User data analysis reasoning method, electronic device and computer program product

Abstract

The embodiment of the invention provides a user data analysis and reasoning method, an electronic device and a computer program product. The method comprises the following steps: converting an analysis and reasoning request from a consumer network function NF into a request identifier through an NWDAF; the NWDAF obtains an analysis reasoning model and knowledge data according to the request identifier; and the NWDAF performs analysis and reasoning according to the analysis and reasoning model and the knowledge data. The problem of poor expandability of data acquisition of the NWDAF in related technologies is solved, and the effect of improving the expandability of data acquisition of the NWDAF is achieved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of communications, and in particular, to a user data analysis and inference method, an electronic device and a computer program product. BACKGROUND

[0002] In existing core network artificial intelligence technology, a network data analysis function (NWDAF) determines user data to be collected according to an analytics identifier (Analytics ID) carried by a consumer network function (Consumer NF), and then generates analysis or inference and returns it to the Consumer NF. The problem with this approach is that the standard needs to standardize the Analytics ID to be collected from which network functions (NFs) in advance, so the scalability is poor. SUMMARY

[0003] Embodiments of the present application provide a user data analysis and inference method, an electronic device and a computer program product to at least solve the problem of poor scalability of data collection by the NWDAF in related technologies.

[0004] According to an embodiment of the present application, a user data analysis and inference method is provided, comprising: a network data analysis function (NWDAF) converting an analysis and inference request from a consumer network function (NF) into a request identifier; the NWDAF obtaining an analysis and inference model and knowledge data according to the request identifier; and the NWDAF performing analysis and inference according to the analysis and inference model and the knowledge data.

[0005] According to another embodiment of the present application, an electronic device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor being configured to run the computer program to perform the steps in any of the method embodiments described above.

[0006] According to still another embodiment of the present application, a computer program product is also provided, comprising a computer program, which, when executed by a processor, implements the steps in any of the method embodiments described above.

[0007] Through the embodiment of the present application, a user data analysis inference method is provided, wherein the NWDAF converts an analysis inference request from a consumer network function (NF) into a request identifier; the NWDAF acquires an analysis inference model and knowledge data according to the request identifier; and the NWDAF performs analysis inference according to the analysis inference model and the knowledge data. The problem of poor scalability of data collection by the NWDAF in the related art is solved, and the effect of improving the scalability of data collection by the NWDAF is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 is an architecture diagram of 5G basic services in the related art;

[0009] Figure 2 is a flowchart of data collection and analysis inference by the NWDAF in the related art;

[0010] Figure 3 is a hardware structure block diagram of a computer terminal of the user data analysis inference method of the embodiment of the present application;

[0011] Figure 4 is a flowchart of the user data analysis inference method of the embodiment of the present application;

[0012] Figure 5 is a flow of the user data analysis inference method of the embodiment of the present application Figure 1 ;

[0013] Figure 6 is a flow of the user data analysis inference method of the embodiment of the present application Figure 2 ;

[0014] Figure 7 is a flow of the user data analysis inference method of the embodiment of the present application Figure 3 . DETAILED DESCRIPTION

[0015] Hereinafter, the embodiment of the present application will be described in detail with reference to the accompanying drawings and in conjunction with the embodiments.

[0016] It should be noted that the terms “first”, “second”, etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0017] Figure 1 is an architecture diagram of 5G basic services in the related art, as Figure 1 shown, including:

[0018] (1) User equipment (User Equipment, UE).

[0019] (2) Radio Access Network (RAN): The RAN manages radio resources, transmits user data received through N3 to the UE, and transmits user data from the UE through the N3 interface. The RAN maps between Quality of Service (QoS) traffic in Dedicated Radio Bearer (DRB) and Protocol Data Unit (PDU) sessions.

[0020] (3) Access and Mobility Management Function (AMF), which includes the following functions: registration management, connection management, reachability management, and mobility management. This function also performs access authentication and access authorization. AMF is a Non-Access Stratum (NAS) secure terminal that forwards SM NAS, etc., between the UE and the Session Management Function (SMF).

[0021] (4) Session Management Function (SMF): This function includes the following: session establishment, modification, and release; UE IP address allocation and management (including optional authorization functions); UP function selection and control; downlink data notification, etc. The SMF controls the User Plane Function (UPF) through N4. The SMF provides the UPF with Packet Detection Rules (PDR) to indicate how to detect user data traffic; it also provides Forwarding Action Rules (FAR), QoS Enforcement Rules (QER), and Usage Reporting Rules (URR) to indicate how the UPF performs user data traffic forwarding, QoS processing, and usage reporting on user data traffic detected using the PDR.

[0022] (5) User Plane Functions (UPF) include the following functions: serving as an anchor point for movement within / between Radio Access Technology (RAT), packet routing and forwarding, traffic usage reporting, user plane QoS processing, downlink packet buffering, and downlink data notification triggering. A GPRS Tunneling Protocol User Plane (GTP-U) tunnel is used for the N3 interface between the RAN and UPF. GTP-U tunnels operate on a per Protocol Data Unit (PDU) session basis. For downlink traffic, the UPF binds the downlink traffic to the QoS traffic within the PDU session GTP-U tunnel using the AR received from the SMF. For uplink traffic, the RAN transmits user plane traffic to the QoS stream identified by the UE.

[0023] (6) Policy Control Function (PCF): The PCF provides QoS policy rules to control plane functions for enforcement. The PCF translates AF requests into Policy and Charging Control (PCC) rules applicable to PDU sessions.

[0024] (7) Unified Data Management (UDM). The UDM performs 3GPP authentication and key agreement (AKA) credential generation, access authorization based on subscription data, UE service NF registration management (e.g., storing UE storage AMF, storing UE PDU session storage SMF), and subscription management. The UDM accesses the UDR to retrieve UE subscription data and stores the UE context in the UDR. The UDM and the Unified Data Repository (UDR) can be deployed together.

[0025] (8) Network Exposure Function (NEF) The main function is to securely expose the capabilities and events of network functions to third-party applications, edge computing, etc., while providing a way for external applications to provide information to 3GPP network security, and translating information between application functions and internal network functions.

[0026] (9) Network Repository Function (NRF) The main function is to maintain the NF configuration files of available network function instances in the 5G core network and the services they support, and to allow other network functions or SCP instances to subscribe to and receive notifications for the discovery and management of network functions.

[0027] In related technologies, to support the application of artificial intelligence in networks, Network Data Analysis Function (NWDAF) has been introduced. NWDAF is a 5G Core Network Function (5GC NF) located in the control plane, performing statistical data and machine learning tasks within the 5G System (5GS). NWDAF can interact with different entities for various purposes, such as collecting data from AMF, SMF, UPF, PCF, UDM, AF (directly or via NEF), or Operations, Administration, and Maintenance (OAM) using provided event subscriptions. It also stores the collected data in the Analytic Data Repository Function (ADRF) or reads stored data from the ADRF. NWDAF uses AI models and the collected data for computation and returns analysis / inference results to consumers on demand.

[0028] NWDAF can include the following logical functions: Analysis Logic Function (AnLF): A logical function in NWDAF used to perform inference, derive analytical information (i.e., derive statistical data and / or predictions based on analysis consumer requests), and expose analytical services (i.e., Nnwdaf_AnalyticsSubscription or Nnwdaf_AnalyticsInfo). Model Training Logic Function (MTLF): A logical function in NWDAF used to train machine learning (ML) models and expose new training services (such as providing pre-trained ML models).

[0029] In related technologies, an NWDAF can contain an MTLF or an AnLF, or both of these logical functions.

[0030] Figure 2 This is a flowchart of NWDAF's data acquisition and analysis inference process in related technologies, such as... Figure 2 As shown, it includes the following steps:

[0031] Step 1: Consumers of the NWDAF service send an analytics / inference request to NWDAF / AnLF. This request carries a standardized analytics identifier (Analytics ID) and a corresponding analytics context. This context includes the time frame of the analytics (to ensure the timeliness and relevance of the results), the domain name and address identifier (DNAI) (to ensure the analytics results are relevant to a specific region or network slice), and other request parameters. For example, it may include specific QoS requirements or data types to meet the specific needs of the consumer network function (Consumer NF).

[0032] Step 2: NWDAF / AnLF obtains a model request from NWDAF / MTLF. This request includes an ML Model Identifier, which is used to identify the required machine learning model, ensuring that NWDAF uses the correct model for analysis. The Analytics ID is an ID associated with the requested analysis type, helping MTLF provide a suitable model. Model filtering information is used to select the appropriate model version or type, ensuring that the model matches the current analysis requirements.

[0033] Step 3: NWDAF / MTLF matches the corresponding machine learning model and returns machine learning model information. This model information may include the actual model or a link to the model. After receiving it, NWDAF / AnLF can download the corresponding model from the link.

[0034] Step 4: NWDAF / AnLF sends a data collection request to the NF. This request includes the data source identifier (identifying the data producer NF (such as AMF, UPF, etc.) to ensure the accuracy of data collection and the reliability of the source), the requested data type (specifying the data type to be collected (such as traffic data, user behavior data) to meet the input requirements of the analysis), the timestamp (ensuring the timeliness and relevance of the data and helping data alignment during the analysis process), and the data collection strategy (defining the frequency and method of data collection to ensure the integrity and timeliness of the data).

[0035] Step 5: NF returns the corresponding data based on the data collection request.

[0036] Steps 6 and 7: NWDAF / AnLF collects data from another NF.

[0037] Steps 8 and 9: As needed, NWDAF / AnLF stores the collected data and collection timestamp together into ADRF for use in subsequent data collection. If the time requirement is met, the data can be directly obtained from ADRF without collecting data from NF again.

[0038] Step 10: NWDAF / AnLF performs analysis / inference based on the model and data, as follows: Input dataset: Data collected from NF, serving as the basic input for analysis / inference; ML model: A model obtained from MTLF, used for analysis / inference to ensure the accuracy and effectiveness of the analysis / inference; Analysis / inference algorithm: The specific algorithm applied to the data and model, determining the logic and output of the analysis / inference; Accuracy monitoring parameters: Used to evaluate the accuracy and reliability of the analysis / inference results, ensuring that the analysis / inference results meet expectations.

[0039] Step 11: NWDAF / AnLF returns the analysis / inference results, the process is as follows: Analysis / inference results, the analysis / inference output based on the model and data, provided to the consumer to support their decision-making; confidence level, the credibility index of the analysis / inference results, helps the consumer evaluate the reliability of the results.

[0040] In the aforementioned process of related technologies, the Consumer first needs to query the NRF that supports the Analytics ID based on the Analytics ID, and then send the Analytics ID to the NWDAF / AnLF. Only then can the NWDAF / AnLF determine which NFs to collect data from. In this embodiment of the invention, the NWDAF first uses an AI large model to translate the Consumer's NF request question, determine which NFs need to collect data from, and then collect data from the network NFs.

[0041] The methods and embodiments provided in this application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Taking running on a computer terminal as an example, Figure 3 This is a hardware structure block diagram of the computer terminal for the user data analysis and reasoning method according to an embodiment of the present invention, such as... Figure 3 As shown, a computer terminal may include one or more ( Figure 3 Only one is shown in the diagram. A processor 302 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 304 for storing data are also shown. The computer terminal may further include a transmission device 306 for communication functions and an input / output device 308. Those skilled in the art will understand that... Figure 3 The structure shown is for illustrative purposes only and does not limit the structure of the computer terminal described above. For example, the computer terminal may also include components that are more complex than those described above.Figure 3 The more or fewer components shown, or having the same Figure 3 The different configurations shown.

[0042] The memory 304 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the user data analysis and reasoning method in this embodiment of the invention. The processor 302 executes various functional applications and data processing by running the computer program stored in the memory 304, thereby implementing the above-described method. The memory 304 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 304 may further include memory remotely located relative to the processor 302, and these remote memories can be connected to the mobile terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0043] The transmission device 306 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider for the computer terminal. In one example, the transmission device 306 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 306 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0044] This invention provides a user data analysis and reasoning method. Figure 4 This is a flowchart of the user data analysis and reasoning method according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps:

[0045] In step S402, NWDAF converts the analysis and reasoning request from the consumer NF into a request identifier.

[0046] In this embodiment of the invention, the consumer NF is referred to as Consumer NF.

[0047] In one exemplary embodiment, NWDAF converts analytics inference requests from consumer NFs into request identifiers, including: NWDAF using an artificial intelligence (AI) model to convert analytics inference requests into request identifiers.

[0048] In one exemplary embodiment, the request identifier includes at least one of the following: a first identifier corresponding to the analytical reasoning model; and a second identifier corresponding to the knowledge data.

[0049] In this embodiment of the invention, the identifier corresponding to the analysis and reasoning model is designated as the first identifier, i.e., the model identifier; and the identifier corresponding to the knowledge data is designated as the second identifier, i.e., the data identifier.

[0050] In one exemplary embodiment, before the NWDAF converts the analytics reasoning request from the consumer NF into a request identifier, the method further includes: the NWDAF receiving the analytics reasoning request from the consumer NF, the analytics reasoning request including an analytics reasoning question and an analytics reasoning request context.

[0051] In this embodiment of the invention, the Consumer NF of the NWDAF service sends an analytics inference request to NWDAF / AnLF. This request carries a question related to the analytics inference request, expressed in a way that the Consumer NF can understand. The question also includes the context of the analytics inference request, including the time range, service area, and QoS requirements for analytics / inference. One way to express the analytics inference request is as a string, such as "Please provide predictions for all data traffic in Yuhuatai District, Nanjing City, Jiangsu Province on May 13, 2024". The Consumer NF can ask any question, and the question does not need to include a standardized Analytics ID.

[0052] In step S404, NWDAF obtains the analysis reasoning model and knowledge data based on the request identifier.

[0053] In one exemplary embodiment, NWDAF obtains an analytical reasoning model and knowledge data based on a request identifier, including: NWDAF obtaining the analytical reasoning model based on a first identifier and obtaining knowledge data based on a second identifier.

[0054] In an exemplary embodiment, the NWDAF obtains the analytical reasoning model based on the first identifier, including: the NWDAF's analytical logic function AnLF sending a model request to the NWDAF's model training logic function function MTLF, the model request carrying the first identifier; and the NWDAF's AnLF receiving model information from the MTLF, the model information including the analytical reasoning model or a model link corresponding to the analytical reasoning model.

[0055] In this embodiment of the invention, NWDAF / AnLF sends a model acquisition request to NWDAF / MTLF, which includes a model identifier, i.e., a first identifier. NWDAF / MTLF matches the corresponding analytical inference model based on the first identifier in the model request and returns the model information of the analytical inference model. This model information may include the actual model or a link to the model. After receiving it, NWDAF / AnLF can download the corresponding model from the link.

[0056] In one exemplary embodiment, the NWDAF acquires knowledge data based on a second identifier, including: the NWDAF's AnLF determining the network function NF that collects user data based on the second identifier and sending a first data collection request to the NF; the NWDAF's AnLF receiving user data from the NF; and the NWDAF's AnLF using an AI model to convert the user data into knowledge data.

[0057] In this embodiment of the invention, NWDAF can sequentially collect data from multiple NFs.

[0058] In this embodiment of the invention, NWDAF / AnLF collects data to the selected NF, that is, NWDAF / AnLF sends a first data collection request to the selected NF, and the NF returns the corresponding data, i.e., user data, to NWDAF / AnLF.

[0059] In one exemplary embodiment, the NWDAF acquires knowledge data based on a second identifier, including: the NWDAF's AnLF sending a second data acquisition request to the Analysis Data Storage Function (ADRF), the second data acquisition request carrying the second identifier; and the NWDAF's AnLF receiving knowledge data from the ADRF.

[0060] In this embodiment of the invention, NWDAF / AnLF sends a data request to ADRF, namely the second data acquisition request in the above embodiment. The data request carries a data identifier, and optionally, in one embodiment, it also carries model information related to the embedded model and the analysis and reasoning model.

[0061] Step S406: NWDAF performs analysis and reasoning based on the analysis and reasoning model and knowledge data.

[0062] In one exemplary embodiment, before the NWDAF performs analysis and reasoning based on the analysis and reasoning model and knowledge data, the method further includes: the NWDAF's AnLF sending a second identifier, along with the corresponding user data and AI model, to the ADRF, so that the ADRF uses the AI ​​model to convert the user data into knowledge data and stores it.

[0063] In one exemplary embodiment, the method further includes: if the ADRF fails to match knowledge data based on the second identifier, the AnLF of the NWDAF receives a third data acquisition request from the ADRF, the third data acquisition request being the address information of the ADRF; the AnLF of the NWDAF sends the third data acquisition request to the corresponding NF, so that the NF sends the user data to the corresponding ADRF.

[0064] In one exemplary embodiment, after the NWDAF performs analysis and reasoning based on the analysis and reasoning model and knowledge data, the method further includes: the NWDAF sending the analysis and reasoning results to the consumer NF.

[0065] This invention provides a user data analysis and reasoning method. The method involves using a Network Data Analyzer (NWDAF) to convert analysis and reasoning requests from a Consumer Network Function (NF) into request identifiers; the NWDAF then obtains an analysis and reasoning model and knowledge data based on the request identifiers; and finally, the NWDAF performs analysis and reasoning based on the analysis and reasoning model and knowledge data. This solves the problem of poor scalability in data collection using NWDAF in related technologies, thereby improving the scalability of NWDAF data collection.

[0066] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0067] This embodiment also provides a user data analysis and inference apparatus for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0068] The user data analysis and inference device provided in this embodiment of the invention can be installed in the core network or a functional unit of the core network, for example, in the Network Data Analysis Function (NWDAF). It includes a conversion module, an acquisition module, and an analysis and inference module. The conversion module is configured to convert analysis and inference requests from consumer NFs into request identifiers. The acquisition module is configured to acquire analysis and inference models and knowledge data based on the request identifiers. The analysis and inference module is configured to perform analysis and inference based on the analysis and inference models and knowledge data.

[0069] In this embodiment of the invention, the user data analysis and reasoning device may further include different modules, and the naming and functional division of the modules may be selected in different ways according to the actual situation, without specific limitations.

[0070] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0071] Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to perform the steps in any of the above method embodiments when executed.

[0072] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0073] Embodiments of the present invention also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0074] In one exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0075] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0076] In one exemplary embodiment, the computer program product described above includes a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the methods described in various embodiments of this application.

[0077] Specific examples in this embodiment can be found in the examples described in the above embodiments and exemplary implementations, and will not be repeated here.

[0078] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0079] To enable those skilled in the art to better understand the technical solutions of the present invention, the following description is provided in conjunction with different embodiments.

[0080] Example 1

[0081] In this embodiment, the process by which NWDAF / AnLF determines to directly collect data from NF is described.

[0082] Figure 5 This is the flow chart of the user data analysis and reasoning method according to an embodiment of the present invention. Figure 1 ,like Figure 5 As shown, it includes the following steps:

[0083] Step 1: The Consumer NF of the NWDAF service sends an analytics inference request to NWDAF / AnLF. This request carries an analytics inference question, expressed in a way that the Consumer NF can understand. The question also includes the context of the analytics inference request, including the time range, service area, and QoS requirements for analytics / inference. One way to express the analytics inference request is as a string, such as "Please provide a prediction of all data traffic in Yuhuatai District, Nanjing City, Jiangsu Province on May 13, 2024". The Consumer NF can ask any question; the question does not need to include the standardized Analytics ID.

[0084] Step 2: NWDAF obtains the embedded model based on the local embedded model (AI model) or from other storage model network elements (e.g., NWDAF / MTLF), and converts the consumer's input question, i.e. the analysis and reasoning request, into one or more identifiers, such as a model identifier and / or a data identifier.

[0085] In this embodiment of the invention, the identifier corresponding to the analysis and reasoning model is designated as the first identifier, i.e., the model identifier; and the identifier corresponding to the knowledge data is designated as the second identifier, i.e., the data identifier.

[0086] In this embodiment of the invention, the data identifier can be used as the Analytics ID, and NWDAF / AnLF can directly use the existing IE to transmit the Analytics ID.

[0087] Step 3: NWDAF / AnLF sends a model retrieval request to NWDAF / MTLF, which includes the model identifier, i.e., the first identifier. NWDAF / MTLF matches the corresponding analytical inference model based on the first identifier in the model request and returns the model information of the analytical inference model. This model information may include the actual model or a link to the model. After receiving it, NWDAF / AnLF can download the corresponding model from the link.

[0088] Step 4: Based on the output of the embedded model (AI model) in Step 2, NWDAF / AnLF determines which NFs need to collect data from, what data to collect, and information such as the time and strategy.

[0089] Step 5: NWDAF / AnLF collects data from the selected NF, i.e., NWDAF / AnLF sends the first data collection request to the selected NF, and the NF returns the corresponding data, i.e., user data, to NWDAF / AnLF.

[0090] Step 6: NWDAF / AnLF collects data from the next selected NF, and the NF returns the corresponding data, i.e., the user data, to NWDAF / AnLF.

[0091] Step 7: Optionally, NWDAF / AnLF converts all data collected from multiple NFs into knowledge data / vectors based on the embedding model in Step 2.

[0092] Step 8: NWDAF / AnLF sends the generated knowledge data / vectors and the AI ​​model generated in Step 2 to ADRF for storage. Optionally, NWDAF / AnLF can also send the collected raw user data and data identifier (i.e., the second identifier) ​​to ADRF together.

[0093] Step 9: If the NWDAF / AnLF sends the original user data, ADRF will use the embedding model, i.e. the AI ​​model, to convert the original user data into knowledge data / vectors and save them together with the data identifier.

[0094] Step 10: ADRF returns a save data response to NWDAF / AnLF.

[0095] Step 11: NWDAF / AnLF performs analysis and reasoning based on the analysis and reasoning model and user data / knowledge data.

[0096] Step 12: NWDAF / AnLF returns the analysis and inference results to the consumer or consumer NF.

[0097] Example 2

[0098] In this embodiment, the process by which NWDAF / AnLF determines and directly matches data collected from ADRF is described.

[0099] Figure 6 This is the flow chart of the user data analysis and reasoning method according to an embodiment of the present invention. Figure 2 ,like Figure 6 As shown, it includes the following steps:

[0100] Step 1: The Consumer NF of the NWDAF service sends an analytics inference request to NWDAF / AnLF. This request carries an analytics inference question, expressed in a way that the Consumer NF can understand. The question also includes the context of the analytics inference request, including the time range, service area, and QoS requirements for analytics / inference. One way to express the analytics inference request is as a string, such as "Please provide a prediction of all data traffic in Yuhuatai District, Nanjing City, Jiangsu Province on May 13, 2024". The Consumer NF can ask any question; the question does not need to include the standardized Analytics ID.

[0101] Step 2: NWDAF obtains the embedded model based on the local embedded model (AI model) or from other storage model network elements (e.g., NWDAF / MTLF), and converts the consumer's input question, i.e. the analysis and reasoning request, into one or more identifiers, such as a model identifier and / or a data identifier.

[0102] In this embodiment of the invention, the identifier corresponding to the analysis and reasoning model is designated as the first identifier, i.e., the model identifier; and the identifier corresponding to the knowledge data is designated as the second identifier, i.e., the data identifier.

[0103] In this embodiment of the invention, the data identifier can be used as the Analytics ID, and NWDAF / AnLF can directly use the existing IE to transmit the Analytics ID.

[0104] Step 3: NWDAF / AnLF sends a model retrieval request to NWDAF / MTLF, which includes the model identifier, i.e., the first identifier. NWDAF / MTLF matches the corresponding analytical inference model based on the first identifier in the model request and returns the model information of the analytical inference model. This model information may include the actual model or a link to the model. After receiving it, NWDAF / AnLF can download the corresponding model from the link.

[0105] Step 4: Based on the output of the embedded model (AI model) in Step 2, NWDAF / AnLF determines which NFs need to collect data from, what data to collect, and information such as the time and strategy.

[0106] Step 5: NWDAF / AnLF sends a data request to ADRF, namely the second data acquisition request in the above embodiment. The data request carries the data identifier generated in step 2. Optionally, in one embodiment, it also carries model information related to the embedded model and the analysis and reasoning model.

[0107] Step 6: ADRF performs knowledge matching within the database based on data identifiers and model information, extracting knowledge data that highly matches the data identifiers.

[0108] Step 7: ADRF returns the knowledge data matched by NWDAF / AnLF.

[0109] Step 8: NWDAF / AnLF performs analysis and reasoning based on the model and user data / knowledge data.

[0110] Step 9: NWDAF / AnLF returns the analysis and inference results to the consumer NF.

[0111] Example 3

[0112] In this embodiment, the process of NWDAF / AnLF determining data acquisition from ADRF and matching failure is described.

[0113] Figure 7 This is the flow chart of the user data analysis and reasoning method according to an embodiment of the present invention. Figure 3 ,like Figure 7 As shown, it includes the following steps:

[0114] Step 1: The Consumer NF of the NWDAF service sends an analytics inference request to NWDAF / AnLF. This request carries an analytics inference question, expressed in a way that the Consumer NF can understand. The question also includes the context of the analytics inference request, including the time range, service area, and QoS requirements for analytics / inference. One way to express the analytics inference request is as a string, such as "Please provide a prediction of all data traffic in Yuhuatai District, Nanjing City, Jiangsu Province on May 13, 2024". The Consumer NF can ask any question; the question does not need to include the standardized Analytics ID.

[0115] Step 2: NWDAF obtains the embedded model based on the local embedded model (AI model) or from other storage model network elements (e.g., NWDAF / MTLF), and converts the consumer's input question, i.e. the analysis and reasoning request, into one or more identifiers, such as a model identifier and / or a data identifier.

[0116] In this embodiment of the invention, the identifier corresponding to the analysis and reasoning model is designated as the first identifier, i.e., the model identifier; and the identifier corresponding to the knowledge data is designated as the second identifier, i.e., the data identifier.

[0117] In this embodiment of the invention, the data identifier can be used as the Analytics ID, and NWDAF / AnLF can directly use the existing IE to transmit the Analytics ID.

[0118] Step 3: NWDAF / AnLF sends a model retrieval request to NWDAF / MTLF, which includes the model identifier, i.e., the first identifier. NWDAF / MTLF matches the corresponding analytical inference model based on the first identifier in the model request and returns the model information of the analytical inference model. This model information may include the actual model or a link to the model. After receiving it, NWDAF / AnLF can download the corresponding model from the link.

[0119] Step 5: NWDAF / AnLF sends a data request to ADRF, which carries the data identifier generated in step 2. Optionally, in one embodiment, it also carries model information related to the embedded model and the analysis / inference model.

[0120] Step 6: ADRF performs knowledge matching within the database based on data identifiers and model information, and finds that no matching knowledge data can be found.

[0121] Step 7: ADRF returns a data acquisition request, namely the third data acquisition request in the above embodiment, which carries the address information of ADRF.

[0122] Step 8: Based on the output of the embedded model in Step 2, NWDAF / AnLF determines which NFs need to collect data from, what data to collect, and information such as time and strategy.

[0123] Step 9: The NWDAF / AnLF collects data from the selected NF, carrying the data identifier from Step 2 and the corresponding ADRF address information in the request. In one implementation, this data identifier can be sent as an association identifier within the data collection request message. The NF then sends this association identifier along with the collected data to the ADRF. In this way, the ADRF uses the association identifier to associate the data returned by multiple NFs as a single data collection task. After collecting data, the NF sends the corresponding data and the received data identifier to the ADRF. The NWDAF / AnLF then collects data from the next selected NF, carrying the data identifier from Step 2 and the corresponding ADRF address information in the request. After the NF collects data, the ADRF sends the corresponding data and the received data identifier.

[0124] Step 10: ADRF uses an embedding model to convert all received raw user data related to the data identifier into knowledge data / vectors, and saves them together with the data identifier.

[0125] Step 11: ADRF returns a data response to NWDAF / AnLF, which carries the newly generated knowledge data / vectors.

[0126] Step 12: NWDAF / AnLF performs analysis and reasoning based on the model and the received knowledge data.

[0127] Step 13: NWDAF / AnLF returns the analysis and inference results to the consumer.

[0128] In summary, this invention provides a user data analysis and reasoning method. The NWDAF / AnLF receives a question request from the Consumer / Consumer NF, converts the requested question into a model identifier and a data identifier using an AI big model, then uses the model identifier to request model information from the NWDAF / MTLF, and uses the requested data identifier to request knowledge data from the ADRF, or directly collects data from the NF. The NWDAF / AnLF performs analysis / reasoning based on the model information and the collected data, and then provides the analysis / reasoning results to the Consumer / Consumer NF.

[0129] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A user data analysis inference method, characterized by, The method comprises: The network data analytics function (NWDAF) converts an analytics inference request from a consumer network function (NF) into a request identifier; The NWDAF acquires an analytics inference model and knowledge data according to the request identifier; The NWDAF performs analytics inference according to the analytics inference model and the knowledge data.

2. The method of claim 1, wherein, The network data analytics function (NWDAF) converts an analytics inference request from a consumer network function (NF) into a request identifier, comprising: The NWDAF converts the analytics inference request into the request identifier by using an artificial intelligence (AI) model.

3. The method of claim 1, wherein, The request identifier comprises at least one of: A first identifier corresponding to the analytics inference model; and a second identifier corresponding to the knowledge data. The NWDAF acquires an analytics inference model and knowledge data according to the request identifier, comprising:

4. The method of claim 3, wherein, The NWDAF acquires the analytics inference model according to the first identifier and acquires the knowledge data according to the second identifier. The NWDAF acquires the analytics inference model according to the first identifier, comprising:

5. The method of claim 4, wherein, The analytics logic function (AnLF) of the NWDAF sends a model request to the model training logic function (MTLF) of the NWDAF, wherein the model request carries the first identifier; The AnLF of the NWDAF receives model information from the MTLF, wherein the model information comprises the analytics inference model or a model link corresponding to the analytics inference model. The NWDAF acquires the knowledge data according to the second identifier, comprising:

6. The method of claim 4, wherein, The AnLF of the NWDAF determines a network function (NF) that collects user data according to the second identifier and sends a first data collection request to the NF; The AnLF of the NWDAF receives the user data from the NF; The AnLF of the NWDAF converts the user data into the knowledge data by using an AI model. The NWDAF acquires the knowledge data according to the second identifier, comprising:

7. The method of claim 4, wherein, The AnLF of the NWDAF sends a second data collection request to an analytics data storage function (ADRF), wherein the second data collection request carries the second identifier; The AnLF of the NWDAF receives the knowledge data from the ADRF. Before the NWDAF performs analytics inference according to the analytics inference model and the knowledge data, the method further comprises:

8. The method of claim 7, wherein, The AnLF of the NWDAF sends the knowledge data and an AI model to the ADRF, so that the ADRF stores the knowledge data and the AI model. Before the NWDAF performs analytics inference according to the analytics inference model and the knowledge data, the method further comprises:

9. The method of claim 7, wherein, The AnLF of the NWDAF sends the second identifier, corresponding user data and an AI model to the ADRF, so that the ADRF converts the user data into the knowledge data by using the AI model and stores the knowledge data. The method further comprises:

10. The method of claim 7, wherein, ​ In the case that the ADRF fails to match the knowledge data according to the second identification, the AnLF of the NWDAF receives a third data collection request from the ADRF, the third data collection request including address information of the ADRF; The AnLF of the NWDAF sends the third data collection request to a corresponding NF, so that the NF sends user data to the corresponding ADRF.

11. The method of claim 1, wherein, Before the NWDAF converts an analysis inference request from a consumer NF into a request identification, the method further comprises: The NWDAF receives the analysis inference request from the consumer NF, the analysis inference request including an analysis inference question and an analysis inference request context.

12. The method of claim 1, wherein, After the NWDAF performs analysis inference according to the analysis inference model and the knowledge data, the method further comprises: The NWDAF sends an analysis inference result to the consumer NF.

13. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the method in any one of claims 1-12 when executing the computer program.

14. A computer program product comprising a computer program, characterized in that, The computer program, when executed by the processor, implements the method in any one of claims 1-12.

Images