Network service method and system and related network elements

Abstract

This disclosure proposes a network service method and system, as well as related network elements, relating to the field of communications. The network service method includes: a Model Storage Function (MSF) network element storing model file information trained by a Network Data Analysis Function (NWDAF) network element; the MSF network element responding to model requests related to analysis services from model-demanding network elements, determining whether the stored existing models meet the model requests of the model-demanding network elements; and the MSF network element providing the target model file corresponding to the model request to the model-demanding network element based on the model file information that meets the model request. By storing and managing models using the added model storage function network element, models can be shared, avoiding resource waste caused by repeated model training.

Description

Technical Field

[0001] This disclosure relates to the field of communications, and in particular to a network service method, a network service system, a model storage function network element, a model requirement network element, and a network data analysis function network element, as well as other related network elements. Background Technology

[0002] One or more Network Data Analytics Function (NWDAF) network elements can be deployed in a public terrestrial mobile network. When multiple NWDAF network elements exist, each NWDAF network element does not need to provide support for all analytics service types, but can provide support for some analytics service types. The analytics service types supported by each NWDAF network element can be identified by an Analytics ID.

[0003] Each NWDAF element can execute an Analysis Logic Function (AnLF) and / or a Model Training Logic Function (MTLF). NWDAF elements with AnLF primarily perform inference functions and feed the final analysis results back to subscribers. NWDAF elements with MTLF primarily perform training functions, training the corresponding model based on the analysis ID and then registering the model with the Network Repository Function (NRF) to support discovery, subscription, and requesting within MTLF-enabled NWDAF elements.

[0004] Research has revealed that current standards do not support model sharing among multiple NWDAF network elements with MTLF functionality. This leads to redundant model training and wasted resources. For example, an NWDAF network element in province A may have generated a model for a specific analysis service, but an NWDAF network element in province B may not be able to access this model and would need to repeat training for the same type of analysis service, resulting in wasted resources. This situation may also occur between private networks providing the same service but located in different areas. Summary of the Invention

[0005] This disclosure enhances the management of network data analysis function models by mobile networks. Based on the added model storage function network element, the model is stored and managed, enabling the model to be shared. For example, the model can be shared among multiple network data analysis function network elements, between different regions, or between different private networks. This achieves the effect of training the model once and calling it multiple times, avoiding the waste of resources caused by repeated model training.

[0006] This disclosure provides a network service method through some embodiments, including:

[0007] Model storage function: MSF network element stores network data; NWDAF network element training model file information;

[0008] MSF network element response to model requirements: The network element proposes an analysis service related model request and determines whether the stored existing model meets the model requirements of the network element's model request.

[0009] MSF network elements provide the target model file corresponding to the model request to the model requesting network elements based on the model file information that meets the model request.

[0010] In some embodiments, the MSF network element stores model file information trained by the NWDAF network element, including:

[0011] The MSF network element receives a model storage request sent by the NWDAF network element, which includes description information of the model file to be stored;

[0012] The MSF network element decides whether to accept the model storage request and sends the corresponding model storage response to the NWDAF network element.

[0013] If the MSF network element decides to accept the model storage request, it downloads the corresponding model file from the NWDAF network element according to the model file address in the description information;

[0014] MSF network elements store model files and their description information as model file information.

[0015] In some embodiments, the MSF network element determines whether the stored existing model meets the model requirements of the network element's model request. The model request of the network element includes: the MSF network element determines whether the existing model meets the model requirements of the network element based on the analysis service type of the stored existing model and in combination with at least one of model performance, model validity period, and model regionality.

[0016] In some embodiments, the MSF network element provides the target model file corresponding to the model request to the model request network element based on the model file information that satisfies the model request, including: the MSF network element selects the model file with the best model performance from the model files that satisfy the model request and provides it to the model request network element as the target model file.

[0017] In some embodiments, the MSF network element provides the target model file corresponding to the model request to the model requesting network element, including:

[0018] MSF network element sends the model file address of the target model file to the network element that requires the model;

[0019] MSF network element receives model download requests initiated by the model requesting network element based on the model file address of the target model file;

[0020] MSF network elements send the target model file to the network element that requires the model through the model download request response.

[0021] In some embodiments, if there is no model file information to satisfy the model request, the MSF network element returns a model request failure response to the model requesting network element.

[0022] The NWDAF network element receives the model request related to the analysis service sent by the model request failure response from the model request network element, which carries an indication of whether the model request to the MSF network element has failed.

[0023] NWDAF network element responds to model requests related to analysis services submitted by network elements that require model services, and provides the target model file corresponding to the model request to the network elements that require model services.

[0024] According to the instructions in the instruction information, if the model requesting network element fails to request a model from the MSF network element, the NWDAF network element initiates a process to store the target model file information into the MSF network element.

[0025] In some embodiments, the model file information includes a model file and its description information. The description information of the model file includes the analysis service type, and also includes at least one of the following: model performance, model validity period, model regionality, and model file address.

[0026] This disclosure provides a network service method through some embodiments, including:

[0027] Submit a model request related to the analysis service to the MSF network element;

[0028] Receive either a successful model request response or a failed model request response from the MSF network element;

[0029] Based on the model file address of the target model file in the successful model request response, initiate the model download process of the target model file to the MSF network element;

[0030] Based on the model request failure response, the system sends a model request related to the analysis service to the NWDAF network element and receives the target model file corresponding to the model request returned by the NWDAF network element. The model request carries an indication of whether the model request to the MSF network element has failed, so as to instruct the NWDAF network element to initiate the process of storing the target model file information to the MSF network element when the model requesting network element fails to request the model from the MSF network element.

[0031] This disclosure provides some embodiments of a model storage function network element, including:

[0032] The storage module is configured to store model file information for training network elements in the NWDAF network data analysis function;

[0033] The judgment module is configured to respond to model requests related to analysis services raised by network elements that require model services, and to determine whether the existing stored models meet the model requests of the network elements that require model services.

[0034] The module is configured to provide the target model file corresponding to the model request to the network element that requires the model, based on the model file information that satisfies the model request.

[0035] This disclosure provides some embodiments of a network element with network data analysis functionality, including:

[0036] The receiving module is configured to receive model requests related to analysis services sent by the model request failure response from the model request network element based on the MSF network element, which carries indication information as to whether the request to the MSF network element for the model has failed.

[0037] The first providing module is configured to respond to model requests related to analysis services raised by model requesting network elements and provide the target model file corresponding to the model request to the model requesting network elements.

[0038] The second providing module is configured to, according to the instructions in the instruction information, initiate a process of storing the target model file information into the MSF network element when the model requesting network element fails to request a model from the MSF network element.

[0039] This disclosure proposes a model requirement network element in some embodiments, including:

[0040] The request module is configured to submit model requests related to analysis services to MSF network elements;

[0041] The receiving module is configured to receive either a successful model request response or a failed model request response sent by the MSF network element.

[0042] The first model acquisition module is configured to initiate a model download process for the target model file to the MSF network element based on the model file address of the target model file in the successful response to the model request.

[0043] The second model acquisition module is configured to send a model request related to the analysis service to the NWDAF network element based on the model request failure response, and receive the target model file corresponding to the model request returned by the NWDAF network element. The model request carries an indication of whether the model request to the MSF network element has failed, so as to instruct the NWDAF network element to initiate the process of storing the target model file information to the MSF network element when the model requesting network element fails to request the model from the MSF network element.

[0044] Some embodiments of this disclosure propose a model storage function network element, including: a memory; and a processor coupled to the memory, the processor being configured to execute a network service method based on instructions stored in the memory.

[0045] Some embodiments of this disclosure propose a model requirement network element, including: a memory; and a processor coupled to the memory, the processor being configured to execute a network service method based on instructions stored in the memory.

[0046] This disclosure provides some embodiments of a network service system, including: a model storage function network element, a model demand network element, and a network data analysis function network element.

[0047] Some embodiments of this disclosure propose a non-transitory computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of a network service method. Attached Figure Description

[0048] The accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. This disclosure can be more clearly understood from the following detailed description with reference to the accompanying drawings.

[0049] Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0050] Figure 1 The diagram illustrates the service architecture of the relevant technologies.

[0051] Figure 2 This diagram illustrates a service architecture for adding MSF network elements and service-oriented interfaces (Nmsf) according to some embodiments of this disclosure.

[0052] Figure 3 This diagram illustrates a process for storing model file information for NWDAF network element training in MSF network elements, according to some embodiments of this disclosure.

[0053] Figure 4 The diagram illustrates a flowchart of a network service method based on MSF network elements according to some embodiments of this disclosure.

[0054] Figure 5 This diagram illustrates a method for an MSF network element to provide network services for an NWDAF with AnLF functionality, according to some embodiments of this disclosure.

[0055] Figure 6This diagram illustrates a method for an MSF network element, according to some embodiments of the present disclosure, to provide network services for a model of an NWDAF with AnLF functionality located in different regions.

[0056] Figure 7 A schematic diagram of a model storage function network element 700 according to some embodiments of the present disclosure is shown.

[0057] Figure 8 A schematic diagram of a network data analysis function network element 800 according to some embodiments of the present disclosure is shown.

[0058] Figure 9 A schematic diagram illustrating the model requirement network element 900 of some embodiments of this disclosure is shown.

[0059] Figure 10 A schematic diagram of a model storage function network element 1000 according to some embodiments of the present disclosure is shown.

[0060] Figure 11 A schematic diagram illustrating the model requirements of network element 1100 according to some embodiments of this disclosure is shown.

[0061] Figure 12 A schematic diagram of a network service system 1200 according to some embodiments of the present disclosure is shown. Detailed Implementation

[0062] The technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0063] Unless otherwise stated, the terms "first," "second," etc., used in this disclosure are used to distinguish different objects and are not used to indicate size or sequence.

[0064] According to relevant technologies, such as Figure 1 As shown, an NWDAF network element with AnLF functionality and an NWDAF network element with MTLF functionality are connected via an Nnwdaf interface. The NWDAF network element with AnLF functionality obtains the trained model from the NWDAF network element with MTLF functionality through the Nnwdaf interface. This results in duplicate model training, leading to a waste of resources.

[0065] To enhance the management of network data analysis function models by mobile networks and enable model sharing among multiple NWDAF network elements, a Model Storage Function (MSF) network element and a service interface Nmsf are added. The Nmsf interface enables the interaction between MSF and other network functions (NFs).

[0066] Figure 2This diagram illustrates the service architecture of adding MSF network elements and the service-oriented interface Nmsf. (For example...) Figure 2 As shown, in a service-oriented architecture, there are originally NSSF (Network Slice Selection Function), NEF (Network Exposure Function), NRF (NF Repository Function), PCF (Policy Control Function), UDM (Unified Data Management), AF (Application Function), NSSAAF (Network Slice-Specific Authentication and Authorization Function), AUSF (Authentication Server Function), AMF (Access and Mobility Management Function), SMF (Session Management Function), SCP (Service Communication Proxy), NSACF (Network Slice Admission Control Function), NWDAF (Network Data Analytics Function), and UPF (User Plane Function). Network elements such as Function (user plane function) and corresponding interfaces, such as Nnssf, Nnef, Nnrf, Npcf, Nudm, etc., are added on this basis, and MSF network element and service interface Nmsf are added. Figure 2 It also shows AN (Access Network) or RAN (Radio Access Network), DN (Data Network), UE (User Equipment), etc.

[0067] MSF network elements can support functions including but not limited to the following.

[0068] (1) Store information about some or all of the trained model files in the network. Model file information includes, but is not limited to, the model files and their descriptions as shown in Table 1. The model file descriptions include the analysis service type, model performance, model validity period, model regionality, and model file address.

[0069] (2) Analyze and evaluate whether the existing model can meet the current demand of the client for analysis services.

[0070] (3) Provide the client with the optimal model file or the address of the model file after evaluation.

[0071] Table 1 Model File Information

[0072]

[0073]

[0074] NWDAF sends the trained model to MSF, thus implementing model copy storage. For example... Figure 3 As shown, the MSF network element stores the model file information trained by the NWDAF network element, including, for example, the following steps.

[0075] In step 310, NWDAF calls the Nmsf_MLModelStorage_Request (model storage request based on the Nmsf interface) operation to initiate a model storage request to MSF, which includes descriptive information of the model file to be stored, such as analysis ID, model file address, model performance and other related parameters.

[0076] The models in the various embodiments of this disclosure include, but are not limited to, machine learning (ML) models, artificial intelligence (AI) models, and active learning (AL) models.

[0077] In step 320, MSF invokes the Nmsf_MLModelStorage_Request Response (Nmsf interface-based model storage request and response) operation to provide a model storage response. MSF can decide whether to accept the model storage request based on model file information and resource availability. If MSF decides to accept the model storage request, it initiates step 330.

[0078] In step 330, MSF initiates a model file download request to the NWDAF based on the model file address provided in step 310.

[0079] In step 340, NWDAF responds to the download request and transfers the model file to MSF, which can be done using transfer protocols such as FTP (File Transfer Protocol) or SFTP (SSH File Transfer Protocol).

[0080] like Figure 4 As shown, the network service method based on MSF network elements includes the following steps.

[0081] In step 410, the MSF network element can, for example, use the method in steps 310-340 to store the model file information trained by the NWDAF network element, including the model file and its description information.

[0082] In step 420, the MSF network element responds to the model request related to the analysis service proposed by the model requirement network element and determines whether the stored existing model meets the model request of the model requirement network element.

[0083] Model requesting network elements located in different regions or on different private networks can submit model requests to the MSF network element. Each NWDAF network element with AnLF functionality can also submit model requests to the MSF network element as a model requesting network element. Model requesting network elements include, but are not limited to, each NWDAF network element with AnLF functionality.

[0084] When a network element requests a model, it will have certain requirements regarding the model's analysis service type, model performance, model validity period, and model regionality. The MSF network element will determine whether the existing model meets the model request of the network element based on the analysis service type of the stored existing models and, as needed, at least one of the following: model performance, model validity period, and model regionality.

[0085] In step 430, if there is model file information that satisfies the model request, the MSF network element provides the target model file corresponding to the model request to the model request network element based on the model file information that satisfies the model request.

[0086] MSF network elements select the model file with the best model performance from the model files that meet the model request and provide it as the target model file to the network elements that require the model.

[0087] The MSF network element provides the target model file corresponding to the model request to the model requesting network element, including: the MSF network element sending the model file address of the target model file to the model requesting network element; the MSF network element receiving the model download request initiated by the model requesting network element based on the model file address of the target model file; and the MSF network element sending the target model file to the model requesting network element through the model download request response.

[0088] In step 440, if there is no model file information that satisfies the model request, the MSF network element returns a model request failure response to the model requesting network element.

[0089] This embodiment enhances the management of network data analysis function models by the mobile network. Based on the added model storage function network element, the model is stored and managed, enabling the model to be shared. For example, the model can be shared among multiple network data analysis function network elements, between different regions, or between different private networks. This achieves the effect of training the model once and calling it multiple times, avoiding the waste of resources caused by repeated model training.

[0090] like Figure 5 As shown, the method by which the MSF network element provides network services for an NWDAF with AnLF functionality includes the following steps.

[0091] In step 510, when an NWDAF with AnLF functionality invokes the Nmsf_MLModelInfo_Request (Model File Information Request based on the Nmsf interface) operation, it requests from MSF one or more models associated with one or more analysis IDs. The model description information provided by the NWDAF with AnLF functionality includes, but is not limited to, the model description information in Table 1, such as analysis ID, model accuracy, and model precision.

[0092] In step 520, MSF analyzes and evaluates the existing model based on the model description information provided by NWDAF with AnLF functionality. The evaluation includes, but is not limited to, generalization error, mean absolute error, R-squared value, and mean variance.

[0093] In step 530, if the existing model can meet the analysis service requirements of an NWDAF with AnLF functionality, the MSF sends model file information to the NWDAF with AnLF functionality, including but not limited to the model file address, analysis ID, and model file storage format. If the existing model cannot meet the analysis service requirements of an NWDAF with AnLF functionality, the MSF sends a model request failure response to the NWDAF with AnLF functionality.

[0094] In step 540, if an NWDAF with AnLF functionality fails to request a model from the MSF, it can invoke the Nnwdaf_MLModelInfo_Request (a model file information request based on the Nnwdaf interface) operation to initiate a model request process to an NWDAF with MTLF functionality. The model request includes an indication of whether the model request to the MSF network element failed (i.e., the model request indication parameter, denoted as Model_Request). A value of 0 indicates that the NWDAF has not initiated a model request process to the MSF, while a value of 1 indicates that the NWDAF's model request to the MSF network element failed. Additionally, the model request may include model performance parameters such as model accuracy and model precision, as well as other model description information from Table 1.

[0095] In step 550, the NWDAF with MTLF functionality responds to the model request and sends the corresponding model file information to the NWDAF with AnLF functionality, including but not limited to the model file address, analysis ID, and model file storage format.

[0096] In step 560, following step 530, if the existing model can meet the analysis service requirements of an NWDAF with AnLF functionality, the NWDAF with AnLF functionality can initiate a model download request to MSF based on the model file address provided to it by MSF.

[0097] In step 570, following step 560, the MSF responds to the model download request and sends the model file to the NWDAF with AnLF functionality.

[0098] In step 580, following step 550, the NWDAF with AnLF functionality can initiate a model download request to the NWDAF with MTLF functionality based on the model file address provided to it by the NWDAF with MTLF functionality.

[0099] In step 590, following step 580, the NWDAF with MTLF functionality responds to the model download request and sends the model file to the NWDAF with AnLF functionality.

[0100] In step 5100, the NWDAF with MTLF functionality can, based on the indication information of whether the model has failed to be requested from the MSF network element by the NWDAF with AnLF functionality added in step 540, choose whether to initiate a request to the MSF. Figure 3 The model storage process shown stores the newly trained model in MSF.

[0101] Specifically, according to the indication information, when the indication information is 1, it indicates that the NWDAF with AnLF function failed to request the model from the MSF network element. In other words, the MSF network element does not store the corresponding model file. Then, the NWDAF with MTLF function initiates the model storage process to the MSF to store the corresponding model file information to the MSF.

[0102] This embodiment enhances the management of network data analysis function models by the mobile network. Based on the added model storage function network element, the model is stored and managed, enabling the model to be shared. For example, the model can be shared among multiple network data analysis function network elements, between different regions, or between different private networks. This achieves the effect of training the model once and calling it multiple times, avoiding the waste of resources caused by repeated model training.

[0103] like Figure 6 As shown, the method by which MSF network elements provide model network services for NWDAFs with AnLF functionality located in different regions includes the following steps.

[0104] Prerequisite: MSF is deployed centrally or distributed across multiple provinces.

[0105] In step 610, the NWDAF in Province A with AnLF functionality initiates a model file information request to the MSF associated with the network element load data analysis service, and provides parameter information (i.e., model file description information): Analytics ID = NF loadinformation (analysis ID is network element load information), Target of analytics = UPF 1 (analysis target is UPF1), Model accuracy = 99% (required model accuracy is 99%).

[0106] In step 620, MSF analyzes and evaluates the existing model based on the provided parameter information and finds that the existing model cannot meet its analysis service requirements.

[0107] In step 630, MSF returns a model file information request failure message.

[0108] In step 640, the NWDAF with AnLF functionality in province A initiates a model file information request to the NWDAF with MTLF functionality in province A, and provides parameter information (i.e. model file description information): Analytics ID = NF loadinformation, Model accuracy = 99%, Model_Request = 1.

[0109] In step 650, the NWDAF in province A with MTLF functionality initiates a data collection process and performs model training as needed. After the model training is completed, a model file UPF99.xml is generated, and the address information of the model file is provided to the NWDAF in province A with AnLF functionality: ML model information = sftp: / / 10.69.0.1 / model / NFloadinformation.

[0110] In step 660, the NWDAF with AnLF functionality in province A initiates a model file download request to it based on the received model file address information.

[0111] In step 670, the NWDAF with MTLF functionality in province A responds to its model file download request, allowing the NWDAF with AnLF functionality in province A to download the model file.

[0112] In step 680, the NWDAF with MTLF functionality in province A initiates a storage request for the model to the MSF based on the parameter Model_Request=1 in step 640, providing the parameters: Analytics ID=NF load information, Model accuracy=99%, ML model information=sftp: / / 10.69.0.1 / model / NFloadinformation (the address of the model in the NWDAF with MTLF functionality).

[0113] In step 690, MSF, after combining information such as model file information and resource usage, allows the storage request.

[0114] In step 6100, MSF initiates a download request for the model to the NWDAF with MTLF functionality in Province A based on the parameter information provided in step 680.

[0115] In step 6110, the NWDAF in Province A with MTLF functionality agrees to the download request and performs model file transfer. The model is stored in the MSF, and the stored content includes: Analytics ID = NF load information, Model accuracy = 99%, ML model address = sftp: / / 10.69.0.2 / model / NFloadinformation.

[0116] In step 6120, the NWDAF with AnLF function in Province B initiates a model file information request associated with the network element load data analysis service to the MSF and provides parameter information (i.e. model file description information): Analytics ID = NF loadinformation, Target of analytics = UPF 3, Model accuracy = 94%.

[0117] In step 6130, based on the provided parameter information, MSF found that two existing models meet the analysis requirements:

[0118] Model 1: UPF96.xml, Analytics ID=NF load information, Model accuracy=96%.

[0119] Model 2: UPF99.xml, Analytics ID=NF load information, Model accuracy=99%.

[0120] After analyzing and evaluating existing models, MSF determined that model 2UPF99.xml was the optimal model.

[0121] In step 6140, the MSF sends the file information of the evaluated optimal model UPF99.xml to the NWDAF in Province B that has AnLF functionality. The information includes: Analytics ID = NF load information, ML model address = sftp: / / 10.69.0.2 / model / NFloadinformation (the address of the model in the MSF), Storage type = xml, Model accuracy = 99%.

[0122] In step 6150, the NWDAF with AnLF functionality in Province B initiates a model file download request to the MSF based on the received model information.

[0123] In step 6160, the MSF responds to its model file download request, allowing the NWDAF with AnLF functionality in Province B to download the model file.

[0124] This embodiment enhances the mobile network's management of network data analysis function models. By adding a model storage function network element to store and manage models, models can be shared. For example, models can be shared among multiple network data analysis function network elements, across different regions, and between different private networks. This achieves the effect of training a model once and using it multiple times, avoiding the resource waste caused by repeated model training. Furthermore, the model storage function network element can provide the optimal model to the requester, increasing the availability of network data analysis function models for the requester.

[0125] like Figure 7 As shown, a model storage function network element 700 includes modules 710-730.

[0126] The storage module 710 is configured to store model file information trained by NWDAF network elements, including: receiving a model storage request sent by an NWDAF network element, including description information of the model file to be stored; deciding whether to accept the model storage request and sending a corresponding model storage response to the NWDAF network element; if the decision is made to accept the model storage request, downloading the corresponding model file from the NWDAF network element according to the model file address in the description information; and storing the model file and its description information as model file information.

[0127] The judgment module 720 is configured to respond to model requests related to analysis services proposed by model-requiring network elements, and to determine whether the stored existing models meet the model requests of the model-requiring network elements, including: judging whether the existing models meet the model requests of the model-requiring network elements based on the analysis service type of the stored existing models, and in combination with at least one of model performance, model validity period, and model regionality.

[0128] The module 730 is configured to provide the target model file corresponding to the model request to the model requesting network element based on the model file information that satisfies the model request. For example, from the model files that satisfy the model request, the model file with the best model performance is selected as the target model file and provided to the model requesting network element. Providing the target model file corresponding to the model request to the model requesting network element includes: sending the model file address of the target model file to the model requesting network element; receiving the model download request initiated by the model requesting network element based on the model file address of the target model file; and sending the target model file to the model requesting network element through the model download request response.

[0129] like Figure 8 As shown, a network data analysis function network element 800 includes modules 810-830.

[0130] The receiving module 810 is configured to receive a model request related to the analysis service sent by the model request network element based on the model request failure response of the MSF network element, which carries an indication information of whether the model request network element requested the MSF network element to determine whether the model failed.

[0131] The first providing module 820 is configured to respond to model requests related to analysis services raised by model requesting network elements and provide the target model file corresponding to the model request to the model requesting network elements.

[0132] The second providing module 830 is configured to, according to the instruction information, initiate a process of storing the target model file information into the MSF network element when the model requesting network element fails to request a model from the MSF network element.

[0133] like Figure 9 As shown, a model requires network element 900, which includes modules 910-940.

[0134] Request module 910 is configured to submit model requests related to analysis services to MSF network elements;

[0135] The receiving module 920 is configured to receive a model request success response or a model request failure response sent by the MSF network element.

[0136] The first model acquisition module 930 is configured to initiate a model download process for the target model file to the MSF network element based on the model file address of the target model file in the successful model request response.

[0137] The second model acquisition module 940 is configured to send a model request related to the analysis service to the NWDAF network element based on the model request failure response, and receive the target model file corresponding to the model request returned by the NWDAF network element. The model request carries indication information as to whether the model request to the MSF network element has failed, so as to instruct the NWDAF network element to initiate the process of storing the target model file information in the MSF network element when the model requesting network element fails to request the model from the MSF network element.

[0138] like Figure 10 As shown, a model storage function network element 1000 includes: a memory 1010; and a processor 1020 coupled to the memory, the processor being configured to execute network service methods executed by the model storage function network element in various embodiments based on instructions stored in the memory.

[0139] like Figure 11As shown, a model demand network element 1100 includes: a memory 1110; and a processor 1120 coupled to the memory, the processor being configured to execute network service methods performed by the model demand network element (demand side) in various embodiments based on instructions stored in the memory.

[0140] The memory may include, for example, system memory, fixed non-volatile storage media, etc. System memory may store, for example, the operating system, application programs, the boot loader, and other programs.

[0141] The processor can be implemented using general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete hardware components such as discrete gates or transistors.

[0142] In addition, the model storage function network element 1000 and the model requirement network element 1100 may also include input / output interfaces, network interfaces, and storage interfaces. These interfaces, as well as the connection between the memory and the processor, can be via a bus. The input / output interfaces provide connection interfaces for input / output devices such as displays, mice, keyboards, and touchscreens. The network interfaces provide connection interfaces for various networked devices. The storage interfaces provide connection interfaces for external storage devices such as SD cards and USB flash drives. The bus can use any bus architecture from various bus structures. For example, bus architectures include, but are not limited to, Industry Standard Architecture (ISA) buses, Micro Channel Architecture (MCA) buses, and Peripheral Component Interconnect (PCI) buses.

[0143] like Figure 12 As shown, a network service system 1200 includes: model storage function network elements 700 and 1000, model requirement network elements 900 and 1100, and network data analysis function network element 800.

[0144] Some embodiments of this disclosure provide a non-transitory computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the network service methods of various embodiments.

[0145] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more non-transitory computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer program code.

[0146] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0147] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0148] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0149] The above description is only a preferred embodiment of this disclosure and is not intended to limit this disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.

Claims

1. A network service method characterized by, include: Model storage function: MSF network element stores network data; NWDAF network element training model file information; MSF network element response to model requirements: The network element proposes an analysis service related model request and determines whether the stored existing model meets the model requirements of the network element's model request. MSF network elements provide the target model file corresponding to the model request to the model requesting network elements based on the model file information that meets the model request.

2. The method of claim 1, wherein, MSF network element storage includes the model file information for NWDAF network element training, including: The MSF network element receives a model storage request sent by the NWDAF network element, which includes description information of the model file to be stored; The MSF network element decides whether to accept the model storage request and sends the corresponding model storage response to the NWDAF network element. If the MSF network element decides to accept the model storage request, it downloads the corresponding model file from the NWDAF network element according to the model file address in the description information; MSF network elements store model files and their description information as model file information.

3. The method of claim 1, wherein, MSF network elements determine whether the existing stored models meet the model requirements. The model requests of the network element include: MSF network elements determine whether the existing models meet the model requirements of the network element's model requests based on the analysis service type of the stored existing models and at least one of the following: model performance, model validity period, and model regionality.

4. The method of claim 1, wherein, MSF network elements provide the target model files corresponding to the model request to the model requesting network elements based on the model file information that meets the model request, including: MSF network elements select the model file with the best model performance from the model files that meet the model request and provide it as the target model file to the network elements that require the model.

5. The method of claim 1, wherein, MSF network elements provide the target model files corresponding to the model request network elements, including: MSF network element sends the model file address of the target model file to the network element that requires the model; MSF network element receives model download requests initiated by the model requesting network element based on the model file address of the target model file; MSF network elements send the target model file to the network element that requires the model through the model download request response.

6. The method according to claim 1, characterized in that, If there is no model file information to satisfy the model request, the MSF network element returns a model request failure response to the model requesting network element; The NWDAF network element receives the model request related to the analysis service sent by the model request failure response from the model request network element, which carries an indication of whether the model request to the MSF network element has failed. NWDAF network element responds to model requests related to analysis services submitted by network elements that require model services, and provides the target model file corresponding to the model request to the network elements that require model services. According to the instructions in the instruction information, if the model requesting network element fails to request a model from the MSF network element, the NWDAF network element initiates a process to store the target model file information into the MSF network element.

7. The method according to any one of claims 1 to 6, characterized in that, The model file information includes the model file and its description information. The description information of the model file includes the analysis service type, and also includes at least one of the following: model performance, model validity period, model regionality, and model file address.

8. A network service method characterized by, include: Submit a model request related to the analysis service to the MSF network element; Receive either a successful model request response or a failed model request response from the MSF network element; Based on the model file address of the target model file in the successful model request response, initiate the model download process of the target model file to the MSF network element; Based on the model request failure response, the system sends a model request related to the analysis service to the NWDAF network element and receives the target model file corresponding to the model request returned by the NWDAF network element. The model request carries an indication of whether the model request to the MSF network element has failed, so as to instruct the NWDAF network element to initiate the process of storing the target model file information to the MSF network element when the model requesting network element fails to request the model from the MSF network element.

9. A model storage function network element, characterized by, include: The storage module is configured to store model file information for training network elements in the NWDAF network data analysis function; The judgment module is configured to respond to model requests related to analysis services raised by network elements that require model services, and to determine whether the existing stored models meet the model requests of the network elements that require model services. The module is configured to provide the target model file corresponding to the model request to the network element that requires the model, based on the model file information that satisfies the model request.

10. A network data analytics function network element, characterized by, include: The receiving module is configured to receive model requests related to analysis services sent by the model request failure response from the model request network element based on the MSF network element, which carries indication information as to whether the request to the MSF network element for the model has failed. The first providing module is configured to respond to model requests related to analysis services raised by model requesting network elements and provide the target model file corresponding to the model request to the model requesting network elements. The second providing module is configured to, according to the instructions in the instruction information, initiate a process of storing the target model file information into the MSF network element when the model requesting network element fails to request a model from the MSF network element.

11. A model requirement network element, characterized by, include: The request module is configured to submit model requests related to analysis services to MSF network elements; The receiving module is configured to receive either a successful model request response or a failed model request response sent by the MSF network element. The first model acquisition module is configured to initiate a model download process for the target model file to the MSF network element based on the model file address of the target model file in the successful response to the model request. The second model acquisition module is configured to send a model request related to the analysis service to the NWDAF network element based on the model request failure response, and receive the target model file corresponding to the model request returned by the NWDAF network element. The model request carries an indication of whether the model request to the MSF network element has failed, so as to instruct the NWDAF network element to initiate the process of storing the target model file information to the MSF network element when the model requesting network element fails to request the model from the MSF network element.

12. A model storage function network element, comprising: Memory; as well as A processor coupled to the memory, the processor being configured to execute the network service method of any one of claims 1-7 based on instructions stored in the memory.

13. A model requirement network element, comprising: Memory; as well as a processor coupled to the memory, the processor configured to perform the network service method of claim 8 based on instructions stored in the memory.

14. A network service system comprising: the model storage function network element of claim 9 or 12, the model requirement network element of claim 11 or 13, the network data analysis function network element of claim 10.

15. A non-transitory computer readable storage medium having stored thereon a computer program, which, when executed by a processor, implements the steps of the network service method of any one of claims 1-8.

Images