that can meet the requirements of the second NWDAF, the first NWDAF calculates the similarity between the training dataset information of the existing pre-trained models (such as Model 1 and Model 2) and the training dataset information requested by the second NWDAF (i.e., the first training dataset information included in the model acquisition request); if there is a model with a calculated similarity higher than the threshold (for example, the similarity between the training dataset information of Model 1 and the training dataset information requested by the second NWDAF is higher than the threshold), the first NWDAF fine-tunes the model (such as Model 1) based on the pre-trained model (such as Model 1) and the training dataset information requested by the second NWDAF to obtain a model that meets the requirements of the second NWDAF, that is, a model that meets the requirements of the model requested in the model acquisition request, and then feedbacks the model through the model acquisition response, such as the ML model information of the fine-tuned model is included in the model acquisition response; if the similarity between the training dataset information of all the pre-trained models of the first NWDAF and the training dataset information requested by the second NWDAF is lower than the threshold, the first NWDAF needs to re-train a new model and feedback the model through the model acquisition response, or the first NWDAF can reject the model acquisition request of the second NWDAF, that is, inform the second NWDAF that the model acquisition request fails through the model acquisition response.

[0162] In another alternative implementation, when the model acquisition request includes the model identifier (Model ID(s)) (such as the second NWDAF includes the Model ID(s) information in the model acquisition request, that is, the model identified by the Model ID(s) information is a pre-trained model of the first NWDAF and the model meets the requirements of transfer learning), the first NWDAF determines the model acquisition response according to the model acquisition request, including: the first NWDAF directly fine-tunes the model using transfer learning based on the model and the training dataset information provided by the second NWDAF to obtain a model that meets the requirements of the second NWDAF. Among them, the ML model information of the fine-tuned model is included in the model acquisition response.

[0163] Optionally, the first NWDAF can determine a first model according to a model acquisition request. For example, the first model can be a trained model of the first NWDAF that meets the requirements of the second NWDAF, or a model retrained by the first NWDAF that meets the requirements of the second NWDAF, or a model fine-tuned by the first NWDAF. Then, the ML model information in the model acquisition response is a mandatory parameter. Optionally, the model acquisition response may further include, but is not limited to, one or more of the following optional parameters: (1) result indication information and / or a first reason. The result indication information is used to indicate that the model acquisition request of the second NWDAF fails, and the first reason is used to indicate the reason for the failure of the model acquisition request. Optionally, if the model acquisition response does not include the result indication information and the first reason, it means that the first NWDAF can successfully respond to the model acquisition request of the second NWDAF. Optionally, the model acquisition response includes result indication information, and the result indication information is used to indicate whether the model acquisition request of the second NWDAF is successful. Optionally, if the model acquisition response does not include the result indication information, it means that the first NWDAF can successfully respond to the model acquisition request of the second NWDAF. Optionally, the model acquisition response includes a first reason, and the first reason is used to indicate the reason for the failure of the model acquisition request. Optionally, if the model acquisition response does not include the first reason, it means that the first NWDAF can successfully respond to the model acquisition request of the second NWDAF; (2) second indication information (such as Transfer Learning indication), and the second indication information is used to indicate that the model fed back by the first NWDAF through the model acquisition response is generated by transfer learning; (3) second similarity calculation value information, and the second similarity calculation value information is used to indicate the calculation result value (or calculation value) of the similarity between the training data set information used by the first model and the first training data set information. Among them, the first model is a model in the first NWDAF that can generate the model expected to be obtained by the model acquisition request using the transfer learning method. (4) second similarity calculation method information, and the second similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the first model and the first training data set information. (5) second transfer learning method information, and the second transfer learning method information is used to indicate the transfer learning method used by the first NWDAF to generate the model requested by the second NWDAF to obtain.

[0164] It can be seen that Figure 3In the model acquisition method shown, first indication information and related parameters can be added to the model acquisition request. Thus, when the second NWDAF acquires a model, by including the first indication information and related parameters in the model acquisition request, the first NWDAF can use transfer learning to generate the required model and return the transferred model through the model acquisition response. Therefore, it is beneficial to improve the efficiency of model training. In addition, second indication information and related parameters can be added to the model acquisition response, which can also enable the second NWDAF to know whether the acquired model is obtained through transfer learning.

[0165] Please refer to Figure 4 , Figure 4 which is a schematic flowchart of another model acquisition method provided by an embodiment of the present application. Among them, Figure 4 the model acquisition method shown and Figure 3 the difference between the model acquisition method shown lies in that Figure 4 in the model acquisition method described, in the model acquisition request, the first indication information is used to instruct the second NWDAF to request to acquire a model that meets the requirements of transfer learning in the first NWDAF. The following describes this embodiment in conjunction with the accompanying drawings. As Figure 4 shown, the model acquisition method may include but is not limited to the following steps:

[0166] S201. The second NWDAF sends a model acquisition request to the first NWDAF. Correspondingly, the first NWDAF receives the model acquisition request.

[0167] Optionally, the model acquisition request may be a model subscription request, that is, the second NWDAF may subscribe to the models supported by the first NWDAF through the model subscription request.

[0168] In an alternative embodiment, the model acquisition request may include but is not limited to one or more of the following parameters: (1) Analytics ID(s), which has the same definition as in Figure 3 the embodiment described above and will not be elaborated here. (2) MLmodel filter information, which has the same definition as in Figure 3 the embodiment described above and will not be elaborated here. (3) Target of ML Model Reporting, which has the same definition as in Figure 3The definitions in the foregoing embodiments are the same and will not be elaborated herein. (4) First indication information. The first indication information is used to indicate that the second NWDAF requests to obtain a model in the first NWDAF that meets the requirements of transfer learning. For the convenience of description, the model in the first NWDAF that meets the requirements of transfer learning is simply referred to as the model to be transferred. The second NWDAF can use the model to be transferred and local data for fine-tuning to obtain the final model. Optionally, if the model acquisition request includes the first indication information, it can be used to indicate that the first NWDAF can only provide the model to be transferred, which means that if the first NWDAF cannot provide the model to be transferred (for example, the first NWDAF finds that the similarity between the training dataset information of all trained models and the dataset of the consumer is lower than the threshold), the model acquisition request of the second NWDAF will be rejected. Optionally, if the model acquisition request includes the first indication information, it can be used to indicate that the first NWDAF can provide the model to be transferred, which means that after receiving this indication, the first NWDAF first considers providing the model to be transferred. When the first NWDAF finds that it can provide the model to be transferred (for example, the first NWDAF finds that the similarity between the training dataset information of all trained models and the training dataset information requested by the second NWDAF is lower than the threshold, where the training dataset information requested by the second NWDAF can be provided through the model acquisition request, such as the first training dataset information described below), it can also retrain a model that meets the request of the second NWDAF. (5) First training dataset information, which is used to indicate the training dataset information that the model requested by the second NWDAF expects to use. This parameter is related to Figure 3The definitions are the same as above and will not be elaborated here. (6) First threshold information, which is used to indicate the range that the similarity between the training dataset information used by the first model in the first NWDAF and the first training dataset information needs to meet; wherein, the first model is the model that meets the requirements of transfer learning and is expected to be obtained by the second NWDAF in the first NWDAF, that is, the model to be transferred. Among them, the range indicated by the first threshold information can specify a lower limit value, or specify an upper limit value, or specify both a lower limit value and an upper limit value. The model acquisition request includes the first threshold information, which can indicate that the value of the similarity between the training dataset information in the first NWDAF and the first training dataset information needs to meet the range indicated by the first threshold information to be used as the model to be transferred. Optionally, if the first threshold information is not included in the model acquisition request, the first NWDAF can determine whether the trained model can be used as the model to be transferred according to the local configuration. For example, the first NWDAF can configure a range locally, and only the model whose similarity value between the training dataset information in the first NWDAF and the first training dataset information meets this range can be fed back to the second NWDAF as the model to be transferred. (7) First similarity calculation method information, which is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the first model and the first training dataset information. Optionally, if the first similarity calculation method information is not included in the model acquisition request, the first NWDAF determines the similarity calculation method to be used by itself.

[0169] S202. The first NWDAF determines a model acquisition response according to the model acquisition request.

[0170] S203. The first NWDAF sends the model acquisition response to the second NWDAF. Correspondingly, the second NWDAF receives the model acquisition response.

[0171] The first NWDAF determines a model acquisition response according to a model acquisition request, including: the first NWDAF determines, according to information such as ML Model Filter Information and Target of ML Model Reporting in the model acquisition request, whether there is a trained model that can meet the requirements of the second NWDAF (that is, whether there is a trained model that meets the requirements of the model requested by the model acquisition request); if so, directly return the corresponding model through the model acquisition response, where the model acquisition response includes ML model information. If there is no trained model that can meet the requirements of the second NWDAF, the first NWDAF calculates the similarity between the training dataset information of the existing trained models (such as Model 1 and Model 2) and the training dataset information requested by the second NWDAF (that is, the first training dataset information included in the model acquisition request); if there is a model with a calculated similarity higher than the threshold (for example, the similarity between the training dataset information of Model 1 and the training dataset information requested by the second NWDAF is higher than the threshold), the first NWDAF takes it as the model to be migrated and feedbacks the model to be migrated through the model acquisition response, such as the model acquisition response includes the ML model information of the model to be migrated; if the similarity between the training dataset information of all the trained models of the first NWDAF and the training dataset information requested by the second NWDAF is lower than the threshold, the first NWDAF needs to retrain a new model and feedback the model through the model acquisition response, or the first NWDAF can reject the model acquisition request of the second NWDAF, that is, inform the second NWDAF that the model acquisition request fails through the model acquisition response.

[0172] Optionally, the first NWDAF can determine a first model according to a model acquisition request. For example, the first model can be a trained model of the first NWDAF that meets the requirements of the second NWDAF, or a model retrained by the first NWDAF that meets the requirements of the second NWDAF, or a model to be migrated determined by the first NWDAF. Then, the ML model information in the model acquisition response is a mandatory parameter. Optionally, the model acquisition response may further include, but is not limited to, one or more of the following optional parameters: (1) Result indication information and / or a first reason. The result indication information is used to indicate that the model acquisition request of the second NWDAF fails, and the first reason is used to indicate the reason for the failure of the model acquisition request. Optionally, if the result indication information and the first reason are not included in the model acquisition response, it means that the first NWDAF can successfully respond to the model acquisition request of the second NWDAF. Optionally, the model acquisition response includes result indication information, and the result indication information is used to indicate whether the model acquisition request of the second NWDAF is successful. Optionally, if the result indication information is not included in the model acquisition response, it means that the first NWDAF can successfully respond to the model acquisition request of the second NWDAF. Optionally, the model acquisition response includes a first reason, and the first reason is used to indicate the reason for the failure of the model acquisition request. Optionally, if the first reason is not included in the model acquisition response, it means that the first NWDAF can successfully respond to the model acquisition request of the second NWDAF; (2) Second indication information (such as Transfer Learning indication), and the second indication information is used to indicate that the model fed back by the first NWDAF through the model acquisition response is a model that meets the requirements of transfer learning; 3) Second similarity calculation value information, and the second similarity calculation value information is used to indicate the calculation result value (or calculated value) of the similarity between the training data set information used by the first model and the first training data set information. Among them, the first model is a model to be migrated determined by the first NWDAF. 4) Second similarity calculation method information, and the second similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the first model and the first training data set information.

[0173] Optionally, if the first indication information is included in the model acquisition request and the model fed back by the model acquisition response is a model to be migrated, then, as Figure 3 shown, the method further includes the following steps:

[0174] S204. The second NWDAF uses the model to be migrated fed back by the model acquisition response and performs fine-tuning in a transfer learning manner to obtain the required model.

[0175] Among them, since the second NWDAF needs to fine-tune the model, or rather, needs to retrain the model, the second NWDAF is a NWDAF with the MTLF function.

[0176] It can be seen that Figure 4 in the model acquisition method shown, first indication information and related parameters can be added to the model acquisition request. Thus, when the second NWDAF acquires a model, by including the first indication information and related parameters in the model acquisition request, the first NWDAF can be enabled to provide the model to be migrated and return the model to be migrated through the model acquisition response. In this way, the second NWDAF can fine-tune the model to be migrated to obtain the required model, which is conducive to improving the efficiency of model training. In addition, second indication information and related parameters can be added to the model acquisition response, which can also enable the second NWDAF to know that the acquired model is a model to be migrated.

[0177] In another embodiment, the present application further provides a network element registration and discovery method. In this method, the first NWDAF can register migration learning ability information with the NRF, so that the second NWDAF can discover the first NWDAF that supports migration learning ability through the NRF, and then be able to send a model acquisition request to the first NWDAF. Please refer to Figure 5 , Figure 5 which is a schematic flowchart of a network element registration and discovery method provided by an embodiment of the present application. As Figure 5 shown, the network element registration and discovery method may include but is not limited to the following steps:

[0178] S301. The first NWDAF sends a network element registration request to the NRF. The network element registration request includes fourth indication information, and the fourth indication information is used to indicate whether the first NWDAF supports generating a model using the migration learning method.

[0179] Alternatively, the fourth indication information is used to indicate whether the first NWDAF has migration learning ability.

[0180] Optionally, the network element registration request is used to register the ability information of the first NWDAF, and the network element registration request may be an NRF_NF Management_NF Register Request (Nnrf_NFManagement_NFRegister Request).

[0181] The fourth indication information is used to indicate whether the first NWDAF has the ability of transfer learning. The network element registration request may further include an Analytics ID, which represents the analysis identifier supported by the first NWDAF, such as the identifier of the service experience analysis as described above, indicating that the first NWDAF supports providing a model for service experience analysis. Optionally, the network element registration request further includes one or more of the following parameters: (1) The second training data set information. If the first NWDAF supports transfer learning, optionally, register the training data set information. The third training data set information may be at the per Model ID per Analytics ID granularity, that is, the first NWDAF registers the training data set information of the model for each trained model. Among them, the training data set information may include the training data set and / or the description information of the training data set. For the description information of the training data set, reference may be made to the foregoing, and details are not described herein again. (2) The NWDAF Serving Area information, for example, it may be an area composed of one or more tracking areas (TAs) and / or cells. (3) The ML model Filter information (such as S-NSSAI(s) and Area(s) of Interest) for the trained ML model(s) per Analytics ID corresponding to each analysis identifier, such as S-NSSAI(s) or Area(s) of Interest, indicating the applicable scope of the NWDAF trained model. (4) The ML Model Interoperability indicator per Analytics ID corresponding to each analysis identifier, which is composed of a list of Vendor ID(s), indicating that the NWDAF supports providing the model to the manufacturers identified by these Vendor IDs. (5) The NF Set ID and / or NF Type of the NF data sources of the network function data source.

[0182] S302. The NRF sends a network element registration response to the first NWDAF. Correspondingly, the first NWDAF receives the network element registration response, and the network element registration response is used to indicate whether the registration is successful.

[0183] Optionally, the network element registration response may be an NRF_NF Management_NF Register Response (Nnrf_NFManagement_NFRegister Response). Among them, the network element registration response may include result indication information to indicate whether the registration is successful.

[0184] S303. The second NWDAF sends a network element discovery request (Discovery Request) to the NRF. Correspondingly, the NRF receives the network element discovery request. Among them, the network element discovery request includes fifth indication information, and the fifth indication information is used to indicate a request to discover MTLF network elements with migration learning capabilities.

[0185] Among them, the network element discovery request also includes an Analytics ID, which represents the analysis identifier requested by the second NWDAF, such as the identifier of the service experience analysis described above, indicating that the second NWDAF requests to support the provision of MTLF for models used in service experience analysis. Optionally, the network element discovery request may also include one or more of the following parameters: (1) Third training dataset information, which is the training dataset information that the model supported by the MTLF requested to be discovered is expected to use. (2) Third threshold information, which is used to indicate the range that the similarity between the training dataset information used by the model in the MTLF requested to be discovered and the third training dataset information needs to meet. When the network element discovery request includes the third training dataset information, the NRF will compare the third training dataset information with the training dataset information registered by the MTLF (i.e., the second training dataset information), such as calculating the similarity between the two. The NRF will feedback MTLF instances with a similarity higher than the threshold to the second NWDAF. For example, assuming that the similarity between the second training dataset information registered by the first NWDAF and the third training dataset information requested by the second NWDAF is higher than the threshold requested by the second NWDAF, the NRF may feedback the first NWDAF as the requested MTLF. (3) Third similarity calculation method information, which is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the model in the MTLF (i.e., the training dataset information registered by the MTLF) and the third training dataset information. For example, if the training dataset information registered by the first NWDAF is training dataset A and the third training dataset information is training dataset B, then the similarity between the training datasets can be calculated by some methods. For example, it is the normalized distance between the training dataset matrices A and B; if the training dataset information registered by the first NWDAF and the third training dataset information both also include the description information of the training dataset, then the overlap degree of the description information of the training dataset can also be calculated and normalized with the similarity between the training datasets to obtain the similarity value.

[0186] S304. The NRF discovers a MTLF network element with transfer learning capabilities, such as the first NWDAF, based on this network element discovery request.

[0187] Optionally, step S304 is: The NRF discovers a MTLF (or NWDAF) with transfer learning capabilities and the highest similarity or similarity higher than a threshold between the registered training dataset information and the third training dataset information based on the network element discovery request.

[0188] S305. The NRF sends a Network Element Discovery Response to the second NWDAF. Correspondingly, the second NWDAF receives this network element discovery response.

[0189] Among them, the network element discovery response is returned for the network element discovery request. The network element discovery response includes the Analytics ID requested by the second NWDAF and one or more NWDAF instances (such as the identifiers of one or more model training function network elements) that meet the second NWDAF network element discovery request. Optionally, the network element discovery response may also include the identifiers of one or more model training function network elements. For each model training function network element identified by the identifier, the network element discovery response may also include one or more of the following information: model identifier, the value of the similarity between the training dataset information used by the model and the third training dataset information, and the similarity calculation method for calculating the similarity (a list of <Model ID, Domain similarity, Similarity measurement method>). If the network element discovery request includes the third threshold information, for a specific Analytics ID and a specific NWDAF instance, there may be multiple models for this NWDAF (or MTLF) that can meet the similarity requirements of the third threshold information. For each model that meets the requirements, the NRF provides the Model ID of this model, the calculated similarity value corresponding to this Model ID, and the method used to calculate the similarity. Optionally, in addition to providing the Model ID, the network element discovery response may also provide the address information of the model file (e.g., URL or FQDN) or the identifier of the analytics data storage network element (set) storing this model (ADRF(Set)ID).

[0190] S306. The second NWDAF selects one NWDAF from the network element discovery response, such as the first NWDAF.

[0191] S307. The second NWDAF sends a model acquisition request to the first NWDAF.

[0192] S308. The first NWDAF returns a model acquisition response to the second NWDAF.

[0193] Among them, the second NWDAF can, according to Figure 3 the embodiments described above, obtain the model after transfer learning from the first NWDAF, or can also, according to Figure 4 the embodiments described above, obtain the model to be transferred from the first NWDAF, and then, fine-tune the model in a transfer learning manner to obtain a model that meets the requirements. Details are not described herein again.

[0194] It can be seen that in the embodiments of the present application, the NWDAF registers transfer learning ability information with the NRF, so that the consumer can discover the MTLF that supports the transfer learning ability through the NRF, avoiding the situation that the NWDAF does not register information related to the transfer learning ability, resulting in the MTLF discovered by the consumer not necessarily having the transfer learning ability and being unable to train the model using the transfer learning method, thereby facilitating the improvement of the model training efficiency. The embodiments of the present application enable the consumer, such as the second NWDAF, to directly discover the MTLF with transfer learning ability, such as the first NWDAF. In addition, the NWDAF can also register the training dataset information of the trained model, so that the directly discovered MTLF is the MTLF of a model that not only has the transfer learning ability but also meets the transfer learning requirements (such as similarity requirements).

[0195] In another embodiment, if the NRF provides the Model ID that meets the transfer learning requirements to the second NWDAF in the network element discovery response, the second NWDAF can directly obtain the ML Model Information from the first NWDAF according to the Model ID (wherein, the network element discovery response includes the address information of the model file (e.g., URL or FQDN) or the ADRF (Set) ID storing the model), and then, download the model based on the ML Model Information, or obtain the model from the ADRF, and fine-tune the model using the transfer learning method to obtain a model that meets the requirements. If the NRF provides the Model ID that meets the transfer learning requirements to the second NWDAF in the network element discovery response and provides the ML Model Information at the same time, the second NWDAF can directly download the model based on the ML Model Information, or obtain the model from the ADRF, and fine-tune the model using the transfer learning method to obtain a model that meets the requirements.

[0196] In another embodiment, the present application also provides a method for model / data storage and retrieval. In this method, the first NWDAF can store the model / data information in the ADRF, so that after the first NWDAF deletes the local model and training data, it can also retrieve the required model through the ADRF. Please refer to Figure 6, Figure 6 is a schematic flowchart of a method for model / data storage and retrieval provided by an embodiment of the present application. As Figure 6 shown, the method for model / data storage and retrieval may include but is not limited to the following steps:

[0197] S401. The first NWDAF sends a model or data storage request to the ADRF. Correspondingly, the ADRF receives the model or data storage request.

[0198] Optionally, the model or data storage request may be an ADRF_ML model management_storage request (Nadrf_MLModelManagement_StorageRequest Request).

[0199] Among them, the model or data storage request includes the model identifier of the model to be stored, the model file address of the model, and the training dataset information used by the model (list of <Model ID, address (e.g., URL or FQDN) of Model file, Training dataset Info>). Optionally, the model or data storage request further includes the NF instance identifier (NF instance ID of the NWDAF containing MTLF) of the NWDAF containing MTLF, such as the identifier of the first NWDAF, and the analytics identifier (Analytics ID).

[0200] S402. The first NWDAF receives a model or data storage response from the ADRF. The model or data storage response is used to indicate whether the information requested to be stored by the model or data storage request is successfully stored.

[0201] Optionally, the model or data storage response may be an ADRF_ML model management_storage request response (Nadrf_MLModelManagement_StorageRequest Response).

[0202] S403. The second NWDAF sends a model acquisition request to the first NWDAF. Correspondingly, the first NWDAF receives the model acquisition request.

[0203] Among them, for the description of step S403, reference may be made to the relevant content of the above Figure 3 , Figure 4 described embodiment, which will not be elaborated here.

[0204] S404. The first NWDAF determines a first model according to the model acquisition request;

[0205] Among them, the first model is asFigure 3 In the embodiment, the first NWDAF can use transfer learning to generate a model to obtain the model expected to be obtained by the model acquisition request, or as Figure 4 In the embodiment, the model in the first NWDAF that meets the requirements of transfer learning expected to be obtained by the second NWDAF, that is, the model to be transferred.

[0206] The first NWDAF determines that the trained model is Figure 3 or Figure 4 After the model obtained by the model acquisition request, if the first NWDAF stores the trained model in the ADRF and deletes the local model, the first NWDAF can execute step S405.

[0207] S405. The first NWDAF sends a model retrieval request to the ADRF. Correspondingly, the ADRF receives the model retrieval request. The model retrieval request is used to request the ADRF to retrieve the first model; wherein, the model retrieval request includes third indication information, and the third indication information is used to indicate the training dataset information used for requesting the first model.

[0208] Optionally, the model retrieval request can be an ADRF_ML model management_retrieval request (Nadrf_MLModelManagement_Retrieval Request).

[0209] Wherein, the model retrieval request includes the following parameters: Storage Transaction Identifier or one or more tuples of unique ML Model identifier(s), that is, the first NWDAF can retrieve the stored model from the ADRF through the Storage Transaction Identifier, or can also retrieve one or more models through the ML Model ID(s). In addition, the model retrieval request can also include the Training dataset Info Indication parameter, which is used to indicate that when obtaining the model, the training dataset information corresponding to the model needs to be obtained.

[0210] S406. The ADRF returns a model retrieval response to the first NWDAF. Correspondingly, the first NWDAF receives the model retrieval response. The model retrieval response is used to feedback the retrieved first model and the training dataset information used by the first model.

[0211] Optionally, the model retrieval response can be an ADRF_ML model management_retrieval response (Nadrf_MLModelManagement_Retrieval Response).

[0212] Among them, the model retrieval response also includes the following parameters: one or more tuples of unique ML Model identifiers and address (e.g., URL or FQDN) of Model file stored in ADRF.

[0213] S407. The first NWDAF sends a model acquisition response to the second NWDAF. Correspondingly, the second NWDAF receives the model acquisition response.

[0214] Among them, the model acquisition response may include the model information of the first model.

[0215] Among them, for the elaboration of step S407, reference can be made to the relevant content of the above Figure 3 、 Figure 4 The relevant content of the described embodiments will not be elaborated here.

[0216] It can be seen that in the embodiment of the present application, by saving the correspondence between the ML model and the model training dataset information in the ADRF, when the first NWDAF retrieves the model from the ADRF, it can also retrieve the corresponding model training dataset information. The first NWDAF can determine whether the model meets the requirements of transfer learning based on the model training dataset information and the training dataset information provided by the second NWDAF.

[0217] In the above embodiments provided by the present application, the solutions of the various methods provided by the embodiments of the present application are introduced from the perspective of each node itself and the interaction between each node. It can be understood that each node, such as the above-mentioned first node, second node, etc., in order to implement the above functions, includes the corresponding hardware structure and / or software unit for executing each function. Those skilled in the art should easily realize that, combining the units and algorithm steps of each example described in the embodiments disclosed in the present application, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a certain function is executed in the way of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0218] Please refer to Figure 7 , Figure 7 which is a schematic structural diagram of a communication device provided by an embodiment of the present application. Figure 7 The shown communication device includes a transceiver module 701 and a processing module 702.

[0219] In a design of this embodiment, the communication device is the first NWDAF or a related device in the first NWDAF:

[0220] Exemplarily, a transceiver module 701 is configured to receive a model acquisition request from a second NWDAF. The model acquisition request is used to request to acquire a model of the first NWDAF, and the model acquisition request includes first indication information. The first indication information is used to indicate that the second NWDAF requests to acquire a model generated by the first network element using transfer learning, or is used to indicate that the second NWDAF requests to acquire a model in the first NWDAF that meets the transfer learning requirements. The transceiver module 701 is further configured to send a model acquisition response to the second NWDAF.

[0221] Optionally, the communication device may further include a processing module 702. The processing module 702 is configured to generate the model requested by the model acquisition request using transfer learning, or determine a model that meets the transfer learning requirements, and then the transceiver module 701 sends the generated or determined model to the second NWDAF through the model acquisition response.

[0222] Optionally, when the communication device is the first NWDAF or a related device of the first NWDAF, it is used to implement Figures 1 to 6 the functions of the first NWDAF and optional implementation manners in the illustrated embodiment.

[0223] In a design of this embodiment, the communication device is the second NWDAF or a related device in the second NWDAF:

[0224] The transceiver module 701 is configured to send a model acquisition request to a first network element. The model acquisition request is used to request to acquire a model of the first network element, and the model acquisition request includes first indication information. The first indication information is used to indicate that the second network element requests to acquire a model generated by the first network element using transfer learning, or is used to indicate that the second network element requests to acquire a model in the first network element that meets the transfer learning requirements. The transceiver module 701 is further configured to receive a model acquisition response from the first network element.

[0225] Optionally, the communication device further includes a processing module 702. The processing module 702 is configured to determine the first NWDAF, so that the communication unit sends a model acquisition request to the first NWDAF. Optionally, the processing module 702 may further generate a required model using transfer learning based on the model returned in the model acquisition response.

[0226] Optionally, when the communication device is the second NWDAF or a related device of the second NWDAF, it is used to implement Figures 1 to 6 the functions of the second NWDAF and optional implementation manners in the illustrated embodiment.

[0227] Please refer toFigure 8 , Figure 8 is a schematic structural diagram of another communication device provided by an embodiment of the present application. Figure 8 The shown communication device includes at least one processor 801 and a memory 802. Optionally, a transceiver 803 may also be included. In the embodiments of the present application, the specific connection medium between the above-mentioned processor 801 and the memory 802 is not limited. Figure 8 Taking the connection between the memory 802 and the processor 801 through a bus 804 as an example, the bus 804 is represented by a thick line in the figure. The connection manners between other components are only for illustrative purposes and are not limited thereto. The bus 804 may be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 8 only one thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0228] The processor 801 may have a data transceiver function and be capable of communicating with other devices. In the device shown in Figure 8 , an independent data transceiver module, such as a transceiver 803, may also be provided for receiving and transmitting data; when the processor 801 communicates with other devices, data transmission may be performed through the transceiver 803.

[0229] In one example, when the first NWDAF adopts the Figure 8 shown form, Figure 8 the processor 801 in Figures 1 to 6 may execute the method performed by the first NWDAF in any one of the embodiments in

[0230] In one example, when the second NWDAF adopts the Figure 8 shown form, Figure 8 the processor 801 in Figures 1 to 6 may execute the method performed by the second NWDAF in any one of the embodiments in

[0231] The embodiments of the present application also provide a communication system, which may include Figures 1 to 6 the first NWDAF in

[0232] and at least two second NWDAFs, and specific reference may be made to the method embodiments described above. Optionally, the system may also include the above-mentioned NRF and / or ADRF, etc., and specific reference may be made to the method embodiments described above.The solutions described in this application can be implemented in various ways. For example, these technologies can be implemented in hardware, software, or a combination of hardware and software. For hardware implementation, the processing modules for executing these technologies at a communication device (e.g., a base station, a terminal, a network entity, or a chip) can be implemented in one or more general-purpose processors, digital signal processors (DSPs), digital signal processing devices, application-specific integrated circuits (ASICs), programmable logic devices, field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination thereof. The general-purpose processor can be a microprocessor. Optionally, the general-purpose processor can also be any conventional processor, controller, microcontroller, or state machine. The processor can also be implemented by a combination of computing devices, such as a digital signal processor and a microprocessor, multiple microprocessors, one or more microprocessors combined with a digital signal processor core, or any other similar configuration.

[0233] It can be understood that the memory in the embodiments of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include but not be limited to these and any other suitable types of memory.

[0234] The present application also provides a computer-readable medium, on which instructions are stored, and when the instructions are executed by a computer, the computer device realizes the functions of any one of the above method embodiments through a transceiver and a processor.

[0235] The present application also provides a computer program product, and when the computer program product is executed by a computer, the computer device realizes the functions of any one of the above method embodiments through a transceiver and a processor.

[0236] The present application also provides a processing device, which can be a product such as a chip device, and the processing device is used to determine Figures 1 to 6The relevant information required in the embodiments described in any of the figures. For example, the processing device is used to determine the model acquisition response. For another example, the processing device is used to fine-tune the model to be migrated through transfer learning to obtain the required model, etc. Optionally, the processing device may be a baseband processing module, and the information determined by the processing device may be sent out through the radio frequency processing module. For example, the migrated model or the model to be migrated. Or, receive information sent by other devices through the radio frequency processing module, and the processing device determines other information based on this information.

[0237] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that the computer can access, or a data storage device such as a server or data center that includes one or more integrated available media. The available media may be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as high-definition digital video discs (DVDs)), or semiconductor media (such as solid state disks (SSDs)), etc.

[0238] It can be understood that some optional features in the embodiments of the present application can, in some scenarios, be implemented independently without relying on other features, such as the current scheme they are based on, to solve the corresponding technical problems and achieve the corresponding effects. In some scenarios, they can also be combined with other features according to requirements. Correspondingly, the devices given in the embodiments of the present application can also implement these features or functions accordingly, which will not be elaborated here.

[0239] Those skilled in the art can also understand that the various illustrative logical blocks and steps listed in the embodiments of the present application can be implemented by electronic hardware, computer software, or a combination of both. Whether such a function is implemented by hardware or software depends on the specific application and the design requirements of the entire system. For the corresponding application, those skilled in the art can use various methods to implement the described function, but such implementation should not be construed as exceeding the scope protected by the embodiments of the present application.

[0240] It can be understood that the "embodiments" mentioned throughout the specification mean that the specific features, structures, or characteristics related to the embodiments are included in at least one embodiment of the present application. Therefore, the various embodiments throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics can be combined in one or more embodiments in any suitable manner. It can be understood that in the various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not mean the sequence of execution, and the execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0241] It can be understood that in the present application, "when", "if", and "in case" all refer to the situation where the device will perform corresponding processing under certain objective circumstances, not to limit time, and do not require the device to have a judgment action during implementation, nor does it mean there are other limitations.

[0242] In the present application, elements represented in the singular are intended to mean "one or more", rather than "one and only one", unless otherwise specified. In the present application, unless otherwise specified, "at least one" is intended to mean "one or more", and "a plurality" is intended to mean "two or more". In the written description of the present application, "including at least one of A, B, and C" can mean: including A; including B; including C; including A and B; including A and C; including B and C; including A, B, and C. In the written description of the present application, "and / or" is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. Here, A can be singular or plural, and B can be singular or plural.

[0243] It can be understood that the various numbers involved in the embodiments of the present application are only for the convenience of description and are not used to limit the scope of the embodiments of the present application. The magnitudes of the serial numbers of the above processes do not mean the sequence of execution, and the execution sequence of each process should be determined by its function and internal logic. In addition, the terms "system" and "network" are often used interchangeably in this article.

[0244] The predefined in this application can be understood as definition, predefined, stored, prestored, pre-negotiated, preconfigured, solidified, or pre-fired.

[0245] Those of ordinary skill in the art can understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0246] The same or similar parts among the various embodiments of this application can be referred to each other. In the various embodiments of this application, as well as in each implementation manner / implementation method / realization method in each embodiment, if there is no special explanation and logical conflict, the terms and / or descriptions between different embodiments, as well as between each implementation manner / implementation method / realization method in each embodiment, are consistent and can be mutually referred to. The technical features in different embodiments, as well as in each implementation manner / implementation method / realization method in each embodiment, can be combined according to their internal logical relationships to form new embodiments, implementation manners, implementation methods, or realization methods. The implementation manners of this application described above do not constitute a limitation on the protection scope of this application.

[0247] As described above, the above are only the specific implementation manners of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application.

Claims

1. A method for obtaining a model, characterized in that, the method includes: A first network element receives a model acquisition request from a second network element, where the model acquisition request is used to request to obtain the model of the first network element, and the model acquisition request includes first indication information; wherein, the first indication information is used to indicate that the second network element requests to obtain the model generated by the first network element using transfer learning, or is used to indicate that the second network element requests to obtain the model in the first network element that meets the requirements of transfer learning; The first network element sends a model acquisition response to the second network element.

2. A method for obtaining a model, characterized in that, the method includes: A second network element sends a model acquisition request to a first network element, where the model acquisition request is used to request to obtain the model of the first network element, and the model acquisition request includes first indication information; wherein, the first indication information is used to indicate that the second network element requests to obtain the model generated by the first network element using transfer learning, or is used to indicate that the second network element requests to obtain the model in the first network element that meets the requirements of transfer learning; The second network element receives a model acquisition response from the first network element.

3. The method according to claim 1 or 2, characterized in that, the model acquisition request further includes first training dataset information, where the first training dataset information is the training dataset information that the requested model is expected to use.

4. The method according to claim 3, characterized in that, the model acquisition request further includes first threshold information, where the first threshold information is used to indicate the range that the similarity between the training dataset information used by the first model in the first network element and the first training dataset information needs to meet; wherein, the first model is the model in the first network element that can generate the model expected to be obtained by the model acquisition request using transfer learning, or is the model in the first network element that meets the requirements of transfer learning expected to be obtained by the second network element.

5. The method according to claim 4, characterized in that, the model acquisition request further includes first similarity calculation method information, where the first similarity calculation method information is used to indicate the similarity calculation method for calculating the similarity between the training dataset information used by the first model and the first training dataset information.

6. The method according to any one of claims 1 to 5, characterized in that, the model acquisition request further includes first transfer learning method information, where the first transfer learning method information is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

7. The method according to any one of claims 1 to 6, characterized in that, the model acquisition response includes result indication information and / or a first reason, where the result indication information is used to indicate that the model acquisition request of the second network element fails, and the first reason is used to indicate the reason for the failure of the model acquisition request.

8. The method according to any one of claims 1 to 7, characterized in that, the model acquisition response includes second indication information, The second indication information is used to indicate that the model through which the first network element obtains a response feedback is generated by transfer learning, or is used to indicate that the model through which the first network element obtains a response feedback is a model that meets the requirements of transfer learning.

9. The method according to claim 4 or 5, wherein, the response obtained by the model further includes second similarity calculation value information, the second similarity calculation value information is used to indicate the calculation result value of the similarity between the training data set information used by the first model and the first training data set information.

10. The method according to claim 9, wherein, the response obtained by the model further includes second similarity calculation method information, the second similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training data set information used by the first model and the first training data set information.

11. The method according to any one of claims 1 to 10, wherein, the response obtained by the model further includes second transfer learning method information, the second transfer learning method information is used to indicate the transfer learning method used by the first network element to generate the model requested by the second network element.

12. The method according to claim 1, wherein, after the first network element receives a model acquisition request from a second network element, the method further includes: the first network element determines a first model according to the model acquisition request; the first network element sends a model retrieval request to an analysis data storage function network element, the model retrieval request is used to request the analysis data storage function network element to retrieve the first model; the model retrieval request includes third indication information, and the third indication information is used to indicate the training data set information used to request the first model; the first network element receives a model retrieval response from the analysis data storage function network element, and the model retrieval response is used to feedback the retrieved first model and the training data set information used by the first model; wherein, the first model is a model in the first network element that can generate the model expected to be obtained by the model acquisition request in a transfer learning manner, or is a model in the first network element that meets the transfer learning requirements expected to be obtained by the second network element.

13. The method according to claim 12, wherein, before the first network element sends a model retrieval request to an analysis data storage function network element, the method further includes: the first network element sends a model or data storage request to the analysis data storage function network element, and the model or data storage request includes the model identifier of the model to be stored, the model file address of the model, and the training data set information used by the model; the first network element receives a model or data storage response from the analysis data storage function network element, and the model or data storage response is used to indicate whether the information requested to be stored by the model or data storage request is successfully stored.

14. The method according to claim 1, or 12, or 13, wherein, the method further includes: The first network element sends a network element registration request to the network storage function network element. The network element registration request includes fourth indication information, and the fourth indication information is used to indicate whether the first network element supports using the transfer learning method to generate a model; The first network element receives a network element registration response from the network storage function network element, and the network element registration response is used to indicate whether the registration is successful.

15. The method according to claim 14, wherein, the network element registration request further includes second training dataset information, and the second training dataset information is the training dataset information of the model supported by the first network element for providing.

16. The method according to claim 2, wherein, before the second network element sends a model acquisition request to the first network element, the method further includes: The second network element sends a network element discovery request to the analysis data storage function network element. The network element discovery request includes fifth indication information, and the fifth indication information is used to indicate a request to discover a model training function network element with transfer learning ability; The second network element receives a network element discovery response from the analysis data storage function network element.

17. The method according to claim 16, wherein, the network element discovery request further includes third training dataset information, and the third training dataset information is the training dataset information that the model training function network element to be discovered is expected to use for the model.

18. The method according to claim 17, wherein, the network element discovery request further includes third threshold information, and the third threshold information is used to indicate the range that the similarity between the training dataset information used by the model in the model training function network element and the third training dataset information needs to meet; The model training function network element is the model training function network element that the network element discovery request expects to obtain in the analysis data storage function network element.

19. The method according to claim 18, wherein, the network element discovery request further includes third similarity calculation method information, and the third similarity calculation method information is used to indicate the similarity calculation method used to calculate the similarity between the training dataset information used by the model in the model training function network element and the third training dataset information.

20. The method according to any one of claims 17 to 19, wherein, the network element discovery response includes the identifiers of one or more model training function network elements. For each model training function network element identified by an identifier, the network element discovery response may further include one or more of the following information: model identifier, the value of the similarity between the training dataset information used by the model and the third training dataset information, and the similarity calculation method for calculating the similarity.

21. A communication system, wherein, the system includes: at least one first network element for executing the method according to claim 1, or any one of claims 3 to 15; and at least one second network element for executing the method according to claim 2 to 11, or any one of claims 16 to 20.

22. A communication device, wherein, Comprising a processor and a transceiver, the processor invokes a computer program stored in a memory through the transceiver, so that the communication device implements the method according to any one of claims 1, or 3 to 15, or implements the method according to any one of claims 2 to 11, or 16 to 20.

23. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, and when the computer program is run, it enables a computer device to implement the method according to any one of claims 1, or 3 to 15, or implement the method according to any one of claims 2 to 11, or 16 to 20 through a transceiver and a processor.

Images