Communication method and device

Abstract

The present application provides a communication method and communication device. The communication method includes: an operator's data analysis network element determines the device manufacturer information of terminal service experience data through association information; the operator's data analysis network element obtains the address information of the device manufacturer's data analysis network element; the operator's data analysis network element, in conjunction with the device manufacturer's data analysis network element, determines a service experience model corresponding to the device manufacturer based on the device manufacturer's corresponding service experience data and network-side data. The method in the embodiment of the present application can effectively establish a service experience model corresponding to a device manufacturer in a scenario where the core network network element is cross-device manufacturer.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of communications, and more specifically, to a communication method and a communication device. Background Art

[0002] 5G networks need to accurately measure the service experience of services within the network and adjust the network if the service experience does not meet requirements. The prerequisite for measuring service experience is to train a service experience model. This service experience model can be determined based on terminal data about the service on the radio access network (RAN), core network (CN), and service providers.

[0003] However, in the CN of the same operator, when network devices come from different equipment manufacturers and are unwilling to share their private data, the 5G network cannot obtain complete data that affects the service experience, and thus cannot complete the training of the service experience model. Summary of the Invention

[0004] The present application provides a communication method and a communication device, which enable effective training of service experience models in a cross-vendor scenario of network elements within a core network domain.

[0005] In a first aspect, an embodiment of the present application provides a communication method, which includes: a first data analysis network element obtains service experience data of a terminal service on an application function network element and associated information corresponding to the service experience data, wherein the service is provided by a core network element of an equipment manufacturer; the first data analysis network element determines the address information of a second data analysis network element corresponding to the equipment manufacturer based on the associated information; the first data analysis network element determines the service experience model of the service based on the address information of the second data analysis network element, the service experience data and the associated information in conjunction with the second data analysis network element.

[0006] In a possible implementation, the first data analysis network element includes a data analysis network element of an operator, and the second data analysis network element includes a data analysis network element of an equipment manufacturer.

[0007] In a possible implementation, the service experience model is a service experience model of the device manufacturer.

[0008] According to the communication method of the embodiment of the present application, the first data analysis network element obtains the service experience data and associated information on the application function network element, and the first data analysis network element determines the service experience data in the service experience data set based on the associated information. The service experience data is the service experience data generated by the terminal when accessing the service provided by the core network network element of the equipment manufacturer, and the associated information is used to associate the network data generated by the terminal on the core network network element of the equipment manufacturer when accessing the service with the service experience data generated on the application function network element. After the first data analysis network element determines the address information of the second data analysis network element based on the associated information, the first data analysis network element then determines the service experience model of the service based on the service experience data and the associated information in conjunction with the second data analysis network element. The communication method of the embodiment of the present application can determine the service experience data of the equipment manufacturer through associated information in the scenario where the network elements are across equipment manufacturers within the core network domain, thereby training a service experience model of the service.

[0009] In some embodiments, the association information includes association information between UPF and AF. Specifically, as an example and not a limitation, AF and UPF can pairwise associate data belonging to the same terminal on two network elements through association information of timestamp and IP address 5-tuple.

[0010] In combination with the first aspect, in a first possible implementation method, the network storage function network element stores the correspondence between the association information and the address information of the second data analysis network element, and the first data analysis network element determines the address information of the second data analysis network element based on the association information, including: the first data analysis network element sends a first request to the network storage function network element, the first request is used to request the address information of the second data analysis network element, and the first request includes the association information; the first data analysis network element receives a first response from the network element storage function network element, and the first response includes the address information of the second data analysis network element.

[0011] According to the communication method of the embodiment of the present application, the second data analysis network element will carry the associated information to register with the network storage function network element in advance, and the first data analysis network element can determine the address information of the second data analysis network element through the associated information, so that the first data analysis network element can jointly complete the training of the service experience model with the second data analysis network element.

[0012] In combination with the first aspect, in a second possible implementation method, the association information corresponds to the identification information of the equipment manufacturer, the network storage function network element stores the correspondence between the identification information of the equipment manufacturer and the address information of the second data analysis network element corresponding to the equipment manufacturer, and the first data analysis network element determines the address information of the second data analysis network element based on the association information, including: the first data analysis network element determines the identification information of the equipment manufacturer based on the association information; the first data analysis network element sends a second request to the network storage function network element, the second request is used to request the address information of the second data analysis network element, and the second request includes the identification information of the equipment manufacturer; the first data analysis network element receives a second response from the network element storage function network element, and the second response includes the address information of the second data analysis network element.

[0013] According to the communication method of the embodiment of the present application, the second data analysis network element will register with the network storage function network element in advance with the identification information of the device manufacturer. The first data analysis network element first determines the identification information of the device manufacturer based on the associated information, and then determines the address information of the second data analysis network element based on the identification information of the device manufacturer, so that the first data analysis network element can jointly complete the training of the service experience model with the second data analysis network element.

[0014] In combination with the second possible implementation method, in a third possible implementation method, before the first data analysis network element determines the identification information of the equipment manufacturer based on the association information, the method also includes: the first data analysis network element obtains the correspondence between the association information and the identification information of the equipment manufacturer from the core network network element of the equipment manufacturer.

[0015] In combination with the first aspect or any one of the above possible implementation methods, in a fourth possible implementation method, the first data analysis network element determines the service experience model of the service in conjunction with the second data analysis network element based on the address information of the second data analysis network element, the service experience data and the association information, including: the first data analysis network element sends the association information and indication information to the second data analysis network element based on the address information of the second data analysis network element, and the indication information is used to instruct the second data analysis network element to perform distributed machine learning model training based on the association information; the first data analysis network element receives the sub-model corresponding to the association information from the second data analysis network element, and the sub-model is determined by the second data analysis network element based on the first network data of the terminal on the core network element of the equipment manufacturer; the first data analysis network element determines the service experience model based on the service experience data and the sub-model.

[0016] In combination with the fourth possible implementation method, in the fifth possible implementation method, the method also includes: the first data analysis network element receives the second network data on the core network network element of the equipment manufacturer corresponding to the association information; the first data analysis network element determines the business experience model based on the business experience data, the sub-model and the second network data.

[0017] In some embodiments, the above-mentioned second network data can be in a fourth possible implementation method. When the first data analysis network element obtains the correspondence between the association information and the identification information of the equipment manufacturer from the core network network element of the equipment manufacturer, it also obtains the second data network from the core network network element of the equipment manufacturer to save a certain amount of resource overhead.

[0018] In combination with the fourth possible implementation manner or the fifth possible implementation manner, in a sixth possible implementation manner, the first network data includes private network data corresponding to the service of the terminal on the core network element of the equipment manufacturer.

[0019] The sixth possible implementation method described above can allow the private network data of the service to participate in the training of the service experience model while ensuring data privacy, thereby improving the generalization ability of the service experience model and ensuring the service experience.

[0020] In combination with the fourth possible implementation manner or the fifth possible implementation manner, in a seventh possible implementation manner, the second network data includes public network data corresponding to the service of the terminal on the core network element of the equipment manufacturer.

[0021] In the seventh possible implementation method above, in addition to the private network data and service experience data corresponding to the service, the public network data corresponding to the service is also involved in the training of the service experience model, further improving the accuracy of the service experience model, allowing the service provider to accurately measure its service experience and effectively monitor the service quality, so that the service experience requirements and network resources can be accurately matched.

[0022] In combination with the first aspect or any possible implementation method above, in an eighth possible implementation method, the method for training the distributed machine learning model includes one or more of the following methods: vertical federated learning, horizontal federated learning, transfer learning, or sharing learning.

[0023] In the above implementation, the distributed machine learning model training method can enable the private network data of the business to be retained locally at the data source to participate in the training of the business experience model, thereby ensuring the privacy of the data and the accuracy and effectiveness of the data experience model, thereby ensuring the business experience.

[0024] In any of the above implementations, the determined service experience model is the service experience model corresponding to the device manufacturer.

[0025] In combination with the first aspect or any possible implementation method mentioned above, in the eighth possible implementation method, the core network network element of the equipment manufacturer includes one or more of the following network elements: access and mobility management function AMF network element, session management function SMF network element, policy control PCF network element, user plane function UPF network element or unified data management UDM network element.

[0026] In the second aspect, an embodiment of the present application provides a communication method, which includes: a first network element obtains service experience data of a terminal service on an application function network element and associated information corresponding to the service experience data, wherein the service is provided by a core network network element of an equipment manufacturer; the first network element determines the address information of a data analysis network element of the equipment manufacturer based on the associated information; the first network element sends the associated information and the service experience data to the data analysis network element of the equipment manufacturer based on the address information of the data analysis network element of the equipment manufacturer, and the service experience data is used to determine a service experience model of the service.

[0027] According to the communication method of an embodiment of the present application, the first network element obtains service experience data and associated information on the application function network element, and the first network element determines the service experience data in the service experience data set based on the associated information. The service experience data is the service experience data generated when the terminal accesses the service provided by the core network network element of the equipment manufacturer, and the associated information is used to associate the network data generated by the terminal on the core network network element of the equipment manufacturer when accessing the service with the service experience data generated on the application function network element. After the first network element determines the address information of the data analysis network element of the equipment manufacturer based on the associated information, the first network element sends the associated information and the service experience data to the data analysis network element of the equipment manufacturer for training the service experience model by the data analysis network element of the equipment manufacturer.

[0028] The communication method of the embodiment of the present application can determine the service experience data of the equipment manufacturer through the associated information in a scenario where the network elements are across equipment manufacturers within the core network domain, so that the data analysis network element of the equipment manufacturer can obtain the service experience data corresponding to the equipment manufacturer, thereby enabling the equipment manufacturer to train the service experience model corresponding to the equipment manufacturer.

[0029] In some embodiments, the association information includes association information between UPF and AF. Specifically, as an example and not a limitation, AF and UPF can pairwise associate data belonging to the same terminal on two network elements through association information of timestamp and IP address 5-tuple.

[0030] In combination with the second aspect, in a first possible implementation method, the network storage function network element stores the correspondence between the association information and the address information of the data analysis network element of the equipment manufacturer, and the first network element determines the address information of the data analysis network element of the equipment manufacturer based on the association information, including: the first network element sends a first request to the network storage function network element, the first request is used to request the address information of the data analysis network element of the equipment manufacturer, and the first request includes the association information; the first network element receives a first response from the network storage function network element, and the first response includes the address information of the data analysis network element of the equipment manufacturer.

[0031] According to the communication method of the embodiment of the present application, the data analysis network element of the equipment manufacturer will carry the associated information to register with the network storage function network element in advance. The first network element can determine the address information of the data analysis network element of the equipment manufacturer through the associated information, so that the first network element can send the service experience data corresponding to the equipment manufacturer to the data analysis network element of the equipment manufacturer, so that the data analysis network element of the equipment can complete the training of the service experience model.

[0032] In combination with the second aspect, in a second possible implementation method, the association information corresponds to the identification information of the device manufacturer, the network storage function network element stores the correspondence between the identification information of the device manufacturer and the address information of the data analysis network element of the device manufacturer, and the first network element determines the address information of the data analysis network element of the device manufacturer based on the association information, including: the first network element determines the identification information of the device manufacturer based on the association information; the first network element sends a second request to the network storage function network element, the second request is used to request the address information of the data analysis network element of the device manufacturer, and the second request includes the identification information of the device manufacturer; the first network element receives a second response from the network storage function network element, and the second response includes the address information of the data analysis network element of the device manufacturer.

[0033] According to the communication method of the embodiment of the present application, the data analysis network element of the equipment manufacturer will carry the identification information of the equipment manufacturer to register with the network storage function network element in advance. The first network element first determines the identification information of the equipment manufacturer based on the associated information, and then determines the address information of the data analysis network element of the equipment manufacturer based on the identification information of the equipment manufacturer. In this way, the first network element can send the service experience data corresponding to the equipment manufacturer to the data analysis network element of the equipment manufacturer, so that the data analysis network element of the equipment can complete the training of the service experience model.

[0034] In combination with the second aspect or any one of the above possible implementations, in a third possible implementation, the first network element includes a network capability exposure function network element or an operator's data analysis network element.

[0035] On the third aspect, an embodiment of the present application provides a communication method, which includes: the data analysis network element of the equipment manufacturer obtains related information and first network data of the service of the terminal on the core network network element of the equipment manufacturer, and the service is provided by the core network network element of the equipment manufacturer; the data analysis network element of the equipment manufacturer jointly determines the service experience model of the service with the data analysis network element of the operator based on the related information and the first network data.

[0036] According to the communication method of the embodiment of the present application, in the scenario where network elements in the core network domain are distributed across different equipment manufacturers, the data analysis network element of the equipment manufacturer of the core network element can work with the data analysis network element of the operator to train a service experience model corresponding to the equipment manufacturer.

[0037] In combination with the third aspect, in a first possible implementation method, the method further includes: the data analysis network element of the equipment manufacturer sends a network element registration request to the network storage function network element, and the network element registration request includes the association information and / or the identification information of the equipment manufacturer.

[0038] According to the communication method of the embodiment of the present application, the data analysis network element of the equipment manufacturer of the core network network element will register with the network storage function network element in advance with the associated information corresponding to the equipment manufacturer and / or the identification information of the equipment manufacturer, so that the first network element can query the network storage function network element through the associated information and / or the identification information of the equipment manufacturer to determine the service experience data corresponding to the equipment manufacturer, so that the service experience data of the equipment manufacturer can participate in the training of the service experience model corresponding to the equipment manufacturer.

[0039] In combination with the third aspect or the first possible implementation method, in a second possible implementation method, the data analysis network element of the equipment manufacturer determines the service experience model of the service based on the association information and the first network data in conjunction with the data analysis network element of the operator, including: the data analysis network element of the equipment manufacturer receives the association information and indication information from the data analysis network element of the operator, and the indication information is used to instruct the data analysis network element of the equipment manufacturer to perform distributed machine learning model training based on the association information; the data analysis network element of the equipment manufacturer determines a sub-model based on the first network data of the terminal on the core network element of the equipment manufacturer; the data analysis network element of the equipment manufacturer sends the sub-model to the data analysis network element of the operator, and the sub-model is used to determine the service experience model.

[0040] According to the communication method of the above embodiment of the present application, for the training of the service experience model of the equipment manufacturer, the first network data of the terminal on the core network network element of the equipment manufacturer participates in the training of the service experience model corresponding to the equipment manufacturer on the data analysis network element of the equipment manufacturer, and the service experience data of the terminal on the application function network element participates in the training on the data analysis network element of the operator. Through the distributed machine learning method, the data analysis network element of the equipment manufacturer only needs to send the sub-model to the data analysis network element of the operator, and avoids directly sending the first network data on the core network network element of the equipment manufacturer. Therefore, the communication method can allow the first network data and service experience data corresponding to the equipment manufacturer to participate in the training of the service experience model corresponding to the equipment manufacturer under the premise of ensuring the data privacy of all parties, thereby improving the generalization ability of the service experience model and ensuring the service experience.

[0041] In combination with the second possible implementation method, in a third possible implementation method, the method of training the distributed machine learning model includes one or more of the following methods: vertical federated learning, horizontal federated learning, transfer learning, or sharing learning.

[0042] In the above implementation, the distributed machine learning model training method can enable the private network data of the business to be retained locally at the data source to participate in the training of the business experience model, thereby ensuring the privacy of the data and the accuracy and effectiveness of the data experience model, thereby ensuring the business experience.

[0043] In combination with the third aspect or the first possible implementation manner, in a fourth possible implementation manner, the data analysis network element of the device manufacturer jointly determines the service experience model of the service with the data analysis network element of the operator based on the association information and the first network data, including: the data analysis network element of the device manufacturer receives the association information and the service experience data from a first network element, where the first network element includes a network capability exposure function network element or an operator's data analysis network element, and the service experience data is used to determine the service experience model of the service;

[0044] The data analysis network element of the equipment manufacturer receives the associated information and indication information from the data analysis network element of the operator, and the indication information is used to instruct the data analysis network element of the equipment manufacturer to train the service experience model based on the associated information; the data analysis network element of the equipment manufacturer determines the service experience model based on the first network data and the service experience data.

[0045] According to the communication method of the above embodiments of the present application, the data analysis network element of the equipment manufacturer can directly obtain the first network data and service experience data corresponding to the equipment manufacturer, without the need for a distributed machine learning method. The training of the service experience model corresponding to the equipment manufacturer can be completed on the data analysis network element of the equipment manufacturer based on the first network data and the service experience data, which not only ensures the generalization ability of the service experience model corresponding to the equipment manufacturer, but also improves the training efficiency of the service experience model.

[0046] In combination with the third aspect or any possible implementation method mentioned above, in a fifth possible implementation method, the data analysis network element of the equipment manufacturer jointly determines the service experience model of the service with the data analysis network element of the operator based on the associated information, the first network data and the second network data, and the second network data includes the public network data of the service of the terminal on the core network network element of the equipment manufacturer.

[0047] In the above fifth possible implementation method, in addition to the first network data and service experience data corresponding to the service, the public network data corresponding to the service is also involved in the training of the service experience model, thereby further improving the accuracy of the service experience model, allowing the service provider to accurately measure its service experience and effectively monitor the service quality, so that the service experience requirements and network resources can be accurately matched.

[0048] In combination with the third aspect or any of the above possible implementations, in a sixth possible implementation, the first network data includes private network data of the service of the terminal on the core network element of the equipment manufacturer.

[0049] The sixth possible implementation method described above can allow the private network data of the service to participate in the training of the service experience model while ensuring data privacy, thereby improving the generalization ability of the service experience model and ensuring the service experience.

[0050] In combination with the third aspect or any possible implementation method mentioned above, in the seventh possible implementation method, the core network network element of the equipment manufacturer includes one or more of the following network elements: access and mobility management function AMF network element, session management function SMF network element, policy control PCF network element, user plane function UPF network element or unified data management UDM network element.

[0051] In a fourth aspect, a communication device is provided, which can be used to perform the operations of the communication device in the first aspect and any possible implementation of the first aspect. Specifically, the communication device includes components (means) corresponding to the steps or functions described in the first aspect, which can be the first communication device in the first aspect. The steps or functions can be implemented through software, hardware, or a combination of hardware and software.

[0052] In a fifth aspect, a communication device is provided, which can be used to perform the operations of the communication device in the second aspect and any possible implementation of the second aspect. Specifically, the device may include means for performing the steps or functions corresponding to the steps or functions described in the second aspect. The steps or functions can be implemented through software, hardware, or a combination of hardware and software.

[0053] In a sixth aspect, a communication device is provided, which can be used to perform the operations of the communication device in the third aspect and any possible implementation of the third aspect. Specifically, the communication device includes means corresponding to the steps or functions described in the third aspect, which can be the first communication device of the third aspect. The steps or functions can be implemented through software, hardware, or a combination of hardware and software.

[0054] In the seventh aspect, a computer-readable medium is provided, which stores a computer program (also referred to as code, or instructions) which, when run on a computer, enables the computer to execute the method in any possible implementation of the first to third aspects above.

[0055] In the eighth aspect, a chip system is provided, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that a communication device equipped with the chip system executes a method in any possible implementation of the above-mentioned first to third aspects.

[0056] In a ninth aspect, a chip is provided, comprising a processor and a communication interface, wherein the communication interface is used to communicate with an external device or an internal device, and the processor is used to implement the method in any possible implementation of the first to third aspects above.

[0057] In one possible implementation, the chip may further include a memory storing instructions, and the processor is configured to execute the instructions stored in the memory or instructions derived from other sources. When the instructions are executed, the processor is configured to implement the method of any possible implementation of the first to third aspects described above. In one possible implementation, the chip may be integrated into an access network device.

[0058] In the tenth aspect, a computer program product is provided, which includes: a computer program (also referred to as code, or instructions), which, when executed, enables a computer to execute a method in any possible implementation of the first to third aspects above.

[0059] In the eleventh aspect, a communication device is provided, including a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the communication device executes the communication method in any possible implementation of the first to third aspects.

[0060] There are one or more processors and one or more memories. The memories may be integrated with the processors, or may be separately provided from the processors.

[0061] In one possible design, a communication device is provided, including a communication interface, a processor, and a memory. The processor is configured to control the communication interface to transmit and receive signals, the memory is configured to store a computer program, and the processor is configured to retrieve and execute the computer program from the memory, so that the communication device performs the method of any possible implementation of the first to fourth aspects or the first to third aspects.

[0062] In a twelfth aspect, a system is provided, comprising the above-mentioned communication device.

[0063] The communication method and communication device of the embodiments of the present invention can determine the equipment manufacturer information corresponding to the service experience data provided by the application function network element, so that in the scenario where the core network network elements are across equipment manufacturers, the service experience model of each equipment manufacturer can be trained, and the accuracy of the service experience model can be improved at the same time, so that the network equipment can accurately measure the service experience of the service in the network, and adaptively adjust the network when the service experience does not meet the requirements, thereby ensuring the user's service experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a schematic diagram of an application scenario to which the method of the embodiment of the present application is applicable.

[0065] Figure 2 Schematic diagram of the federated learning system framework involved in the embodiments of this application.

[0066] Figure 3 This is a schematic diagram of a method of associating terminal data on each network element in an embodiment of the present application.

[0067] Figure 4 This is a schematic flowchart of the service experience model training method provided in an embodiment of the present application.

[0068] Figure 5 This is a schematic flowchart of a service experience model training method provided in another embodiment of the present application.

[0069] Figure 6 This is a schematic flowchart of a service experience model training method provided in another embodiment of the present application.

[0070] Figure 7 This is a schematic flowchart of a service experience model training method provided in another embodiment of the present application.

[0071] Figure 8 This is a schematic flowchart of a service experience model training method provided in another embodiment of the present application.

[0072] Figure 9 This is a schematic block diagram of a communication device provided in an embodiment of the present application.

[0073] Figure 10 This is a schematic block diagram of another communication device provided in an embodiment of the present application.

[0074] Figure 11 It is a structural diagram of a communication device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0075] The technical solution in this application will be described below with reference to the accompanying drawings.

[0076] The technical solutions of the embodiments of the present application can be applied to various communication systems, such as: long term evolution (LTE) system, LTE frequency division duplex (FDD) system, LTE time division duplex (TDD) system, universal mobile telecommunication system (UMTS), fifth generation (5G) system, future fifth and a half generation (5.5G), sixth generation (6G) or new radio (NR), etc.

[0077] The following takes the fifth generation system as an example, combined with Figure 1 A network architecture based on a network data analytics function (NWDAF) applicable to the present application is described.

[0078] like Figure 1 As shown, the communication system includes but is not limited to the following network elements:

[0079] 1. Terminal equipment

[0080] The terminal device in the embodiments of the present application may also be referred to as: user equipment (UE), mobile station (MS), mobile terminal (MT), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication equipment, user agent or user device, etc.

[0081] The terminal device may be a device that provides voice / data connectivity to users, such as a handheld device or vehicle-mounted device with wireless connection function. At present, some examples of terminals are: mobile phones, tablet computers, laptop computers, PDAs, mobile internet devices (MIDs), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self-driving or autopilot, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, computing devices or other processing devices connected to wireless modems, vehicle-mounted devices, wearable devices, terminal devices in future 5G networks or future evolved public land mobile communication networks (PLMNs). Mobile network, PLMN) and other terminal devices, etc., which are not limited in the embodiments of the present application.

[0082] As an example and not a limitation, in the embodiment of the present application, the terminal device may also be a wearable device. Wearable devices may also be called wearable smart devices, which are a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothing, and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are fully functional, large in size, and can achieve complete or partial functions without relying on smartphones, such as smart watches or smart glasses, as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various types of smart bracelets and smart jewelry for vital sign monitoring. In addition, in the embodiment of the present application, the terminal device may also be a terminal device in the Internet of Things (IoT) system.

[0083] In addition, in the embodiment of the present application, the terminal device can also communicate with the terminal devices of other communication systems, for example, inter-device communication, etc. For example, the terminal device can also transmit (for example, send and / or receive) time synchronization messages with the terminal devices of other communication systems.

[0084] 2. Radio access network (RAN)

[0085] A radio access network (RAN) is an access network that implements network access based on wireless communication technologies. It manages radio resources, provides wireless access or air interface access services to terminals, and forwards control signals and user data between terminals and the core network.

[0086] As an example and not a limitation, the wireless access network can be an evolved NodeB (eNB or eNodeB) in an LTE system, or a wireless controller in a cloud radio access network (CRAN) scenario, or the access device can be a relay station, an access point, a vehicle-mounted device, a wearable device, an access device in a 5G network, or an access device in a future evolved PLMN network, etc. It can be an access point (AP) in a WLAN, or a gNB in ​​an NR system. The embodiments of this application are not limited.

[0087] 3. Network data analytics function (NWDAF)

[0088] The network data analysis function network element NWDAF has at least one of the following functions: data collection function, model training function, analysis result inference function, and analysis result feedback function. Among them, the data collection function refers to the function used to collect data from network network elements, third-party service servers, terminal equipment or network management systems; the model training function refers to the analysis and training based on relevant input data to obtain a model, and the analysis result inference function determines the data analysis result based on the trained model and the inference data. Finally, the analysis result feedback function can provide data analysis results to network network elements, third-party service servers, terminal equipment or network management systems. The analysis results can assist the network in selecting service quality parameters for the service, or assist the network in executing traffic routing, or assist the network in selecting background traffic transmission strategies, etc. This application mainly involves the data collection function and model training function of NWDAF.

[0089] In the embodiments of the present application, the NWDAF may be a separate network element or may be co-located with other core network elements. For example, the NWDAF network element may be co-located with an access and mobility management function (AMF) network element or a session management function (SMF) network element.

[0090] Typical application scenarios of NWDAF include: customization or optimization of terminal parameters, that is, NWDAF collects information such as user connection management, mobility management, session management, and accessed services, and uses reliable analysis and prediction models to evaluate and analyze different types of users, build user portraits, determine users' movement trajectories and service usage habits, and predict user behavior. Based on the analysis and prediction data, the 5G network optimizes user mobility management parameters and wireless resource management parameters; service (path) optimization, that is, NWDAF collects information such as network performance, service load in a specific area, and user service experience, and uses reliable network performance analysis and prediction models to evaluate and analyze different types of services, build service portraits, determine the inherent correlations such as service quality of experience (QoE), service experience, service path or 5G service quality (QoS) parameters, and optimize service paths, service routing, 5G edge computing, and service corresponding to 5G QoS, etc.; AF optimizes service parameters. For example, the Internet of Vehicles (IoV) is a key technology in 5G networks. In autonomous driving scenarios, predicting the network performance (such as QoS information and service load) of the base stations that the vehicle will pass plays a vital role in improving the IoV's service quality. For example, IoV servers can determine whether to continue autonomous driving mode based on predicted network performance information. NWDAF collects information such as network performance and service load in specific areas, and utilizes reliable network performance analysis and prediction models to achieve statistical and prediction of network performance, assisting AF in optimizing parameters.

[0091] It should be noted that, in the embodiment of the present application, each equipment manufacturer has its corresponding equipment manufacturer's network data analysis function network element.

[0092] 4. Session Management Function (SMF)

[0093] The session management function network element is mainly used for session management, Internet protocol (IP) address allocation and management of terminal devices, selection of manageable user plane function (UPF) network elements, termination points of policy control and charging function interfaces, and downlink data notification. In the embodiment of the present application, it can be used to implement the functions of the session management network element.

[0094] 5. Access and mobility management function (AMF)

[0095] The access and mobility management function network element is mainly used for mobility management and access management, etc., and can be used to implement other functions of the mobility management entity (MME) in addition to session management, such as lawful interception or access authorization (or authentication). In the embodiment of the present application, it can be used to implement the functions of the access and mobility management network element.

[0096] 6. Policy Control Function (PCF)

[0097] The policy control network element is used to guide the unified policy framework of network behavior and provide policy rule information to the control plane functional network elements (such as AMF, SMF network elements, etc.).

[0098] 7. Application Function Network Element (AF)

[0099] The application function network element is used to provide services, or to route data affected by the application, access the network open function network element, or interact with the NWDAF network element for service data and policy control.

[0100] 8. User plane function (UPF)

[0101] The user plane function network element can be used for packet routing and forwarding, or QoS parameter processing of user plane data, etc. User data can be accessed to the data network (DN) through this network element. In the embodiment of the present application, it can be used to implement the functions of the user plane network element.

[0102] 9. Network repository function (NRF)

[0103] The network storage function network element can be used to support network element services or network element discovery functions, receive NF discovery requests from network function (NF) instances, and provide information about discovered NF instances to the NF instances. It can also support the maintenance of NF configuration files for available NF instances and the services they support. In embodiments of the present application, it can be used to support network element services or network element discovery functions.

[0104] 10. Network exposure function (NEF): used to expose services and network capability information (such as terminal location and session reachability) provided by 3GPP network functions to the outside world.

[0105] In the above network architecture, the N2 interface is the interface between the RAN and AMF network elements, used for sending radio parameters and non-access stratum (NAS) signaling; the N3 interface is the interface between the RAN and UPF network elements, used for transmitting user plane data; the N4 interface is the interface between the SMF network element and the UPF network element, used for transmitting information such as service policies, tunnel identification information of the N3 connection, data cache indication information, and downlink data notification messages. The N6 interface is the interface between the DN network element and the UPF network element, used for transmitting user plane data; Naf is the service-based interface provided by the AF, Nnrf is the service-based interface provided by the NRF, Nnwdaf is the service-based interface provided by the NWDAF, and Nnef is the service-based interface provided by the NEF.

[0106] It should be understood that the above-mentioned network architecture applied to the embodiment of the present application is only an example of the network architecture described from the perspective of traditional point-to-point architecture and service-oriented architecture. The network architecture applicable to the embodiment of the present application is not limited to this. Any network architecture that can realize the functions of the above-mentioned network elements is applicable to the embodiment of the present application.

[0107] It should be understood that Figure 1 The interface names between the various network elements in the embodiment are only examples. The names of the interfaces in the specific implementation may be other names, and this application does not specifically limit this. In addition, the names of the messages (or signaling) transmitted between the above-mentioned network elements are only examples and do not constitute any limitation on the function of the messages themselves.

[0108] It should be noted that the above-mentioned network element may also be referred to as an entity, device, apparatus, or module, etc., and this application does not specifically limit this. Furthermore, in this application, for ease of understanding and explanation, the description of the network element is omitted in some descriptions. For example, the NWDAF network element is referred to as NWDAF. In this case, the "NWDAF" should be understood as the NWDAF network element. The following descriptions of the same or similar situations are omitted.

[0109] It is understandable that the above-mentioned functional network elements can be network elements in hardware devices, software functions running on dedicated hardware, or virtualized functions instantiated on a platform (e.g., a cloud platform).

[0110] It is understandable that in Figure 1 In the communication system shown, the functions of each network element are merely exemplary, and not all functions of each network element are required when applied in the embodiments of the present application.

[0111] It should also be noted that, in the embodiment of the present application, the core network elements in the above network elements come from at least two different equipment manufacturers.

[0112] In order to establish a business experience model, it is necessary to combine relevant data from multiple data sources for training. However, due to the reluctance of different data sources to share their private data, it is difficult to obtain complete data to establish an accurate and effective business experience model. In order to avoid leaking privacy and affecting data compliance, and to protect data privacy and security, it is possible to consider using machine learning methods to extract the characteristics of the private network data of each data source and establish a virtual shared model. In this way, the characteristics of the data can be obtained through the model without obtaining the original data of the data source. The above-mentioned machine learning methods include but are not limited to federated learning methods. Federated learning methods can ensure the information security of data exchange between all parties, protect the privacy of terminal data and personal data, and ensure legality and compliance. Under the premise of ensuring that the original data does not leave the local area, the information and model parameters are encrypted and exchanged, thereby realizing the establishment of a business experience model. The established business experience model only serves local goals.

[0113] It should be noted that federated learning is divided into horizontal federated learning, vertical federated learning, and federated transfer learning. Vertical federated learning is suitable for situations where participants' training sample IDs overlap significantly but data features overlap less. In vertical federated learning, each participant's training data is vertically divided and federated learning is performed by combining the different data features of common samples from multiple participants. Vertical federation increases the feature dimensionality of training samples. Horizontal federated learning is suitable for situations where participants' data features overlap significantly but sample IDs overlap less. In horizontal federated learning, each participant's training data is horizontally divided and federated learning is performed by combining multiple rows of samples with the same features from multiple participants. Consequently, horizontal federated learning increases the total number of training samples.

[0114] Figure 2 The following is a schematic diagram of the federated learning system framework. The system architecture consists of two parts: encrypted sample alignment and encrypted model training. After determining the shared user group, this data can be used to train the machine learning model. To ensure data confidentiality during training, encrypted training is performed with the help of a third-party collaborator C. Taking the linear regression model as an example, the training process can be divided into the following four steps:

[0115] Step 1, collaborator C initializes the model parameters Θ A 、Θ B Distributed to A and B to encrypt the data that needs to be exchanged during training.

[0116] In step 2, A and B exchange the intermediate results of the gradient calculation in encrypted form. The specific process is as follows:

[0117] Client A has a dataset Client B has a data set. Where yi is the label data, then the model to be trained is as follows:

[0118]

[0119] We assume that the objective function for linear regression is:

[0120] Where L is the loss function, as follows:

[0121]

[0122] Since the original data D on A and B A and D B It is impossible to aggregate them together and cannot be based on traditional centralized training methods. The training method based on vertical federation is as follows:

[0123] make Then the L transformation is as follows:

[0124]

[0125] make

[0126] So

[0127] L=L A +L B +L AB Formula 4

[0128] make Then L is about Θ A and Θ B The gradient is as follows:

[0129]

[0130] Accordingly, the model parameters are updated as follows:

[0131]

[0132] The training process of vertical federated learning is as follows:

[0133] (1) Client A and Client B determine the initialization model parameters Θ A and Θ B ;

[0134] (2) ClientA is based on Θ A calculate and L A , and then send it to Client B;

[0135] (3)ClientB is based on ΘB calculate Further based on and y i Calculate d i 、L AB 、L B , and finally based on L A 、L AB 、L B Calculate L. Client B will i Sent to Client A.

[0136] In step 3, A and B perform calculations based on the encrypted gradient values ​​respectively, while B calculates the loss based on its label data and summarizes these results to C. C calculates the total gradient by summarizing the results and decrypts it.

[0137] In step 4, C passes the decrypted gradient back to A and B respectively, and A and B update the parameters of their respective models based on the gradient.

[0138] Iterate the above steps until the model training end conditions are met, such as the number of iterations reaches a certain threshold (such as 10,000 times) and the loss L is less than a certain threshold (such as 0.001). In this way, the entire training process is completed and the final training model is obtained.

[0139] After the model training is completed, for a new set of data A and B are based on the trained model parameters Θ A and Θ B Compute local inference results as well as Client A then sends the local inference result Sent to Client B, which will finalize the inference result

[0140] During the sample alignment and model training process, the data of A and B are retained locally, and the data in training are encrypted and interacted through model parameters, avoiding data privacy leakage.

[0141] It should be noted that the vertical federated learning method is only one possible training method for implementing the training of the business experience model of the embodiment of this application and should not constitute any limitation to this application. This application does not exclude the future definition of other models or the use of other methods to achieve the ability to train a business experience model with centralized and complete data without violating data privacy regulations. As long as the method can achieve the same or similar functions as mentioned above, it is within the scope of protection of this application.

[0142] It should be noted that in the embodiment of the present application, training a service experience model requires combining private network data and service experience data of the same manufacturer for training. The trained service experience model is used to determine the service experience of the services provided by the core network elements of the manufacturer.

[0143] In a possible implementation, the data involved in training the service experience model may also include public network data of the core network elements of the manufacturer.

[0144] It should be understood that the data used for the above-mentioned business experience model training can be targeted at a specific business of the terminal (such as Tencent video business), or it can be targeted at multiple businesses in a category of business of the terminal (such as Tencent Video, YouTube, etc. in video business). This application does not impose any restrictions on the training data used in the business experience model training process.

[0145] It should be understood that the training of the service experience model in the embodiment of the present application only focuses on the data in the CN domain and the data on the AF that are related to the core of the solution of the present application, but the data actually participating in the training of the service experience model includes data on domains such as UE, RAN, CN and AF. The process in which data in domains such as UE and RAN participate in the training of the service experience model is similar to that in the prior art. In order to avoid redundancy, the specific description of the participation of data in domains such as UE and RAN in the training of the service experience model is omitted in the embodiment of the present application.

[0146] It should be understood that the private network data in the embodiments of the present application includes data that cannot be reported by each network element or non-standardized data. The network element may determine which data cannot be reported to the NWDAF based on data privacy, data volume, or equipment manufacturer policies described by the network element. For example, a base station equipment manufacturer may be unwilling to report RAN private parameters such as energy-saving parameters, positioning parameters, and radio resource management parameters to protect product profits.

[0147] It should be understood that the public network data in the embodiment of the present application includes data or standardized data that can be reported to the NWDAF by each network element, wherein the network element can determine which data can be reported to the NWDAF based on data privacy or data volume or the device manufacturer policy described by the network element. For example, the public network data reported by each network element may include: wireless signal quality reported by the RAN: reference signal received power (RSRP), reference signal received quality (RSRQ), signal to interference plus noise ratio (SINR); QoS related parameters reported by the UPF: flow rate (QoS flow bit rate), flow delay (QoS flow packet delay), flow packet error rate (QoS flow packet error rate); service flow related parameters reported by the AF: application layer buffer size corresponding to the service flow, service experience (service experience) of the service flow.

[0148] It should also be noted that when the NWDAF needs to analyze data on the RAN, CN, and AF, it needs to logically associate the data generated on different network elements when the terminal accesses services.

[0149] Figure 3 Table 1 shows a possible way to correlate the data of a terminal on different network elements. Specifically, the data belonging to the same terminal on every two network elements can be correlated in pairs in sequence using different types of correlation information.

[0150] Table 1: Association information used to associate terminal data on two different network elements

[0151]

[0152] Specifically, as an example but not limited to, the association method can be that AF and UPF use the association information as timestamp (Timestamp) and IP address 5-tuple (IP address 5-tuple) to associate the data belonging to the same terminal on the two network elements in pairs, UPF and SMF use the association information as timestamp (Timestamp) and UE IP to associate the data belonging to the same terminal on the two network elements in pairs, SMF and PCF use the association information as timestamp (Timestamp) and user permanent identifier (SUPI, subscription permanent identifier) ​​to associate the data belonging to the same terminal on the two network elements in pairs, SMF and AMF use the association information as timestamp (Timestamp) and user permanent identifier (SUPI, subscription permanent identifier) ​​to associate the data belonging to the same terminal on the two network elements in pairs, AMF and RAN use the association information as timestamp (Timestamp), RAN UE NGAP ID and RAN global unique identifier (Global RAN Node ID) to associate the data belonging to the same terminal on the two network elements in pairs, RAN and UPF use the association information as timestamp (Timestamp) and AN Tunnel Info associates data belonging to the same terminal on two network elements.

[0153] It should be understood that the action of associating the data of the terminal on different network elements can be performed by the equipment manufacturer's NWDAF after the equipment manufacturer's NWDAF collects the data set and corresponding association information on the equipment manufacturer's core network network element, or it can be performed by the operator's NWDAF after the operator's NWDAF obtains the data set and corresponding association information list or set on the equipment manufacturer's core network network element, or it can be performed by other network elements with the same or similar functions, and this application is not limited here.

[0154] It should also be understood that the data generated by the terminal on different network elements when accessing the service can also be logically associated in other ways, such as associating per UE data between AMF and RAN. In addition to the above-mentioned [RAN UE NGAP ID, Global RAN Node ID, Timestamp], the association information can also be [AMF UE NGAP ID, Global RAN Node ID, Timestamp]. The association information and association order in the above association method are only examples, and this application does not limit this.

[0155] In view of the scenario where network elements within the core network domain are distributed across different vendors, the current 5G network lacks a method for establishing a corresponding service experience model. The private network data and public network data for the terminal that can be used to train the service experience model are stored locally in the network elements within the core network domain. The network elements within the core network domain can obtain their own equipment manufacturer information and can only report their own private network data to the NWDAF of their own equipment manufacturer. The service experience data used to train the model is provided by the AF, but because the AF cannot perceive the manufacturer information of the network equipment through which the terminal's service flow passes in the network, these service experience data do not include the corresponding equipment manufacturer information on the network side. Therefore, it is necessary to first consider how to segment the service experience data provided by the AF according to the equipment manufacturer, so that the network data and service experience data belonging to the same equipment manufacturer can be centralized to train the service experience model required by the equipment manufacturer.

[0156] In response to the above problems, an embodiment of the present application provides a method for training a service experience model, so that the training of the service experience model can be completed even when the network elements within the core network domain are from different manufacturers.

[0157] Specifically, the embodiment of the present application first determines the device manufacturer information of the service experience data provided by the AF. Under this premise, if the NWDAF of each device manufacturer can directly obtain the service experience data corresponding to its own manufacturer, the service experience model can be trained directly in the NWDAF of the device manufacturer; but if the NWDAF of each device manufacturer cannot directly obtain the service experience data corresponding to its own manufacturer, a third-party collaborator can be introduced to cooperate in completing the training of the service experience model. In order to avoid the leakage of privacy data of the core network network element of the device manufacturer when the two parties interact with each other during the training process, a machine learning method can be used at this time to ensure the privacy of the network element data within the core network domain while also ensuring the establishment of the service experience model.

[0158] It should be understood that, taking the operator's NWDAF as an example, when training the model, the private network data on the core network elements only participates in the training on the local equipment manufacturer's NWDAF, while the service experience data can be obtained by the operator's NWDAF and then participate in the training on the operator's NWDAF, or it can be obtained by the NEF or the operator's NWDAF and then sent to the equipment manufacturer's NWDAF for training; similarly, if the public network data on the core network elements participates in the training, it can participate in the training on the local equipment manufacturer's NWDAF, or it can be sent from the equipment manufacturer's NWDAF to the operator's NWDAF for training.

[0159] It should be noted that, in the process of describing the embodiments in conjunction with the accompanying drawings below, the figures are only for illustration to facilitate understanding and should not constitute any limitation to this application. In addition, the figure shows that VendorNWDAF may correspond to the data analysis function network element of the equipment manufacturer, 5GC NF corresponds to the internal network element of the 5G core network domain, such as AMF, SMF, UPF, PCF, etc., NRF may correspond to the network storage function network element, AF represents the application function network element, and operator NWDAF represents the operator's data analysis network element. The names of the network elements are only defined to distinguish different functions and should not constitute any limitation to this application. This application does not exclude the possibility of defining other network elements to implement the same or similar functions.

[0160] Figure 4 A schematic flow chart of a training method 100 for a service experience model of the present application is shown.

[0161] When the core network elements come from different equipment manufacturers, the specific method for determining the service experience model of each equipment manufacturer includes S110 to S120, and each step is described in detail below.

[0162] S110 , determining device manufacturer information corresponding to service experience data generated when a terminal uses a service.

[0163] Specifically, methods for determining the device manufacturer information corresponding to the service experience data include but are not limited to the following:

[0164] Method A

[0165] The operator's NWDAF obtains the associated information corresponding to the terminal's service experience data on the AF. This associated information is used to associate the network data generated by the terminal on the core network element with the service experience data generated on the AF.

[0166] The operator's NWDAF obtains the device vendor identifier (such as the Vendor ID) corresponding to the core network element and the associated information corresponding to the network data generated by the terminal on the core network element of the device vendor;

[0167] The operator's NWDAF matches the service experience data with the device manufacturer identifier based on the association information, thereby determining the device manufacturer information corresponding to the service experience data.

[0168] Method B

[0169] When registering with the NRF, the NWDAF of the equipment manufacturer carries the associated information corresponding to the network data generated by the terminal on the core network element of the equipment manufacturer and the information of the equipment manufacturer (such as the address information of the equipment manufacturer, the manufacturer identification information of the equipment manufacturer, etc.);

[0170] The operator's NWDAF obtains the associated information corresponding to the terminal's service experience data on the AF. This associated information is used to associate the network data generated by the terminal on the core network element with the service experience data generated on the AF.

[0171] The operator's NWDAF carries the association information and queries the NRF for the device manufacturer information corresponding to the association information, thereby determining the device manufacturer information corresponding to the service experience data through the association information.

[0172] It should be noted that the operator's NWDAF can carry one or multiple pieces of associated information to query the NRF, and this application does not limit this.

[0173] In one possible implementation, the equipment manufacturer's NWDAF may also carry the associated information corresponding to the network data generated on the core network element of the equipment manufacturer when the terminal uses the service when registering with other network elements with similar functions (such as data collection coordination function network element (DCCF) and data storage function network element (DRF), etc.). Correspondingly, the operator's NWDAF carries the associated information to query the equipment manufacturer information corresponding to the associated information to the network element with similar functions (such as data collection coordination function network element DCCF and data storage function network element DRF, etc.). This application does not limit this.

[0174] It should be understood that the DCCF is responsible for coordinating the NWDAF to collect terminal data on the core network elements corresponding to the equipment manufacturer, and the DRF is used to store terminal data on the core network elements corresponding to the equipment manufacturer. In addition, the DCCF or DRF can be deployed within the NWDAF as an internal logical function.

[0175] Method C

[0176] When registering with the NRF, the NWDAF of the equipment manufacturer carries the associated information corresponding to the terminal's network data on the core network element of the equipment manufacturer;

[0177] NEF obtains the associated information corresponding to the terminal's business experience data on AF. The associated information is used to associate the terminal's data on the core network element with the business experience data;

[0178] The NEF carries the associated information and queries the NRF for the device manufacturer information corresponding to the associated information.

[0179] It should be noted that the NEF may carry one or multiple pieces of associated information to query the NRF, which is not limited in this application.

[0180] In one possible implementation, the NWDAF of the equipment manufacturer may also carry the association information corresponding to the network data generated on the core network element of the equipment manufacturer when the terminal uses the service when registering with other network elements with similar functions (such as data collection coordination function network element (DCCF) and data storage function network element (DRF), etc.). Correspondingly, the NEF may carry the association information to query the equipment manufacturer information corresponding to the association information to the network element with similar functions (such as data collection coordination function network element DCCF and data storage function network element DRF, etc.). This application does not limit this.

[0181] It should be understood that the DCCF is responsible for coordinating the NWDAF to collect terminal data on the core network elements corresponding to the equipment manufacturer, and the DRF is used to store terminal data on the core network elements corresponding to the equipment manufacturer. In addition, the DCCF or DRF can be deployed within the NWDAF as an internal logical function.

[0182] In a possible implementation, other network elements with similar functions may also obtain the service experience data set of the terminal's service on the AF and determine the device manufacturer information corresponding to each service experience data in the service experience data set, which is not limited in this application.

[0183] S120 , combining network data of the terminal on the core network element of the equipment manufacturer and corresponding service experience data to train a service experience model when the terminal uses the service on the core network element of the equipment manufacturer.

[0184] It should be noted that the "network data of the equipment manufacturer" or "network data on the core network element of the equipment manufacturer" appearing in the description below can be understood as the network data generated on the core network element of the equipment manufacturer when the terminal uses the service, where the network data includes private network data and / or public network data.

[0185] It should also be noted that the "service experience data corresponding to the equipment manufacturer" appearing in the following description can be understood as the service experience data corresponding to the manufacturer's equipment on the application function network element when the core network element of the manufacturer's equipment provides services to the terminal.

[0186] It should also be noted that the "service experience model of the equipment manufacturer" that appears in the following description can be understood as the service experience model corresponding to the equipment manufacturer when the core network element of the manufacturer's equipment provides services to the terminal.

[0187] It should also be noted that the "terminal" appearing in this application may refer to one and the same terminal or multiple different terminals, and this application does not limit this.

[0188] Specifically, taking one of the equipment manufacturers as an example, methods for training the service experience model of the equipment manufacturer by combining the network data of the core network elements of the equipment manufacturer and the service experience data corresponding to the equipment manufacturer include but are not limited to the following:

[0189] Method 1

[0190] The equipment manufacturer's NWDAF collects private network data of the equipment manufacturer's core network elements;

[0191] The operator's NWDAF obtains the service experience data of the equipment manufacturer from the AF;

[0192] The operator's NWDAF determines the address information of the device manufacturer's NWDAF;

[0193] The operator's NWDAF and the equipment manufacturer's NWDAF train and determine the equipment manufacturer's service experience model based on the private network data of the equipment manufacturer's core network elements and the equipment manufacturer's corresponding service experience data.

[0194] In one possible implementation, the operator's NWDAF can also work with the equipment manufacturer's NWDAF to determine the equipment manufacturer's service experience model based on the private network data, public network data, and service experience data of the equipment manufacturer's core network elements. In this case, the public network data of the equipment manufacturer's core network elements can be sent by the equipment manufacturer's core network elements to the operator's NWDAF for training on the operator's NWDAF; or the equipment manufacturer's core network elements can send the data to the equipment manufacturer's NWDAF, which in turn sends the data to the operator's NWDAF for training on the operator's NWDAF; or the equipment manufacturer's core network elements can send the data to the equipment manufacturer's NWDAF for training locally on the equipment manufacturer's NWDAF.

[0195] It should be noted that if the core network element of the equipment manufacturer sends public network data to the operator NWDAF and participates in training on the operator NWDAF, the public network data of the core network element of the equipment manufacturer can be obtained at the same time in method A of S110 when the operator NWDAF obtains the equipment manufacturer identifier Vendor ID of the core network element of each equipment manufacturer and the associated information corresponding to the network data of the core network element of each equipment manufacturer, so as to save a certain amount of resource overhead.

[0196] Method 2

[0197] The equipment manufacturer's NWDAF collects private network data of the equipment manufacturer's core network elements;

[0198] The NWDAF of the device manufacturer obtains the service experience data corresponding to the device manufacturer;

[0199] The NWDAF of the equipment manufacturer is trained based on the private network data of the core network elements of the equipment manufacturer and the service experience data corresponding to the equipment manufacturer to determine the service experience model of the equipment manufacturer.

[0200] It should be noted that the way in which the NWDAF of the equipment manufacturer obtains the service experience data corresponding to the equipment manufacturer includes but is not limited to, the NEF sends the service experience data of the equipment manufacturer to the NWDAF of the equipment manufacturer (wherein, the service experience data of the equipment manufacturer on the NEF comes from the AF), or the operator's NWDAF sends the service experience data of the equipment manufacturer to the NWDAF of the equipment manufacturer.

[0201] In one possible implementation, the device manufacturer's NWDAF can also train and determine the device manufacturer's service experience model based on the private network data and public network data of the device manufacturer's core network elements, as well as the device manufacturer's corresponding service experience data. In this case, Method 2 should also include the device manufacturer's NWDAF collecting public network data from the device manufacturer's core network elements.

[0202] As an example and not a limitation, four specific embodiments for implementing service experience model training in a cross-vendor scenario of core network elements are given below.

[0203] In the method of the embodiment of the present application, the data analysis network element NWDAF includes the operator's data analysis network element NWDAF and the core network element equipment manufacturer's data analysis network element NWDAF. The cross-vendor network elements within the core network domain can span two equipment manufacturers or multiple equipment manufacturers. As an example and not a limitation, the method of this embodiment is described as a case where the core network elements belong to two equipment manufacturers.

[0204] As mentioned previously, if a device manufacturer's NWDAF cannot directly access the vendor's service experience data, a third-party collaborator can be brought in to help train the service experience model. Specifically, this third-party collaborator can be a carrier's NWDAF. The following uses a carrier's NWDAF as an example for detailed explanation.

[0205] Figure 5 A schematic flow chart of a method 200 for acquiring a service experience model according to a first specific embodiment of the present application is shown.

[0206] like Figure 5As shown, to obtain the service experience model, it is first necessary to determine the device manufacturer information of the service experience data. The specific determination method includes S201a to S204. Then, the network data of the core network element of the device manufacturer and the corresponding service experience data are combined to train the service experience model of the device manufacturer. The specific training steps include S205a to S207. Each step is described in detail below.

[0207] S201a: The operator's NWDAF sends a request message #a to the AF to subscribe the terminal's service experience data and associated information corresponding to the service experience data to the AF.

[0208] Specifically, in one possible implementation, the operator's NWDAF triggers an event open subscription (Naf_EventExposure_Subscribe) service operation on the Naf interface of the AF. The service operation carries Event ID = service experience information (Service Experience), Event Filter = application identifier or service identifier (Application ID), which is used to subscribe to the service experience data of the service identified by the Application ID to the AF, and to subscribe to the service experience data of the terminal and the associated information corresponding to the service experience data to the AF.

[0209] It should be noted that the operator NWDAF can subscribe to a terminal service experience data and the associated information corresponding to the service experience data from the AF, or can subscribe to multiple terminal service experience data and the associated information corresponding to the service experience data set from the AF at the same time. This application does not limit this.

[0210] It should also be noted that the above-mentioned business experience data may specifically include one or more business experience data, wherein one business experience data corresponds to one associated information. Here, one business experience data may correspond to the business experience data generated by a terminal accessing a certain business, or may correspond to the business experience data generated by a terminal accessing a class of business (including multiple businesses). This application does not limit this.

[0211] It should also be noted that the service experience data may include the service experience and / or terminal location of the terminal at a specific timestamp when using the service, where the service experience may be one or more of the following types: mean opinion score (MOS), round trip time (RTT), bandwidth, jitter, etc.

[0212] S201b: The AF sends a reply message #a to the operator NWDAF. The reply message #a includes the service experience data subscribed by the operator NWDAF and the associated information corresponding to the service experience data set.

[0213] Specifically, in a possible implementation, the AF triggers an event exposure notification (Naf_EventExposure_Notify) service operation on the Naf interface of the operator's NWDAF, and sends the service experience data of the terminal and associated information corresponding to the service experience data to the operator's NWDAF.

[0214] It should be noted that the above-mentioned association information is used to associate the service experience data generated by a terminal on the AF when accessing the service with the network data on the core network element. The service experience data and the network data correspond to the same equipment manufacturer. Specifically, the association information can first associate the service experience data generated by a terminal on the AF when accessing the service with the network data generated on the UPF. At this time, the association information may include: timestamp and IP address 5-tuple, and then associate the data on the SMF, AMF or AN according to the method described above.

[0215] S202a, the operator's NWDAF sends a request message #b1 to the 5GC NF#1 of the equipment manufacturer #1, and subscribes to the public network data #1 on the network element of the equipment manufacturer and the associated information #1 corresponding to the public network data from the 5GC NF#1 of the equipment manufacturer #1. For example, it subscribes to the terminal location information from the AMF through Event ID = terminal location (UE Location), and subscribes to QoS flow related data (such as DNN, S-NSSAI, QFI, QoS flow bit rate (QoS flow bit rate), QoS delay (QoS flow packet delay), number of packet transmissions (packet transmission), number of packet retransmissions (packet retransmission), etc.) from the SMF through Event ID = QFI allocation, and instructs 5GC NF#1 to carry the equipment manufacturer identifier Vendor ID#1 of 5GC NF#1 when reporting public network data #1.

[0216] Specifically, in one possible implementation, the operator's NWDAF triggers the event open subscription (Nnf_EventExposure_Subscribe) service operation on the Nnf interface of 5GC NF#1, subscribes to the public network data #1 and the associated information #1 of the terminal information corresponding to the public network data #1 on the network element of the device manufacturer #1 to the 5GC NF#1 of the device manufacturer #1, and instructs 5GC NF#1 to carry the device manufacturer identifier VendorID#1 of 5GC NF#1 when reporting the public network data #1.

[0217] S202b, 5GC NF#1 sends a reply message #b1 to the operator NWDAF. The reply message #b1 includes the public network data #1 subscribed by the operator NWDAF, the associated information #1 of the terminal information corresponding to the public network data #1, and the device vendor ID #1 of 5GC NF#1.

[0218] Specifically, in one possible implementation, 5GC NF#1 triggers the event exposure notification (Nnf_EventExposure_Notify) service operation on the Nnf interface of the operator NWDAF, and sends the public network data #1 subscribed by the operator NWDAF in step S202a, the associated information #1 corresponding to the public network data #1, and the device manufacturer identifier Vendor ID#1 of 5GC NF#1 to the operator NWDAF.

[0219] It should be understood that different network elements in the 5GC NF may carry different types of associated information when reporting public network data. Specifically, Figure 3 Table 1 shows a possible correspondence between network elements and associated information types.

[0220] In step S202c, the NWDAF of device manufacturer #1 collects private network data on each network element from 5GC NF #1.

[0221] In one possible implementation, Vendor NWDAF#1 triggers the Nnf_EventExposure_Subscribe service operation on the Nnf interface of 5GC NF#1, subscribing to 5GC NF1 for private network data and associated information on the core network element of Device Vendor #1. 5GC NF#1 triggers the Nnf_EventExposure_Notify service operation on the Nnf interface of Vendor NWDAF#1, sending each terminal's public network data and private network data, as well as associated information, to Vendor NWDAF#1.

[0222] It should be understood that VendorNWDAF#1 can subscribe to the private network data and corresponding associated information of one terminal from 5GC NF#1, or can subscribe to the private network data and corresponding associated information of multiple terminals from 5GC NF#1 at the same time, which is not limited in the embodiment of the present application.

[0223] It should be noted that the data corresponding to different network elements in 5GC NF may have different associated information. Figure 3 Or Table 1 shows a possible correspondence between network element data and associated information, wherein the associated information of the terminal data on the UPF may include: a timestamp and an IP address 5-tuple.

[0224] In a possible implementation, the NWDAF of device manufacturer #1 may also collect private network data on the core network elements of device manufacturer #1 through other means (such as hardware probes).

[0225] S202d, the device manufacturer NWDAF#1 registers with the NRF with its own vendor ID.

[0226] Specifically, the way for the equipment manufacturer NWDAF#1 to register with the NRF can be that the equipment manufacturer NWDAF#1 carries the registration information to trigger the network element registration request (Nnrf_NFManagement_NFRegister_request) service operation in the network element management on the Nnrf interface, and initiates a registration request to the NRF. The registration information includes the vendor identifier Vendor ID of the equipment manufacturer NWDAF. After receiving the registration request, the NRF stores the registration information of the equipment manufacturer NWDAF. The NRF triggers the network element registration response (Nnrf_NFManagement_NFRegister_response) service operation in the network element management on the Nnrf interface to send a reply message to the equipment manufacturer NWDAF#1, and the registration is successful.

[0227] It should be noted that when the equipment manufacturer NWDAF registers with the NRF, the network element registration request (Nnrf_NFManagement_NFRegister_request) in the registration service operation Nnrf interface on the network element management includes the NF network element information (NF Profile) of the equipment manufacturer NWD AF, that is, the NWDAF network element information (NWDAF Profile). In addition to carrying the vendor identifier Vendor ID to which it belongs, the NF Profile must also carry other basic information, such as one or more of the following information corresponding to the equipment manufacturer N WDAF: network element type, address information, service area, analysis identifier Analytics ID, etc. This is similar to the prior art and will not be elaborated on here.

[0228] Steps S203a to S203d are actions related to the device manufacturer NWDAF#2. For specific steps, please refer to the description of steps S202a to S202d and will not be repeated here.

[0229] S204, the operator NWDAF associates the service experience data corresponding to the association information with the device manufacturer identifier, thereby determining the device manufacturer identifier of each service experience data, and associates the service experience data corresponding to the association information with the public network data, thereby determining the data set belonging to device manufacturer #1, which includes service experience data #1, public network data #1, association information #1 corresponding to public network data #1, and device manufacturer identifier Vendor ID #1, and data of device manufacturer #2, which includes service experience data #2, public network data #2, association information #2 corresponding to public network data #2, and device manufacturer identifier Vendor ID #2.

[0230] Specifically, a possible way to correlate the service experience data and public network data of the same terminal of the same equipment manufacturer is as follows: Figure 3As shown in Table 1, after the operator NWDAF obtains the public network data and corresponding association information on the 5GC NF, the operator NWDAF first associates the service experience data from the AF with the public network data from the UPF according to the association information of the Timestamp and IP address 5-tuple, and then associates the public network data from the UPF with the public network data from the SMF according to the association information of the Timestamp and UE IP. Then, according to the association information of the Timestamp and SUPI, the public network data from the SMF is associated with the public network data from the AMF, and the public network data from the SMF is associated with the public network data from the PCF. Then, according to the association information of the Timestamp, RAN UE NGAP ID and the RAN global unique identifier Global RAN Node ID, the public network data from the AMF is associated with the public network data from the RAN, and according to the association information of the Timestamp and AN Tunnel Info, the public network data from the RAN is associated with the public network data from the UPF.

[0231] If the operator's NWDAF directly trains the service experience model based on the public network data of the core network elements of each equipment manufacturer and the service experience data corresponding to the equipment manufacturer that has been obtained, the obtained service experience model has poor generalization ability. In order to obtain a more accurate service experience model, the private network data of each equipment manufacturer is required to participate in the training of the model. However, for data privacy reasons, the operator's NWDAF cannot directly obtain the private network data of each equipment manufacturer, but the equipment manufacturer's NWDAF can obtain the private network data on each network element of the equipment manufacturer. As an example and not a limitation, the embodiment of the present application adopts a vertical federation method to allow the operator's NWDAF to jointly with the equipment manufacturer's NWDAF, so that the equipment manufacturer's private network data is retained locally in the equipment manufacturer's NWDAF to participate in the training of the model, so as to improve the generalization ability or model performance of the service experience model.

[0232] It should be noted that the vertical federation method is only an example training method in the embodiment of this application. The training method used in the embodiment of this application may also be called by other names. As long as it is possible to jointly train the business experience model with the data of all parties while protecting the privacy of all parties, it is within the scope of protection of this application.

[0233] S205a, the operator NWDAF sends a request message #c to the NRF, querying the NRF for the address of the device manufacturer NWDAF. The request message #c includes the vendor ID of the device manufacturer.

[0234] Specifically, in one possible implementation method, the operator NWDAF triggers the Nnrf interface network element discovery request (Nnrf_NFDiscovery_Request) service operation, sends a request message #c to the NRF, and the request message #c includes the vendor ID of the device manufacturer, and requests the NRF to query the address of the device manufacturer NWDAF corresponding to the vendor ID.

[0235] It should be noted that the request information #c may include the vendor ID of one device manufacturer, or may include the vendor IDs of multiple device manufacturers at the same time (used to request the address of the device manufacturer's NWDAF corresponding to each vendor ID), which is not limited in this embodiment of the present application.

[0236] S205b, the NRF sends a reply message #c to the operator NWDAF, where the reply message #c includes the address of the device manufacturer NWDAF corresponding to the device manufacturer's vendor identifier VendorID.

[0237] Specifically, in one possible implementation, NRF triggers the Nnrf interface network element discovery request response (Nnrf_NFDiscovery_Request Response) service operation, and sends a reply message #c to the operator NWDAF, where the reply message #c includes the address Vendor NWDAF ID of the device vendor NWDAF corresponding to the device vendor identifier Vendor ID.

[0238] In a possible implementation, the operator NWDAF may obtain the device manufacturer's NWDAF address information in steps S202b and S203b. When the device manufacturer's core network element sends data to the operator NWDAF, it also carries the address of the device manufacturer's NWDAF.

[0239] S206, training a business experience model.

[0240] Next, we will take device manufacturer #1 as an example to describe the process of using the vertical federation method to train the service experience model of device manufacturer #1 in the first embodiment. For ease of reading and understanding, the public network data #1 of device manufacturer #1 is used as the data set. Take the private network data #1 of device manufacturer #1 as the data set Taking the service experience data #1 of device manufacturer #1 as data yi (i=1, 2, 3, ..., N), where the subscript i represents the i-th sample data, and the linear regression (LiR) algorithm used in the service experience model as an example, the detailed training process is introduced:

[0241] S206a: The operator NWDAF sends an initial federated learning parameters provisioning message to the equipment manufacturer NWDAF#1 to trigger the vertical federated learning training process.

[0242] Among them, the Initial Federated Learning parameters provisioning message includes algorithm identification information and an associated information list, which is used to determine the data set participating in model training. Specifically, the associated information can be the association information between AF and UPF, that is, the IP quintuple and timestamp.

[0243] In one possible implementation, the message also includes the longitudinal federated training dataset. Initialization model parameters

[0244] S206b, the device manufacturer NWDAF receives the above message sent by the operator NWDAF and initializes the model parameters calculate And send it to the operator's NWDAF through the machine learning model update notification (Nnwdaf_MLModelUpdate_Notify) service operation on the Nnwdaf interface and a list of associated information, which is used to associate and the training data belonging to device manufacturer #1 on the operator's NWDAF;

[0245] It should be understood that the training data belonging to device manufacturer #1 on the operator's NWDAF in this embodiment includes service experience data and public network data corresponding to device manufacturer #1.

[0246] S206c, operator NWDAF according to the data set Initialization model parameters calculate The business experience models that need to be trained are as follows:

[0247]

[0248] Among them, xi Represents the i-th sample data, where It is the public network data distributed on CN in the sample data. is the private network data distributed on CN in the sample data, Θ A and Θ B They are and The corresponding model parameters. h(x) is based on the data And the model parameters Θ A and Θ B Calculation results.

[0249] Operator NWDAF according to and y i Calculate Θ A and Θ B The residual d i And the overall model loss L, where

[0250]

[0251] The operator NWDAF is based on the residual d i Update model parameters Θ A The specific update process is as follows:

[0252]

[0253] The operator's NWDAF sends the residual d to the device manufacturer's NWDAF through the machine learning model distribution notification (Nnwdaf_MLModelProvision_Notify) service operation on the Nnwdaf interface. i and Corresponding association information, auxiliary local model parameters Θ B The update of the association information is used to associate the residual d i Datasets on device manufacturers NWDAF#1 The specific update process is as follows:

[0254]

[0255] At this point, an iterative process of longitudinal model training is completed.

[0256] S206d, the equipment manufacturer NWDAF receives and calculates the residual d i Update model parameters Θ B and send the updated intermediate results to the operator NWDAF The associated information list is used to associate the updated intermediate results and the training data belonging to device manufacturer #1 on the operator's NWDAF;

[0257] S206e, the operator's NWDAF receives the updated intermediate result The operator NWDAF determines the final service experience model and ends the training process if the training end conditions are met. If the training end conditions are not met, the operator repeats the above training steps until the training is completed.

[0258] It should be understood that the above model parameters Θ A and Θ B is a model parameter vector. Specifically, each model parameter vector may include one or more model parameters.

[0259] It should be noted that the above model parameters Θ A and Θ B The update process will be executed in a loop until the operator's NWDAF determines that the service experience model training end conditions have been met, and the service experience model training is completed.

[0260] It should be understood that the end condition of the service experience model training can be set in advance by the operator NWDAF. As an example but not limited to, it can be that the number of model parameter iterations reaches a certain threshold, such as the number of model parameter iterations reaches 10,000 times, or the loss function L (e.g., based on the updated intermediate results as well as ) is less than a certain threshold, such as the value of the loss function is less than 0.001.

[0261] In one possible implementation, the model training termination can be set by the device manufacturer NWDAF itself. In this case, the device manufacturer NWDAF does not need to send the updated intermediate results to the operator NWDAF. With the associated information list, you can determine whether the model training is terminated.

[0262] It should also be noted that the above specific steps are only examples of the first iteration process. In non-first model parameter iteration, the data sent by the operator NWDAF to the device manufacturer NWDAF in S107a is determined by the updated model parameters corresponding to the previous iteration.

[0263] The training process of the service experience model of device manufacturer #2 can refer to the training process of the service experience model of device manufacturer #1, and will not be repeated here.

[0264] In one possible implementation, after the service experience model training of each device manufacturer is completed, if Vendor NWDAF#1 and Vendor NWDAF#2 both report public network data, step S207 can also be executed. The operator NWDAF performs horizontal federated training on the model parameters of the public network data in the service experience model trained by each device manufacturer, or the model parameter gradient of the public network data. This unifies the model parameters of the public network data for different device manufacturers, further improving the generalization capability of the model parameters of the public network data.

[0265] In the first embodiment described above, the NWDAF of each equipment manufacturer carries the equipment manufacturer identifier and registers with the NRF in advance; the operator NWDAF associates the service experience data provided by the AF with the Vendor ID provided by the core network element based on the same association information, thereby determining the equipment manufacturer information of each service experience data; the operator NWDAF then queries the NRF for the Vendor NWDAF ID corresponding to the equipment manufacturer identifier based on the equipment manufacturer identifier sent by the core network element, and then jointly conducts vertical federated training based on the network-side data and service experience data of each equipment manufacturer to train service experience models for different equipment manufacturers.

[0266] In the first embodiment, the model parameters of the public network data in the service experience model are calculated and updated on the operator's NWDAF after the operator obtains the public network data from the core network elements of the equipment manufacturer. In another possible implementation method, the model parameters of the public network data can also be calculated and updated on the equipment manufacturer's NWDAF after the equipment manufacturer's own NWDAF collects the public network data from the core network elements of the equipment manufacturer.

[0267] like Figure 6 As shown, to obtain the service experience model, it is first necessary to determine the device manufacturer information of the service experience data. The specific determination method includes S301a to S305. Then, the network data of the core network network elements of each device manufacturer and the corresponding service experience data are combined to train the service experience model of each device manufacturer. The specific training steps include S306 to S308. Each step is described in detail below.

[0268] Steps S301a-S301b are the same as steps S201a-S201b in the first embodiment, and are not described in detail here.

[0269] S302a: The NWDAF of device manufacturer #1 sends a request message #b1 to the 5GC NF #1 of device manufacturer #1, subscribing to the public network data and private network data of device manufacturer #1 and the corresponding association information from the 5GCNF #1. The association information is used to identify the public network data and private network data.

[0270] Specifically, in one possible implementation, Vendor NWDAF#1 triggers the Nnf_EventExposure_Subscribe service operation on the Nnf interface of 5GC NF#1, and subscribes to the public network data and private network data corresponding to each terminal and the corresponding associated information from 5GC NF#1.

[0271] It should be understood that Vendor NWDAF#1 can subscribe to the public network data and private network data and corresponding associated information corresponding to one terminal from 5GC NF#1, or can subscribe to the public network data and private network data and corresponding associated information corresponding to multiple terminals from 5GC NF#1 at the same time, which is not limited in the embodiments of the present application.

[0272] S302b: 5GC NF#1 of device manufacturer #1 sends reply information #b1 to NWDAF of device manufacturer #1. Reply information #b1 includes public network data and private network data of device manufacturer #1 and corresponding association information. The association information is used to identify the terminal information of the public network data and private network data.

[0273] Specifically, in one possible implementation, 5GC NF#1 triggers an event exposure notification (Nnf_EventExposure_Notify) service operation on the Nnf interface of Vendor NWDAF#1, and sends the public network data and private network data of each terminal and corresponding associated information to Vendor NWDAF#1.

[0274] It should be noted that the data corresponding to different network elements in 5GC NF may have different associated information. Figure 3 Table 1 shows a possible correspondence between network element data and associated information, wherein the associated information of the terminal data on the UPF may include: a timestamp and an IP address 5-tuple.

[0275] In a possible implementation, the NWDAF of device manufacturer #1 may also collect private network data on the core network elements of device manufacturer #1 through other means (such as hardware probes).

[0276] S302c: The NWDAF of device manufacturer #1 registers with the NRF by carrying the associated information corresponding to the data on the core network element of device manufacturer #1.

[0277] Specifically, the way for the equipment manufacturer NWDAF to register with the NRF can be that the equipment manufacturer NWDAF carries the registration information to trigger the network element registration request (Nnrf_NFManagement_NFRegister_request) service operation in the network element management on the Nnrf interface, and initiates a registration request to the NRF. The registration information includes the association information corresponding to the data on the core network network element of the equipment manufacturer #1. After receiving the registration request, the NRF stores the registration information of the equipment manufacturer NWDAF, and the NRF triggers the network element registration response (Nnrf_NFManagement_NFRegister_response) service operation in the network element management on the Nnrf interface to send a reply message to the equipment manufacturer NWDAF, and the registration is successful.

[0278] It should be noted that when the equipment manufacturer NWDAF registers with the NRF, in addition to carrying the associated information corresponding to the data on the core network network element of the equipment manufacturer #1, it also needs to carry other basic information, such as NF Profile (network element information, such as the address of the equipment manufacturer NWDAF), etc. This is similar to the existing technology and will not be elaborated here.

[0279] S303a-S303c is the process in which the NWDAF of device manufacturer #2 collects the public network data and private network data corresponding to the terminal and the corresponding associated information from the 5GC NF #2 of device manufacturer #2, and registers it with the NRF. Please refer to steps S302a-S302c and will not be repeated here.

[0280] S304a, the operator NWDAF sends a request message #c to the NRF, querying the NRF for the VendorNWDAF ID (Vendor NWDAF identifier, or Vendor NWDAF address information) corresponding to the correlation information. The request message #c includes the correlation information corresponding to the service experience data collected by the operator NWDAF.

[0281] It should be noted that one service experience data corresponds to one association information. The operator NWDAF can request the NRF to query the Vendor NWDAF ID corresponding to one association information, or can simultaneously request to query the Vendor NWDAF IDs corresponding to multiple association information. This application does not limit this.

[0282] Specifically, in one possible implementation, the operator NWDAF triggers the Nnrf interface network element discovery request (Nnrf_NFDiscovery_Request) service operation, sends a request message #c to the NRF, and requests the NRF to query the Vendor NWDAF ID corresponding to each associated information. The request message #c includes the associated information of the service experience data collected by the operator NWDAF, namely, the Timestamp and IP address 5-tuple.

[0283] S304b, the NRF sends a reply message #c to the vendor NWDAF, where the reply message #c includes the vendor NWDAF ID corresponding to the association information.

[0284] Specifically, in a possible implementation, the NRF triggers a network element discovery response (Nnrf_NFDiscovery_Request Response) service operation on the Nnrf interface, and sends a reply message #c to the operator NWDAF. The reply message #c includes the Vendor NWDAF ID corresponding to each association information.

[0285] S305 , the operator's NWDAF determines the device manufacturer information of the service experience data based on the correlation information, thereby determining the service experience data of device manufacturer # 1 and the service experience data of device manufacturer # 2 .

[0286] The operator NWDAF currently obtains the service experience data of equipment manufacturer #1 and equipment manufacturer #2. The VendorNWDAF currently obtains the private network data and public network data of its corresponding equipment manufacturer. In order to ensure the generalization ability of the service experience model and train a more accurate service experience model, the private network data, public network data and service experience data of the equipment manufacturer can be combined to participate in the training of the service experience model corresponding to the equipment manufacturer. However, for data privacy reasons, the operator NWDAF cannot directly obtain the private network data corresponding to the equipment manufacturer. As an example and not a limitation, the embodiment of the present application adopts a vertical federation method to allow the operator NWDAF to jointly with the equipment manufacturer's NWDAF, so that the equipment manufacturer's private network data is retained locally in the equipment manufacturer's NWDAF to participate in the training of the service experience model corresponding to the equipment manufacturer, so as to improve the generalization ability of the service experience model.

[0287] It should be noted that the vertical federation method is only an example training method in the embodiment of this application. The training method used in the embodiment of this application may also be called by other names. As long as it is possible to jointly train the business experience model with the data of all parties while protecting the privacy of all parties, it is within the scope of protection of this application.

[0288] S306: The operator's NWDAF and the vendor's NWDAF jointly train the service experience model for each equipment manufacturer.

[0289] Next, we will take device manufacturer #1 as an example to describe the process of using the vertical federation method to train the service experience model of device manufacturer #1 in the second embodiment. For ease of reading and understanding, the public network data #1 of device manufacturer #1 is used as the data set. Take the private network data #1 of device manufacturer #1 as the data set Take the service experience data #1 of device manufacturer #1 as data y i Taking the business experience model as a linear regression model as an example, the training process is introduced:

[0290] The business experience models that need to be trained are as follows:

[0291]

[0292] Among them, x i Represents the i-th sample data, where It is the public network data distributed on CN in the sample data. is the private network data distributed on CN in the sample data, Θ A and Θ B They are and The corresponding model parameters.

[0293] S306a: The operator NWDAF sends an initial federated learning parameters provisioning message to the equipment manufacturer NWDAF#1 to trigger the vertical federated learning training process.

[0294] The Initial Federated Learning parameters provisioning message includes algorithm identification information and a list of associated information. The associated information list is used to determine the datasets involved in model training. Specifically, the associated information can be the association information between the AF and the UPF, namely the IP quintuple and the timestamp. The algorithm identification is used to determine the algorithm used for vertical federated learning, such as linear regression or neural network.

[0295] In a possible implementation, the algorithm information also includes a data set and datasets Initialization model parameters and

[0296] S306b, the device manufacturer NWDAF receives the above message sent by the operator NWDAF and initializes the model parameters and calculate and And send it to the operator's NWDAF through the machine learning model update notification (Nnwdaf_MLModelUpdate_Notify) service operation on the Nnwdaf interface and a list of associated information, which is used to associate Business experience data with device manufacturer #1;

[0297] S306c, operator NWDAF according to and y i Calculate Θ A and Θ B The residual d i And the overall model loss L, where

[0298]

[0299] The operator's NWDAF sends the residual d to the device manufacturer's NWDAF through the machine learning model update notification (Nnwdaf_MLModelUpdate_Notify) service operation on the Nnwdaf interface. i and a list of associated information, which is used to associate the residual d i Network-side data with device manufacturer #1;

[0300] S306d, the equipment manufacturer NWDAF receives and calculates the residual d i Update Θ A and Θ B The model parameters of , the specific update process is as follows:

[0301]

[0302] At this point, an iterative process of longitudinal model training is completed.

[0303] The device manufacturer's NWDAF sends the updated intermediate results to the operator's NWDAF through the machine learning model update notification (Nnwdaf_MLModelUpdate_Notify) service operation on the Nnwdaf interface. The associated information list is used to associate the updated intermediate results Training data from the operator's NWDAF belonging to device manufacturer #1

[0304] In a possible implementation, the device manufacturer NWDAF may also send the model parameter Θ to the operator NWDAF. A The corresponding model parameter gradient, sample number and corresponding association information list are used for the horizontal federation training of the operator's NWDAF on the parameters related to the public network data. The association information list is used to associate Θ A Corresponding model parameter gradients, number of samples, and device manufacturer #1;

[0305] S306e, the operator's NWDAF receives the updated intermediate result and the intermediate results obtained from local calculations And determine whether the end conditions of service experience model training are met. If the training end conditions are met, the operator's NWDAF determines the final service experience model and ends the training process; if the training end conditions are not met, repeat the above training steps until the training is completed;

[0306] It should be understood that the above model parameters Θ A and Θ B is a model parameter vector. Specifically, each model parameter vector may include one or more model parameters.

[0307] It should be noted that the above model parameters Θ A and Θ B The update process will be executed in a loop until the operator's NWDAF determines that the service experience model training end conditions have been met, and the service experience model training is completed.

[0308] It should be understood that the end condition of the service experience model training can be set in advance by the operator NWDAF. As an example but not limited to, it can be that the number of model parameter iterations reaches a certain threshold, such as the number of model parameter iterations reaches 10,000 times, or the loss function L (e.g., based on the updated intermediate results as well as ) is less than a certain threshold, such as the value of the loss function is less than 0.001.

[0309] In one possible implementation, the model training termination can be set by the device manufacturer NWDAF itself. In this case, the device manufacturer NWDAF does not need to send the updated intermediate results to the operator NWDAF. With the associated information list, you can determine whether the model training is terminated.

[0310] It should also be noted that the above specific steps are only examples of the first iteration process. In non-first model parameter iteration, the data sent by the operator NWDAF to the device manufacturer NWDAF in S306a is determined by the updated model parameters corresponding to the previous iteration.

[0311] The training process of the service experience model of device manufacturer #2 can refer to the training process of the service experience model of device manufacturer #1, and will not be repeated here.

[0312] In one possible implementation, after the service experience model training of each device manufacturer is completed, steps S307 to S308 can also be executed. The operator's NWDAF performs horizontal federated training on the model parameters of the public network data or the model parameter gradients of the public network data in the service experience models trained by each device manufacturer, so that the model parameters of the public network data of different device manufacturers are unified, further improving the generalization capability of the model parameters of the public network data.

[0313] S307: The operator's NWDAF performs horizontal federation training on the model parameters of public network data or the model parameter gradients of public network data in the service experience models trained by different equipment manufacturers.

[0314] It should be noted that horizontal federated training of model parameter gradients for public network data is an optional step. Its purpose is to unify the model parameters for public network data across different equipment manufacturers and further improve the generalization capability of the model parameters for public network data.

[0315] S308 , the operator's NWDAF sends a processing parameter #d to each device manufacturer's NWDAF. The processing parameter #d is a parameter obtained by the operator's NWDAF after horizontal federation of the model parameter of the public network data or the model parameter gradient of the public network data.

[0316] In the second embodiment mentioned above, VendorNWDAF carries the association information corresponding to the network-side data of each device manufacturer and registers with NRF in advance. Therefore, the operator NWDAF can query the NRF for the VendorNWDAF ID corresponding to the association information based on the association information corresponding to the service experience data sent by AF, thereby determining the device manufacturer information of each service experience data. After that, the operator NWDAF will jointly with the NWDAFs of different device manufacturers to perform vertical federated training based on the network-side data and service experience data of each device manufacturer, and train service experience models for different device manufacturers respectively.

[0317] In the first and second embodiments, after determining the device manufacturer information of the service experience data, the operator NWDAF will retain it locally in the operator NWDAF to participate in the training of the service experience model. As an example and not a limitation, in the third embodiment, after determining the device manufacturer information of the service experience data, the operator NWDAF can also directly send the service experience data corresponding to the device manufacturer to the VendorNWDAF. The device manufacturer can then use the public network data and private network data corresponding to the device manufacturer to train the service experience model directly in the Vendor NWDAF of the device manufacturer.

[0318] Figure 7 A schematic flow chart of a method 400 for acquiring a service experience model according to another specific embodiment of the present application is shown.

[0319] like Figure 7 As shown, to obtain the service experience model, it is first necessary to determine the device manufacturer information of the service experience data. The specific determination method includes S401a to S405. Then, the network data of the core network network elements of each device manufacturer and the corresponding service experience data are combined to train the service experience model of each device manufacturer. The specific training steps include S406a to S410. Each step is described in detail below.

[0320] The method for determining the device manufacturer information of the service experience data S401a to S405 may refer to steps S301a to S305 in the second embodiment, and will not be described in detail here.

[0321] In a possible implementation, the method for determining the device manufacturer information of the service experience data may also refer to the determination method in the first embodiment, or other methods that can associate the service experience data with the device manufacturer information, which is not limited in this application.

[0322] S406a: The operator NWDAF sends a training parameter #d1 to the vendor NWDAF#1. The training parameter #d1 includes the service experience data #1 corresponding to the device manufacturer #1 and the associated information #1 corresponding to the service experience data #1.

[0323] S406b: The operator NWDAF sends a training parameter #d2 to the vendor NWDAF#2. The training parameter #d2 includes the service experience data #2 corresponding to the device manufacturer #2 and the associated information #2 corresponding to the service experience data #2.

[0324] In a third embodiment, the operator's NWDAF determines the device manufacturer information of the service experience data and sends each service experience data to the NWDAF of the corresponding device manufacturer, and participates in the service experience model training on the NWDAF of the device manufacturer.

[0325] At present, each equipment manufacturer's NWDAF has obtained the network data on its own core network network elements and the service experience data corresponding to its own equipment manufacturer. Therefore, there is no need to introduce the vertical federation training method. The equipment manufacturer's NWDAF trains the service experience model based on the network data on the core network network elements corresponding to the equipment manufacturer and the corresponding service experience data. At this time, the service experience model training process S407a and S407b are similar to the existing technology and will not be repeated here.

[0326] In one possible implementation, the operator's NWDAF can also execute steps S408a to S410 to perform horizontal federated training on the model parameters of public network data in the service experience models of various equipment manufacturers, or on the model parameter gradients of public network data, so that the model parameters of public network data for different equipment manufacturers are unified, further improving the generalization capability of the model parameters of public network data.

[0327] At this time, steps S406a and S406b also include the operator NWDAF sending an Initial Federated Learning parameters provisioning message to Vendor NWDAF#1 and Vendor NWDAF#2 to trigger the horizontal federated learning training process. The message includes the algorithm identifier and a list of data types of the public data participating in the training.

[0328] At this time, steps S407a and S407b also include each VendorNWDAF based on the local public data corresponding to the data type list of the public data Perform model training and determine the size of the local training dataset n I And the gradient value of local model training

[0329] In S408a and S408b, each Vendor NWDAF sends the local training dataset size n through the machine learning model update notification (Nnwdaf_MLModelUpdate_Notify) service on the Nnwdaf interface. I And the gradient value of local model training To operator NWDAF.

[0330] S409: The operator NWDAF performs weighted aggregation on the local model gradient values ​​reported by each vendor NWDAF, as follows:

[0331]

[0332] S410: The operator's NWDAF sends processing parameters #f to each device manufacturer's NWDAF to assist each VendorNWDAF in updating the model parameters locally. The processing parameters #f are the gradient results after weighted aggregation. The local model parameter update process is as follows:

[0333]

[0334] At this point, a horizontal federation iteration process is completed.

[0335] It should be noted that the above model parameters Θ I The update process will continue to execute in a loop until the model termination condition of horizontal federated learning is reached, and the above iterative process will terminate.

[0336] Specifically, the model termination condition of horizontal federated learning can be reaching the maximum number of iterations (such as 10,000 times). This termination condition can be set and judged in advance by the operator's NWDAF, or by the equipment manufacturer's NWDAF.

[0337] In the third embodiment, the operator's NWDAF first determines the vendor information of the service experience data provided by the AF, then distributes the service experience data to the corresponding device manufacturer's Vendor NWDAF to assist in training the vendor's internal service experience model. After receiving the service experience data corresponding to the device manufacturer, the vendor NWDAF directly completes the service experience model training on the device manufacturer's internal NWDAF. In this embodiment, vertical federated learning is not required.

[0338] In the third embodiment, after determining the device manufacturer information of the service experience data, the operator NWDAF directly sends the service experience data corresponding to the device manufacturer to the Vendor NWDAF. The device manufacturer then combines the public network data and private network data corresponding to the device manufacturer to train the service experience model directly in the Vendor NWDAF of the device manufacturer. As an example and not a limitation, in the fourth embodiment, the NEF can also determine the device manufacturer information of the service experience data sent by the AF, and then send its corresponding service experience data to each Vendor NWDAF. The operator NWDAF assists in the training of the service experience model within the manufacturer. After receiving the service experience data corresponding to the device manufacturer, the Vendor NWDAF of the device manufacturer directly trains the service experience model on the Vendor NWDAF within the device manufacturer, without the need to introduce vertical federated learning.

[0339] Figure 8 A schematic flow chart of a method 500 for acquiring a service experience model according to another specific embodiment of the present application is shown.

[0340] like Figure 8 As shown, to obtain the service experience model, it is first necessary to determine the device manufacturer information of the service experience data. The specific determination method includes S501a to S505. Then, the network data of the core network network elements of each device manufacturer and the corresponding service experience data are combined to train the service experience model of each device manufacturer. The specific training steps include S506a to S510. Each step is described in detail below.

[0341] S501a: NEF sends a request message #a to AF to subscribe to the service experience data of the terminal and the associated information corresponding to the service experience data set.

[0342] Specifically, in a possible implementation, the NEF triggers an event exposure subscription (Naf_EventExposure_Subscribe) service operation on the Naf interface of the AF, and subscribes the service experience data of the terminal and the associated information corresponding to the service experience data to the AF.

[0343] It should be noted that NEF can subscribe to a terminal service experience data and the associated information corresponding to the service experience data from AF, or can subscribe to multiple terminal service experience data and the associated information corresponding to the service experience data from AF at the same time. This application does not limit this.

[0344] It should also be noted that one service experience data corresponds to one piece of associated information. Here, one service experience data can be for a specific service of the terminal (such as Tencent video service), or it can be for multiple services in a category of services of the terminal (such as Tencent Video, YouTube, etc. in video services). This application does not limit this.

[0345] S501b, the AF sends a reply message #a to the NEF. The reply message #a includes the service experience data subscribed by the operator NWDAF and the associated information corresponding to the service experience data.

[0346] Specifically, in a possible implementation, the AF triggers an event exposure notification (Naf_EventExposure_Notify) service operation on the Naf interface of the NEF, and sends the service experience data of the terminal and the associated information corresponding to the service experience data to the NEF.

[0347] Steps S502a to S503c may refer to steps S402a to S403c in the third embodiment, and will not be described in detail here.

[0348] S504a, the NEF sends a request message #c to the NRF, querying the NRF for the VendorNWDAF ID corresponding to each piece of associated information. The request message #c includes the associated information corresponding to the service experience data.

[0349] It should be noted that one piece of associated information corresponds to one piece of business experience data.

[0350] Specifically, in one possible implementation, NEF triggers the Nnrf interface network element discovery request (Nnrf_NFDiscovery_Request) service operation, sends a request message #c to NRF, and requests NRF to query the VendorNWDAF ID corresponding to the associated information. The request message #c includes the associated information Timestamp and IPaddress 5-tuple corresponding to the service experience data.

[0351] S504b, the NRF sends a reply message #c to the NEF, where the reply message #c includes the VendorNWDAF ID corresponding to the association information.

[0352] Specifically, in a possible implementation, the NRF triggers a network element discovery request response (Nnrf_NFDiscovery_Request Response) service operation on the Nnrf interface, and sends a reply message #c to the NEF. The reply message #c includes the Vendor NWDAF ID corresponding to the association information.

[0353] S505 , the NEF determines the device manufacturer information of the service experience data according to the correlation information, thereby determining the service experience data of device manufacturer # 1 and the service experience data of device manufacturer # 2 .

[0354] S506a , the NEF sends a training parameter #d1 to the Vendor NWDAF#1. The training parameter #d1 includes the service experience data #1 corresponding to the device vendor #1 and the association information #1 corresponding to the service experience data #1.

[0355] S506b: The NEF sends a training parameter #d2 to the Vendor NWDAF#2. The training parameter #d2 includes the service experience data #2 corresponding to the device vendor #2 and the association information #2 corresponding to the service experience data #2.

[0356] In a fourth embodiment, the NEF determines the device manufacturer information of the service experience data and sends the service experience data to the device manufacturer's NWDAF for training on the device manufacturer's NWDAF.

[0357] At present, the equipment manufacturer's NWDAF has obtained the network data on the core network network element corresponding to the equipment manufacturer, as well as the service experience data corresponding to the equipment manufacturer. Therefore, there is no need to introduce a federated training method. The service experience model can be trained directly on the equipment manufacturer's NWDAF based on the network data on the core network network element corresponding to the equipment manufacturer and the corresponding service experience data. At this time, the service experience model training processes S508a and S508b are similar to the existing technology and will not be repeated here.

[0358] In one possible implementation, the operator's NWDAF can also perform steps S509a to S511 to perform horizontal federated training on the model parameters or model parameter gradients of public network data in the service experience models of various device manufacturers. This standardizes the model parameters for public network data across different device manufacturers, further improving the generalization capability of the model parameters for public network data. The specific process of horizontal federated training can be found in the detailed description of the third embodiment and is not detailed here.

[0359] In the fourth embodiment, the Vendor NWDAF registers with the NRF in advance, carrying the associated information corresponding to the device manufacturer's network-side data. After the NEF obtains the service experience data and the corresponding associated information from the AF, it uses the associated information to query the NRF to determine the Vendor NWDAF of the service experience data, and then distributes the service experience data to the Vendor NWDAF of the device manufacturer. After receiving the service experience data corresponding to the device manufacturer, the Vendor NWDAF does not need to introduce vertical federated learning and directly trains the service experience model on the Vendor NWDAF of the device manufacturer. The operator NWDAF assists the Vendor NWDAF in training the service experience model corresponding to the device manufacturer.

[0360] It should be understood that the specific examples in the embodiments of the present application are only intended to help those skilled in the art better understand the embodiments of the present application, rather than to limit the scope of the embodiments of the present application.

[0361] It should also be understood that the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0362] It should also be understood that in the various embodiments of the present application, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced to each other, and the technical features in different embodiments can be combined to form new embodiments according to their internal logical relationships.

[0363] It should also be noted that in the embodiments of the present application, "pre-setting", "pre-configuration", etc. can be achieved by pre-saving corresponding codes, tables or other methods that can be used to indicate relevant information in a device (for example, a network device). This application does not limit its specific implementation method, such as the preset rules and preset constants in the embodiments of the present application.

[0364] It can be understood that, in the above embodiments of the present application, the method implemented by the communication device can also be implemented by components (such as chips or circuits) that can be configured inside the communication device.

[0365] Above, combined Figures 4 to 8 The communication method provided by the embodiment of the present application is described in detail. The above communication method is mainly introduced from the perspective of interaction between various network elements. It can be understood that, in order to implement the above functions, each network element includes a hardware structure and / or software module corresponding to the execution of each function. Those skilled in the art should be aware that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software driven hardware manner depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0366] The following, combined Figures 9 to 11 The communication device provided in the embodiment of the present application is described in detail. It should be understood that the description of the device embodiment corresponds to the description of the method embodiment, so that the content not described in detail can be referred to the above method embodiment. For the sake of brevity, some content will not be repeated.

[0367] In the embodiment of the present application, the functional modules of the transmitting device or the receiving device can be divided according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present application is schematic and is only a logical functional division. There may be other division methods in actual implementation. The following is an example of dividing each functional module according to each function.

[0368] Figure 9A schematic structural diagram of a communication device 600 is provided. The communication device includes a receiving unit 610, a processing unit 620, and a sending unit 630. The communication device 600 can be the data analysis device or network capability exposure device in the above method embodiments, or can be a chip for implementing the functions of the data analysis device or network capability exposure device in the above method embodiments.

[0369] It should be understood that the communication device 600 may correspond to the data analysis device in the methods 200 to 500 or the network capability exposure device in the method 500 according to the embodiments of the present application, and the communication device 600 may include a device for executing Figures 5 to 8 Data analysis equipment in Figure 8 Furthermore, each unit in the communication device 600 and the above-mentioned other operations and / or functions respectively implement the method executed by the network capability opening device. Figures 5 to 8 The corresponding processes of method 200 to method 500 in .

[0370] In one possible design, the communication device 600 may implement Figures 4 to 8 Any function possessed by the operator data analysis device NWDAF in any embodiment shown in any figure.

[0371] For example, the receiving unit 610 is configured to receive first information, where the first information includes service experience data of the terminal on the application function device and corresponding associated information;

[0372] The receiving unit 610 is further configured to receive second information, where the second information includes associated information of network data of the terminal on the core network device;

[0373] In a possible implementation, the second information may further include a device manufacturer identifier of the core network device and / or network data of the terminal on the core network device;

[0374] The receiving unit 610 is further configured to receive third information, where the second information includes address information of the device manufacturer data analysis device NWDAF;

[0375] Processing unit 620 is configured to determine the device vendor information of the service experience data based on the association information and / or the device vendor identifier, determine an address of a device vendor data analysis device NWDAF based on the third information, and jointly determine, with the device vendor data analysis device NWDAF, a service experience model of the device vendor for the terminal based on the first information and the second information;

[0376] A sending unit 630 is configured to send a request message, where the request message may be used to request the first information and / or the second information and / or the third information;

[0377] The sending unit 630 is further configured to send training parameters in the process of determining the service experience model, such as training algorithm information of the service experience model, residual values ​​in model training, etc.

[0378] For another example, the receiving unit 610 is configured to receive first information, where the first information includes service experience data of the terminal on the application function device and corresponding associated information;

[0379] The receiving unit 610 is further configured to receive third information, where the second information includes address information of the device manufacturer data analysis device NWDAF;

[0380] Processing unit 620 is configured to determine, based on the first information and the third information, the device manufacturer information of the service experience data and the address of the device manufacturer data analysis device NWDAF, and to collaborate with the device manufacturer data analysis device NWDAF to determine, based on the first information and the third information, the device manufacturer's service experience model for the terminal;

[0381] A sending unit 630 is configured to send a request message, where the request message may be used to request the first information and / or the third information;

[0382] The sending unit 630 is further configured to send the service experience data corresponding to the device manufacturer to the device manufacturer data analysis device NWDAF;

[0383] The sending unit 630 is further configured to send training parameters in the process of determining the service experience model, such as training algorithm information of the service experience model.

[0384] In another possible design, the communication device 600 may implement Figures 4 to 8 Any function possessed by the device manufacturer data analysis device NWDAF in any of the embodiments shown in any of the figures.

[0385] For example, the receiving unit 610 is configured to receive fourth information, where the fourth information includes network data of the terminal on the core network device and corresponding associated information;

[0386] In a possible implementation, the receiving unit 610 is further configured to receive fifth information, where the fifth information includes service experience data of the terminal on the application function device and corresponding associated information;

[0387] The processing unit 620 is configured to jointly determine, with the operator NWDAF, the service experience model of the device manufacturer for the terminal based on the fourth information, or to jointly determine, with the device manufacturer data analysis device NWDAF, the service experience model of the device manufacturer for the terminal based on the fourth information and the fifth information;

[0388] The sending unit 630 is used to send training parameters in the process of determining the service experience model, such as the model parameter gradient of the public network data in the service experience model, the product of the initial model parameters and the original network data, and the updated model parameters.

[0389] In another possible design, the communication device 600 may implement Figure 8 Any function possessed by the network capability exposure device in the embodiment shown in .

[0390] For example, the receiving unit 610 is configured to receive first information, where the first information includes service experience data of the terminal on the application function device and corresponding associated information;

[0391] The receiving unit 610 is further configured to receive third information, where the second information includes address information of the device manufacturer data analysis device NWDAF;

[0392] A processing unit 620 is configured to determine the device manufacturer information of the service experience data based on the first information and the third information, and further configured to determine an address of a device manufacturer data analysis device NWDAF based on the third information;

[0393] A sending unit 630 is configured to send a request message, where the request message may be used to request the first information and / or the third information;

[0394] The sending unit 630 is further configured to send the service experience data corresponding to the device manufacturer to the device manufacturer data analysis device NWDAF.

[0395] Figure 10 A schematic diagram of the structure of a communication device 700 is provided. The communication device includes a receiving unit 710 and a sending unit 720. The communication device 700 can be the application function device, core network device, or network storage function device in the above method embodiments. Alternatively, the communication device 700 can be a chip used to implement the functions of the application function device, core network device, or network storage device in the above method embodiments.

[0396] In one possible design, the communication device 700 may implement Figures 4 to 8 Any function possessed by the application function device in any of the embodiments shown in any figure.

[0397] For example, the receiving unit 710 is configured to receive request information, where the request information is used to request service experience data of the terminal on the application function device;

[0398] The sending unit 720 is configured to send the service experience data of the terminal on the application function device.

[0399] In another possible design, the communication device 700 may implement Figures 4 to 8 Any function possessed by the core network device in any of the embodiments shown in any figure.

[0400] For example, the receiving unit 710 is configured to receive request information, where the request information is used to request network data of the terminal on the core network device;

[0401] The sending unit 720 is configured to send the network data of the terminal on the core network device.

[0402] In another possible design, the communication device 700 may implement Figures 4 to 8 Any function possessed by the network storage device in any embodiment shown in any figure.

[0403] For example, the receiving unit 710 is configured to receive a request message for requesting the address information of the network analysis device NWDAF of the device manufacturer, wherein the request message includes the manufacturer identifier of the device manufacturer and / or the associated information corresponding to the device manufacturer;

[0404] The sending unit 720 is configured to send the address information of the network analysis device NWDAF of the device manufacturer according to the manufacturer identifier of the device manufacturer and / or the associated information corresponding to the device manufacturer.

[0405] Figure 11 It is a structural block diagram of a communication device 800 provided according to an embodiment of the present application. Figure 11 The communication device 800 shown includes a processor 810, a memory 820, and a communication interface 830. The processor 810 is coupled to the memory and is configured to execute instructions stored in the memory to control the communication interface 830 to send and / or receive signals.

[0406] It should be understood that the processor 810 and memory 820 may be combined into a processing device, and the processor 810 is used to execute the program code stored in the memory 820 to implement the above functions. In specific implementations, the memory 820 may also be integrated into the processor 810 or independent of the processor 810.

[0407] In one possible design, the communication device 800 can be the data analysis device or network capability exposure device in the above method embodiment, or it can be a chip used to implement the functions of the data analysis device or network capability exposure device in the above method embodiment.

[0408] Specifically, the communication device 800 may correspond to the operator data analysis device, the manufacturer data analysis device, and the network capability exposure device in the method 500 according to the embodiment of the present application. The communication device 800 may include a device for performing Figures 4 to 8 Units of methods executed by operators and equipment manufacturers in data analysis equipment, Figure 8 The network capability exposure device in the communication device 800 is a unit of the method executed by the network capability exposure device. Furthermore, each unit in the communication device 800 and the other operations and / or functions described above are respectively for implementing the corresponding processes of methods 200 to 500. It should be understood that the specific process of each unit performing the above corresponding steps has been described in detail in the above method embodiment and will not be repeated here for the sake of brevity.

[0409] When the communication device 800 is a chip, the chip includes a transceiver unit and a processing unit. The transceiver unit may be an input / output circuit or a communication interface; the processing unit may be a processor, microprocessor, or integrated circuit integrated on the chip. This embodiment of the application also provides a processing device including a processor and an interface. The processor may be used to execute the method described in the above method embodiment.

[0410] It should be understood that the processing device may be a chip. For example, the processing device may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on chip (SoC), a central processor unit (CPU), a network processor (NP), a digital signal processor (DSP), a microcontroller unit (MCU), a programmable logic device (PLD), or other integrated chips.

[0411] During implementation, each step of the above method can be completed by an integrated logic circuit of the hardware in the processor or by instructions in the form of software. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware processor, or can be executed by a combination of hardware and software modules in the processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in conjunction with its hardware. To avoid repetition, it will not be described in detail here.

[0412] It should be noted that the processor in the embodiments of the present application can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiment can be completed by an integrated logic circuit of the hardware in the processor or by instructions in the form of software. The above processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component. The various methods, steps, and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0413] It is understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and 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), synchronous-link DRAM (SLDRAM), and direct RAM-bus RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0414] According to the method provided in the embodiment of the present application, the present application also provides a computer program product, which includes: a computer program code, which, when executed on a computer, causes the computer to execute Figure 4 and Figure 8 A method according to any one of the embodiments shown.

[0415] According to the method provided in the embodiment of the present application, the present application also provides a computer-readable medium, which stores a program code, and when the program code is run on a computer, the computer executes Figure 4 and Figure 8 A method according to any one of the embodiments shown.

[0416] According to the method provided in the embodiments of the present application, the present application also provides a system, which includes the aforementioned device or equipment.

[0417] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments may be implemented 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, all or part of the processes or functions described in the embodiments of the present application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. 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 via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a high-density digital video disc (DVD)), or a semiconductor medium (eg, a solid state disc (SSD)).

[0418] The network-side devices in the above-mentioned various apparatus embodiments correspond to the network-side devices or terminal devices in the terminal devices and method embodiments, and the corresponding modules or units perform the corresponding steps. For example, the communication unit (communication interface) performs the receiving or sending steps in the method embodiments, and other steps except sending and receiving can be performed by the processing unit (processor). The functions of the specific units can be referred to in the corresponding method embodiments. There can be one or more processors.

[0419] As used in this specification, the terms "component," "module," "system," and the like are used to refer to computer-related entities, hardware, firmware, a combination of hardware and software, software, or software in execution. For example, a component can be, but is not limited to, a process running on a processor, a processor, an object, an executable file, an execution thread, a program, and / or a computer. By way of illustration, both an application running on a computing device and a computing device can be a component. One or more components can reside in a process and / or an execution thread, and a component can be located on one computer and / or distributed between two or more computers. In addition, these components can be executed from various computer-readable media having various data structures stored thereon. Components can communicate, for example, via local and / or remote processes based on signals having one or more data packets (e.g., data from two components interacting with another component across a local system, a distributed system, and / or a network, such as the Internet interacting with other systems via signals).

[0420] It should also be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " as used herein generally indicates that the associated objects are in an "or" relationship.

[0421] It should also be understood that the numbers "first", "second", "#a", "#b", "#1", "#2", etc., introduced in the embodiments of the present application are only for distinguishing different objects, for example, different "information", or "equipment manufacturers", or "equipment", or "units". The understanding of specific objects and the correspondence between different objects should be determined by their functions and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0422] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0423] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0424] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0425] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0426] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0427] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0428] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A communication method, characterized in that: include: The first data analysis network element obtains service experience data of the terminal on the application function network element and associated information corresponding to the service experience data, where the service is provided by the core network element of the equipment manufacturer; The first data analysis network element determines, according to the association information, address information of a second data analysis network element corresponding to the device manufacturer; The first data analysis network element sends the association information and instruction information to the second data analysis network element according to the address information of the second data analysis network element, where the instruction information is used to instruct the second data analysis network element to perform distributed machine learning model training according to the association information; The first data analysis network element receives the sub-model corresponding to the association information from the second data analysis network element; The first data analysis network element determines a service experience model of the service based on the service experience data and the sub-model.

2. The method according to claim 1, characterized in that The first data analysis network element determines the address information of the second data analysis network element according to the association information, including: The first data analysis network element sends a first request to the network storage function network element, where the first request is used to request address information of the second data analysis network element, and the first request includes the association information; The first data analysis network element receives a first response from the network storage function network element, where the first response includes address information of the second data analysis network element.

3. The method according to claim 1, characterized in that The first data analysis network element determines the address information of the second data analysis network element according to the association information, including: The first data analysis network element determines the identification information of the device manufacturer based on the association information; The first data analysis network element sends a second request to the network storage function network element, where the second request is used to request address information of the second data analysis network element, and the second request includes identification information of the device manufacturer; The first data analysis network element receives a second response from the network storage function network element, where the second response includes address information of the second data analysis network element.

4. The method according to claim 3, characterized in that Before the first data analysis network element determines the identification information of the device manufacturer according to the association information, the method further includes: The first data analysis network element obtains the correspondence between the association information and the identification information of the equipment manufacturer from the core network element corresponding to the equipment manufacturer.

5. The method according to any one of claims 1 to 4, characterized in that The method further comprises: The first data analysis network element receives second network data on the core network element corresponding to the device manufacturer corresponding to the association information, The first data analysis network element determines the service experience model of the service according to the service experience data and the sub-model, including: The first data analysis network element determines a service experience model of the service based on the service experience data, the sub-model and the second network data.

6. The method according to any one of claims 1 to 4, characterized in that The sub-model is determined by the second data analysis network element based on the first network data of the terminal on the core network element of the equipment manufacturer, and the first network data includes the private network data of the service of the terminal on the core network element of the equipment manufacturer.

7. The method according to claim 5, characterized in that The second network data includes public network data of the service of the terminal on the core network element of the equipment manufacturer.

8. A communication method, characterized in that: include: The data analysis network element of the equipment manufacturer obtains the associated information and the first network data of the service of the terminal on the core network network element of the equipment manufacturer; The data analysis network element of the equipment manufacturer receives the correlation information and instruction information from the data analysis network element of the operator, where the instruction information is used to instruct the data analysis network element of the equipment manufacturer to perform distributed machine learning model training according to the correlation information; The data analysis network element of the equipment manufacturer determines a sub-model according to the first network data of the terminal on the core network network element of the equipment manufacturer; The data analysis network element of the equipment manufacturer sends the sub-model to the data analysis network element of the operator, and the sub-model is used to determine a service experience model.

9. The method according to claim 8, characterized in that The method further comprises: The data analysis network element of the equipment manufacturer sends a network element registration request to the network storage function network element, and the network element registration request includes the association information and / or the identification information of the equipment manufacturer.

10. The method according to claim 8 or 9, characterized in that The data analysis network element of the equipment manufacturer jointly determines the service experience model of the service with the data analysis network element of the operator based on the associated information, the first network data and the second network data. The second network data includes the public network data of the service of the terminal on the core network network element of the equipment manufacturer.

11. The method according to claim 8 or 9, characterized in that The first network data includes private network data of the service of the terminal on the core network element of the equipment manufacturer.

12. A communication device, characterized in that: include: A sending unit, configured to send a first request, wherein the first request is used to request service experience data of a service provided by a terminal on an application function device and associated information corresponding to the service experience data; a receiving unit, configured to receive a first response, where the first response includes the service experience data and the associated information, and the service is provided by a core network device of a device manufacturer; A processing unit, configured to determine address information of the data analysis device of the device manufacturer according to the association information; The sending unit is further configured to send the association information and instruction information to the data analysis device of the device manufacturer according to the address information of the data analysis device of the device manufacturer, wherein the instruction information is configured to instruct the data analysis device of the device manufacturer to perform distributed machine learning model training according to the association information; The receiving unit is further configured to receive a sub-model corresponding to the association information from a data analysis device of the device manufacturer, where the sub-model is determined by the data analysis device of the device manufacturer based on first network data of the terminal on the core network device of the device manufacturer; The processing unit is specifically configured to determine a service experience model according to the service experience data and the sub-model.

13. The communication device according to claim 12, wherein: The sending unit is further configured to send a second request to the network storage device, wherein the second request is configured to request address information of a data analysis device of the device manufacturer, and the second request includes the association information; The receiving unit is further configured to receive a second response from the network storage device, where the second response includes address information of the data analysis device of the device manufacturer.

14. The communication device according to claim 12, wherein: The processing unit is further configured to determine identification information of the device manufacturer based on the association information; The sending unit is further configured to send a third request to the network storage device, wherein the third request is configured to request address information of the data analysis device of the device manufacturer, and the third request includes identification information of the device manufacturer; The receiving unit is further configured to receive a third response from the network storage device, where the third response includes address information of the data analysis device of the device manufacturer.

15. The communication device according to claim 12, wherein: The sending unit is further configured to send request information for a correspondence between the association information and the identification information of the device manufacturer to the core network device of the device manufacturer; The receiving unit is further configured to receive a correspondence between the association information sent by the core network device of the device manufacturer and the identification information of the device manufacturer.

16. The communication device according to any one of claims 12 to 15, characterized in that: The receiving unit is further configured to receive second network data on a core network element of the device manufacturer corresponding to the association information; The processing unit is configured to determine the service experience model according to the service experience data, the sub-model, and the second network data.

17. A communication device, characterized in that: include: a sending unit, configured to send a first request, wherein the first request is used to request association information and first network data corresponding to a service of the terminal on a core network device of the device manufacturer, where the service is provided by the core network device of the device manufacturer; a receiving unit, configured to receive a first response, where the first response includes the association information and the first network data corresponding to the service of the terminal on the core network device of the device manufacturer; The receiving unit is further configured to receive the association information and instruction information, wherein the instruction information is used to instruct the data analysis network element of the device manufacturer to perform distributed machine learning model training according to the association information; The processing unit is configured to determine a sub-model according to first network data of the terminal on the core network device of the device manufacturer; The sending unit is further configured to send the sub-model, where the sub-model is used to determine a service experience model.

18. The communication device according to claim 17, wherein: The sending unit is further configured to send a registration request to the network element storage device, where the registration request includes the association information and / or identification information of the device manufacturer.

19. The communication device according to claim 17 or 18, characterized in that The processing unit is used to determine the service experience model of the service in conjunction with the operator's data analysis equipment based on the association information, the first network data and the second network data, wherein the second network data includes the public network data of the service of the terminal on the core network equipment of the equipment manufacturer.

20. The communication device according to claim 17 or 18, characterized in that The first network data includes private network data of the service of the terminal on the core network device of the device manufacturer.

21. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed, causes the apparatus to execute the method according to any one of claims 1 to 7, or causes the apparatus to execute the method according to any one of claims 8 to 11.

22. A chip system, characterized in that: include: A processor for calling and running a computer program from a memory so that a communication device equipped with the chip system executes a method as described in any one of claims 1 to 7, or a communication device equipped with the chip system executes a method as described in any one of claims 8 to 11.

23. A communication device, characterized in that: include: memory for storing computer programs; A processor, configured to execute a computer program stored in the memory, so that the communication device performs the communication method according to any one of claims 1 to 7, or so that the communication device performs the communication method according to any one of claims 8 to 11.

24. A communication system, characterized in that: include: A data analysis network element, configured to execute the communication method according to any one of claims 1 to 7, or to execute the communication method according to any one of claims 8 to 11; and a core network element that communicates with the data analysis network element.

Images