Information processing method and device based on model transmission state analysis

Abstract

The application provides an information processing method and device based on model transmission state analysis. The method comprises the following steps: sending a first message to a network data analysis function (NWDAF), wherein the first message is used for requesting to subscribe to analysis information of an artificial intelligence / machine learning (AI / ML) model transmission state in a network; receiving the analysis information sent by the NWDAF; and performing information processing on the AI / ML model transmission according to the analysis information. The information processing based on the AI / ML model transmission can be effectively realized, and then the charging negotiation, charging statistics, policy decision or session management and the like of the AI / ML model transmission can be realized.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to an information processing method and apparatus based on model transmission state analysis. Background Technology

[0002] In recent years, due to technological breakthroughs in artificial intelligence, its applications have become increasingly widespread. In mobile communication systems, mobile devices are increasingly replacing traditional algorithms with artificial intelligence (AI) / machine learning (ML) models (hereinafter referred to as AI / ML models) to improve their intelligence level.

[0003] End devices typically have strict limitations in power consumption, computing power, and memory, making it impossible to run a large number of AI / ML models. Therefore, AI / ML models need to be transmitted to the cloud or other terminals.

[0004] However, transmitting AI / ML models in 5G systems needs to meet the requirements outlined in SA1 R18 to enable information processing such as billing negotiation, billing statistics, policy decisions, and session management related to AI / ML model transmission. However, current technologies cannot effectively perform information processing based on AI / ML model transmission, thus hindering operations such as billing negotiation, billing statistics, policy decisions, and session management. Summary of the Invention

[0005] This application provides an information processing method and apparatus based on model transmission status analysis, which solves the technical problem in the prior art that it is impossible to effectively realize information processing based on AI / ML model transmission, and thus impossible to realize operations such as fee negotiation, fee statistics, strategy decision-making or session management of AI / ML model transmission, thereby ensuring the service experience and performance of AI / ML model transmission.

[0006] In a first aspect, this application provides an information processing method based on model-transmission state analysis, the method being applied to a policy control function (PCF), the method comprising:

[0007] Send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0008] Receive analysis information sent by NWDAF;

[0009] Based on the analysis information, information processing is performed on the AI / ML model transmission.

[0010] In this embodiment, a request for analysis information on the transmission status of AI / ML models in the subscription network is sent to NWDAF. The analysis information determined by NWDAF based on the collected 5GC NF(s) data is received. Then, information processing is performed on the AI / ML model transmission based on the analysis information, thereby realizing operations such as fee negotiation, fee statistics, and strategy decision-making in the AI / ML model transmission.

[0011] Optionally, the information processing includes fee / sponsorship negotiation; the information processing of AI / ML model transmission based on the analysis information includes:

[0012] The system receives a second message, either directly or through the Network Capability Opening Function (NEF), from an Application Service Provider (ASP) or Application Function (AF). This second message requests negotiation on pricing information for AI / ML model transmission. The parameters requested in the second message include at least one of the following: Federated Learning Group Identifier, Identifier of the application using the AI / ML model, Model Size for AI / ML Transmission, Region Information for AI / ML Model Use, Time Period for AI / ML Model Transmission, Quality of Service (QoS) Parameter Set, Transaction Reference Identifier, and Sponsorship Status.

[0013] Based on the analysis information and the second message, determine whether to accept the ASP / AF's fee / sponsorship request, and send a first result directly or through NEF to the ASP / AF; the first result includes accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission or not accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission.

[0014] Optionally, the information processing further includes PDU session charge statistics; the second message is also used to request negotiation of PDU session charge statistics information for AI / ML model transmission;

[0015] The PDU session used for AI / ML model transmission is created by the user equipment (UE). The parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set.

[0016] Accordingly, the parameters requested in the first message include the parameters in the PDU session.

[0017] In this embodiment, a request for negotiating the charging information of AI / ML model transmission sent by ASP / AF is received directly or through NEF. Then, based on the request and the parameters in the corresponding analysis information, it is determined whether to accept ASP / AF as the charging party / sponsor for AI / ML model transmission or not to accept ASP / AF as the charging party / sponsor for AI / ML model transmission, thus realizing charging / sponsor negotiation based on model transmission status analysis.

[0018] Optionally, the step of processing the AI / ML model transmission based on the analysis information further includes:

[0019] Based on the analysis information and the parameters requested in the second message, determine the billing rules for AI / ML model transmission in the PCC rules;

[0020] Send a second result to the Session Management Function (SMF), the second result including the determined PCC rules;

[0021] The system receives information from the User Plane Function (UPF) report and session / user subscription information sent by the SMF, and charges the UE's AI / ML model transmission PDU session based on the information from the UPF report, session / user subscription information, and other charging function entities.

[0022] The information reported by the UPF is determined by the UPF based on the usage reporting rules for data packets used to transmit AI / ML models. The usage reporting rules are determined by the SMF based on the received and determined PCC rules.

[0023] Optionally, the billing rules include at least one of the following:

[0024] No payment required;

[0025] Different rates apply depending on the size of the AI / ML model being transmitted; the larger the AI / ML model being transmitted, the higher the rate for transmitting the AI / ML model.

[0026] Different rates are applied based on the length of the AI / ML model transmission period; the longer the AI / ML model transmission time, the higher the rate applied to AI / ML model transmission.

[0027] Different rates are applied based on a weighted average of the size of the AI / ML model being transmitted, the region where the AI / ML model is used, and the time period for which the AI / ML model is transmitted; where a higher weighted average of these factors results in a higher rate for AI / ML model transmission.

[0028] Different rates are applied based on the set of service quality parameters; among them, the higher the service quality stream bit rate for AI / ML model transmission, the higher the rate applied to AI / ML model transmission.

[0029] Whether or not a fee is charged and the corresponding rate is determined based on whether or not the Federal Learning Group logo is present.

[0030] In this embodiment, the charging rules for AI / ML model transmission in the PCC rules are determined based on the parameters of the request for negotiation of charging information for AI / ML model transmission sent by the ASP / AF and the parameters in the corresponding analysis information. These parameters include the charging mode and rate. Then, the determined (or updated) PCC rules are sent to the SMF. Based on the information in the UPF report sent by the SMF, the session / user subscription information, and other charging function entities, the charging is performed on the UE's AI / ML model transmission PDU session, thus realizing PDU session charging statistics based on model transmission status analysis.

[0031] Optionally, the parameters requested in the first message include at least one of the following: the identifier of the application using the AI / ML model, the network slice of the PDU session used to transmit the AI / ML model service quality flow, the data network name of the PDU session used to transmit the AI / ML model service quality flow, and the service quality parameter set; the analysis information includes at least one of the analysis results corresponding to the parameters requested in the first message; wherein, the information processing includes session management policy decision processing;

[0032] Accordingly, the information processing of the AI / ML model transmission based on the analysis information includes:

[0033] Based on the analysis information, determine the authorized service quality parameters for AI / ML model transmission in the PCC rules;

[0034] The authorized service quality parameters transmitted by the AI / ML model in the determined PCC rules are used as the latest session management policy information, and a third message carrying the latest session management policy information is sent to the session management function SMF. The third message is used to request the SMF to update the session management policy.

[0035] The system receives a second result sent by the SMF, which is determined by the SMF based on the latest session management policy information. The second result may include updating the session management policy or not updating the session management policy.

[0036] Optionally, determining the authorized service quality parameters for AI / ML model transmission in the PCC rules based on the analysis information includes:

[0037] If the data rate of the AI / ML model transmission in the analysis information is detected to be too low, then the priority of the 5G service quality identifier in the authorized service quality parameters of the AI / ML model transmission, or the reflective service quality control, the maximum bit rate of the uplink direction of the transmitted AI / ML model, the maximum bit rate of the transmitted AI / ML model in the downlink direction, the minimum bit rate of the transmitted AI / ML model in the uplink direction, and the minimum bit rate of the transmitted AI / ML model in the downlink direction in the authorized service quality parameters are adjusted.

[0038] In this embodiment of the application, if the data rate of the AI / ML model transmission in the analysis information is detected to be too low, the session management strategy decision processing based on model transmission state analysis is realized by adjusting the priority of the 5G service quality identifier in the authorized service quality parameters of the AI / ML model transmission, or the reflective service quality control, the maximum bit rate of the uplink direction of the transmitted AI / ML model, the maximum bit rate of the transmitted AI / ML model in the downlink direction, the minimum bit rate of the transmitted AI / ML model in the uplink direction, and the minimum bit rate of the transmitted AI / ML model in the downlink direction of the transmitted AI / ML model in the authorized service quality parameters.

[0039] Secondly, this application provides an information processing method based on model transport state analysis, applied to the Session Management Function (SMF), the method comprising:

[0040] Send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0041] Receive analysis information sent by NWDAF;

[0042] Based on the analysis information, information processing is performed on the AI / ML model transmission.

[0043] In this embodiment, a request to NWDAF for analysis information on the transmission status of AI / ML models in the network is sent to NWDAF. The analysis information determined by NWDAF based on the collected 5GC NF(s) data is received. Then, information processing is performed on the AI / ML model transmission based on the analysis information, thereby realizing operations such as session management in the transmission of AI / ML models.

[0044] Optionally, the parameters requested in the first message include parameters from a PDU session for AI / ML model transmission, wherein the PDU session for AI / ML model transmission is created by a user equipment (UE); the parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set; the information processing includes session management;

[0045] The step of processing the AI / ML model transmission based on the analysis information includes:

[0046] Based on the location of each UE, information on the optimal AI / ML model transmission service experience is matched from the analysis information;

[0047] The information of the service session anchor UPF is obtained from the information of the service experience transmitted by the optimal AI / ML model, and the service session anchor UPF is determined to be the new PDU session anchor UPF.

[0048] The selected new PDU session anchor point UPF will be used to provide the optimal path for AI / ML model transmission for each UE.

[0049] Optionally, determining to use the service session anchor UPF as the new PDU session anchor UPF includes:

[0050] In AI / ML model transmission, the optimal path for AI / ML model transmission is determined based on the location of each UE and the information on the best AI / ML model transmission service experience.

[0051] The service session anchor point UPF corresponding to the optimal path is determined as the new PDU session anchor point UPF.

[0052] Optionally, the method further includes:

[0053] Before changing the PDU session anchor point UPF, a fourth message is sent to the application function AF. The fourth message is used to notify the user plane anchor point of the change and the target data network access identifier.

[0054] Receive the confirmation result sent by AF.

[0055] In this embodiment, a request for analysis information on the transmission status of AI / ML models in the subscription network is sent to the NWDAF. The NWDAF sends analysis information determined based on the collected 5GC NF(s) data. The best UPF is then selected based on the analysis information, thereby realizing session management of AI / ML model transmission.

[0056] Optionally, the step of processing the AI / ML model transmission based on the analysis information further includes:

[0057] Based on the location of each UE, information on the optimal AI / ML model transmission service experience is matched from the analysis information;

[0058] The information of the service session anchor point UPF and the analysis results of the service quality parameter set are obtained from the information of the service experience transmitted by the optimal AI / ML model.

[0059] Based on the information of the service session anchor UPF and the analysis results of the service quality parameter set, a new PDU session anchor UPF is determined to be established to transmit the AI / ML model. The new PDU session anchor UPF is the second PDU session anchor UPF, wherein the first PDU session anchor UPF is the PDU session anchor currently transmitting the AI / ML model.

[0060] Select either a User Plane Function (UPF) as the branch point (BP) or the uplink classifier (UL CL) of the PDU session;

[0061] Provide the BP or UL CL with a traffic filter corresponding to the first PDU session anchor UPF and the second PDU session anchor UPF, and instruct the BP or UL CL to forward the quality of service stream / data packet transmitted by the uplink AI / ML model to the second PDU session anchor UPF;

[0062] The created second PDU session anchor point UPF will be used to provide each UE with the optimal path for information related to the quality of service flow in the AI / ML model transmission.

[0063] Optionally, determining the establishment of a new PDU session anchor for transmitting AI / ML models based on the information of the service session anchor UPF and the analysis results of the service quality parameter set includes:

[0064] Based on the information from the service session anchor point UPF and the analysis results of the service quality parameter set, the optimal path for AI / ML model transmission is determined.

[0065] Based on the optimal path, a new PDU session anchor point UPF is determined to be established to transmit the AI / ML model.

[0066] Optionally, the method further includes:

[0067] Before creating a new PDU session anchor point UPF to transmit the AI / ML model, a fifth message is sent to the application function AF, which is used to notify the user of the change in the plane anchor point and the target data network access identifier.

[0068] Receive the confirmation result sent by AF.

[0069] In this embodiment, a request to NWDAF for analysis information on the transmission status of AI / ML models in the subscription network is sent, and analysis information determined by NWDAF based on the collected 5GC NF(s) data is received. Then, BP or UL CL is set according to the analysis information, thereby realizing session management of AI / ML model transmission.

[0070] Thirdly, this application provides an information processing device based on model-driven transmission state analysis, applied to a policy control function (PCF). The device includes: a memory, a transceiver, and a processor.

[0071] A memory for storing computer programs; a transceiver for sending and receiving data under the control of the processor; and a processor for reading the computer programs from the memory and performing the following operations:

[0072] Send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0073] Receive analysis information sent by NWDAF;

[0074] Based on the analysis information, information processing is performed on the AI / ML model transmission.

[0075] Fourthly, this application provides an information processing device based on model-driven transmission state analysis, applied to the Session Management Function (SMF). The device includes: a memory, a transceiver, and a processor.

[0076] A memory for storing computer programs; a transceiver for sending and receiving data under the control of the processor; and a processor for reading the computer programs from the memory and performing the following operations:

[0077] Send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0078] Receive analysis information sent by NWDAF;

[0079] Based on the analysis information, information processing is performed on the AI / ML model transmission.

[0080] Fifthly, this application provides an information processing device based on model-transmission state analysis, applied to the policy control function (PCF), the device comprising:

[0081] The sending unit is used to send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0082] The receiving unit is used to receive the analysis information sent by NWDAF;

[0083] The processing unit is used to process the AI / ML model transmission based on the analysis information.

[0084] Sixthly, this application provides an information processing apparatus based on model transport state analysis, applied to the session management function (SMF), the apparatus comprising:

[0085] The sending unit is used to send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0086] The receiving unit is used to receive the analysis information sent by NWDAF;

[0087] The processing unit is used to process the AI / ML model transmission based on the analysis information.

[0088] In a seventh aspect, this application provides a processor-readable storage medium storing a computer program for causing the processor to perform the method described in either the first or second aspect.

[0089] This application provides an information processing method and apparatus based on model transmission status analysis. The method involves sending a first message to a Network Data Analysis Function (NWDAF) requesting subscription to analysis information on the transmission status of AI / ML models in the network; receiving analysis information from the NWDAF; and processing the AI / ML model transmission based on the analysis information. By sending a request to the NWDAF to subscribe to analysis information on the transmission status of AI / ML models in the network, receiving analysis information determined by the NWDAF based on collected 5GC NF(s) data, and then processing the AI / ML model transmission based on the analysis information, the method enables operations such as fee negotiation, fee statistics, strategy decision-making, or session management in AI / ML model transmission, ensuring the service experience and performance of AI / ML model transmission.

[0090] It should be understood that the description in the foregoing summary section is not intended to limit the key or essential features of the embodiments of the present invention, nor is it intended to restrict the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0091] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0092] Figure 1 Network architecture diagram of the information processing method based on model transmission state analysis provided in the embodiments of this application;

[0093] Figure 2 A network architecture diagram of an information processing method based on model transmission state analysis provided in another embodiment of this application;

[0094] Figure 3 A flowchart illustrating an information processing method based on model transmission state analysis provided in an embodiment of this application;

[0095] Figure 4 A schematic diagram of the signaling flow of the information processing method based on model transmission state analysis provided in the embodiments of this application;

[0096] Figure 5 A schematic diagram of the signaling flow of an information processing method based on model transmission state analysis provided in another embodiment of this application;

[0097] Figure 6 A schematic diagram of the signaling flow of an information processing method based on model transmission state analysis provided in another embodiment of this application;

[0098] Figure 7 A flowchart illustrating an information processing method based on model transmission state analysis is provided in another embodiment of this application;

[0099] Figure 8 A schematic diagram of the signaling flow of an information processing method based on model transmission state analysis provided in another embodiment of this application;

[0100] Figure 9 A schematic diagram of the signaling flow of an information processing method based on model transmission state analysis provided in another embodiment of this application;

[0101] Figure 10 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application;

[0102] Figure 11 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application;

[0103] Figure 12 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application;

[0104] Figure 13 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application. Detailed Implementation

[0105] In this application, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0106] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0107] To clearly understand the technical solution of this application, the solutions of the prior art will first be described in detail. In the prior art, the SA1 R18 requirements approved by SA#93e require AI / ML model transmission in at least the following scenarios:

[0108] Scenario 1: Distribution and sharing of AI / ML models. Due to changes in tasks or environments, mobile terminals have limited memory and cannot pre-load all models. Therefore, mobile terminals need to download new AI / ML models from the network in real time via the 5G system.

[0109] Scenario 2: Federated learning algorithm using 5GS. When training a global model on a cloud server, it is necessary to aggregate the models trained locally on various terminal devices. Each training iteration process: a terminal device downloads a global model from the cloud server and trains it using local data; the terminal reports the intermediate training results to the cloud server; the cloud server aggregates the intermediate training results from all terminals and updates the global model, then redistributes the global model to the terminals; the terminals then execute the next iteration.

[0110] Scenario 3: AI / ML Model Segmentation Between AI / ML Endpoints. An AI / ML model can be segmented into multiple parts based on the current task or environment. The trend is to perform computationally complex and energy-intensive inference on the network, while inference on the terminal for parts requiring privacy protection or sensitive to latency. For example, the terminal downloads / loads a model, infers specific layers / parts first, and then sends the intermediate results to the network; the network then executes the remaining layers / parts and feeds back the inference results to the terminal. This scenario involves transmitting part of the model in the first step or intermediate step, and therefore may include model transmission.

[0111] Therefore, transmitting AI / ML models in a 5G system needs to meet the requirements outlined in SA1 R18 to enable information processing such as billing negotiation, billing statistics, policy decisions, and session management related to AI / ML model transmission. However, current technologies cannot effectively perform information processing based on AI / ML model transmission, thus hindering operations such as billing negotiation, billing statistics, policy decisions, and session management.

[0112] Further research by the inventors revealed that effective analysis of the transmission status of AI / ML models requires interaction between Application Functions (AF), Network Exposure Functions (NEF), Network Data Analytics Functions (NWDAF), and various Network Functions (NF). For example... Figure 1 As shown, the AF can send a request to the NWDAF directly or through the NEF to indicate the analysis of the AI / ML model transmission status in the subscribed network. The NWDAF analyzes the AI / ML model transmission status in the network and feeds back data by collecting data from various Network Functions (NFs) in the 5G Core Network (5GC). This enables effective analysis of the AI / ML model transmission status, allowing the Policy Control Function (PCF) or Session Management Function (SMF) to perform information processing such as charging negotiation, charging statistics, policy decision-making, or session management based on the analysis information of the AI / ML model transmission status.

[0113] Therefore, based on the inventors' inventive research, this application proposes an information processing method based on model transmission status analysis. In this application, a first message is sent to the Network Data Analysis Function (NWDAF), requesting subscription to analysis information on the transmission status of AI / ML models in the network; analysis information is received from the NWDAF; and information processing is performed on the AI / ML model transmission based on the analysis information. By sending a request to the NWDAF to subscribe to analysis information on the transmission status of AI / ML models in the network, receiving analysis information determined by the NWDAF based on collected 5GC NF(s) data, and then performing information processing on the AI / ML model transmission based on the analysis information, operations such as fee negotiation, fee statistics, strategy decision-making, or session management in AI / ML model transmission are realized, ensuring the service experience and performance of AI / ML model transmission.

[0114] Figure 2 A network architecture diagram of an information processing method based on model transmission state analysis provided in another embodiment of this application is shown below. Figure 2 As shown in this embodiment, NWDAF is a network analysis function managed by the operator. NWDAF can provide data analysis services to 5GC network functions, application functions (AF), and operation administration and maintenance (OAM). The analysis results can be historical statistics or predictive information. NWDAF can serve one or more network slices.

[0115] The 5GC also includes several other functions, namely: User Plane Function (UPF), Session Management Function (SMF), Access and Mobility Management Function (AMF), Unified Data Repository (UDR), Network Exposure Function (NEF), AF, Policy Control Function (PCF), and Online Charging System (OCS). These other functions are collectively referred to as NFs. The NWDAF communicates with other functional entities in the 5G core network, including 5GC NF(s) and OAM, through service-oriented interfaces.

[0116] In 5GC, different NWDAF instances can provide different types of specialized analytics. For consumer NFs to discover suitable NWDAF instances to provide specific types of analytics, NWDAF instances must provide their supported Analytic ID when registering with the Network Repository Function (NRF). The Analytic ID represents the analytics type (or analytics identifier). This allows consumer NFs to provide the Analytic ID when querying the NRF for NWDAF instances, indicating the type of analytics needed. 5GC network functions and OAM determine how to use the data analytics provided by the Network Data Analytics Function (NWDAF) to improve network performance.

[0117] In one application scenario, the PCF and SMF send an Nnwdaf_AnalyticsSubscription_Subscribe request to the NWDAF to subscribe to analysis information on the transmission status of AI / ML models in the network. This analysis information includes at least one of the following: the application ID using the AI / ML model, the network data analysis ID, the region information using the AI / ML model, the time period for transmitting the AI / ML model, the size of the transmitted model, the Quality of Service (QoS) information related to the transmission of the AI / ML model, the network slice used for AI / ML model transmission, and the Data Network Name (DNN). If federated learning is involved, it also includes: a group ID (i.e., federated learning group ID), the UE ID or UE group ID participating in the federated learning, and the address information of the application server providing the model or participating in the federated learning. The PCF and SMF make corresponding decisions and optimize session management based on the analysis information on the transmission status of AI / ML models provided by the NWDAF.

[0118] Therefore, PCF and SMF send requests to NWDAF to request analysis information on the transmission status of AI / ML models in the subscription network. PCF and SMF receive analysis information from NWDAF based on the collected 5GC NF(s) data. Then, PCF and SMF process information on AI / ML model transmission based on the analysis information, thereby enabling PCF to perform operations such as fee negotiation, fee statistics, and policy decision-making in AI / ML model transmission, and SMF to perform operations such as session management in AI / ML model transmission.

[0119] The embodiments of this application will now be described with reference to the accompanying drawings.

[0120] Figure 3 A flowchart illustrating an information processing method based on model transmission state analysis provided in an embodiment of this application is shown below. Figure 3 As shown, the execution entity of the information processing method based on model transmission state analysis provided in this embodiment is PCF. Therefore, the information processing method based on model transmission state analysis provided in this application embodiment includes the following steps:

[0121] Step 101: Send the first message to the Network Data Analysis Function (NWDAF).

[0122] The first message is used to request the subscription of analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network.

[0123] In this embodiment, the parameters requested in the first message include at least one of the following: network data analysis identifier (i.e., Analytics ID), identifier of a user equipment (UE) or a group of UEs receiving the AI / ML model, or any UE that meets the analysis conditions (i.e., Target of Analytics Reporting), identifier of the application using the AI / ML model (i.e., Application ID), area of ​​AI / ML model transmission (i.e., Area of ​​Interest (AoI)), network slice (i.e., S-NSSAI) indicating the protocol data unit (PDU) session transmitting the AI / ML model quality of service flow, data network (i.e., DNN) indicating the PDU session transmitting the AI / ML model quality of service flow, time period of AI / ML model transmission (i.e., Model transmission duration), start timestamp of AI / ML model transmission (i.e., Model transmission start), end timestamp of AI / ML model transmission (i.e., Model transmission stop), size of AI / ML transmission model (i.e., Model size), and quality of service requirements (i.e., QoS requirements, including quality of service requirements for indicating the quality of service flow of AI / ML model transmission (i.e., 5QI (5G QoS Identifier)) and / or specific quality of service requirements for indicating the transmission of AI / ML model (i.e., QoS Characteristics)).

[0124] Among these are specific quality of service requirements, such as packet transmission latency and packet error rate.

[0125] If the AI / ML model performs federated learning, the parameters requested in the first message may also include at least one of the following: federated learning (FL) group information; the federated learning (FL) group information may include: the identifier of the federated learning group used to indicate the analysis (i.e., Federated Learning (FL) group ID), the identifier of the UE or UE(s) participating in the federated learning (i.e., Federated Learning (FL) UE ID or UE group ID), and the identifier of the application participating in the federated learning (i.e., Federated Learning (FL) Application ID).

[0126] Step 102: Receive the analysis information sent by NWDAF.

[0127] The analysis information is determined by the NWDAF based on the AI / ML model transmission status data sent by other network functions of the 5G core network, 5GC NF(s). The AI / ML model transmission status data is collected by the NWDAF by sending a data collection request to the 5GC NF(s) based on the parameters requested in the first message received.

[0128] In this embodiment, the parameters requested in the data acquisition request include at least one of the following: the current location of the UE using the AI / ML model (i.e., UE location), the identifier of the application using the AI / ML model (i.e., Application ID, which can be the identifier of the server or the identifier of the AF), the Quality of Service Flow Identifier (i.e., QFI) for transmitting the AI / ML model, the bit rate for the uplink direction (i.e., UL direction) and the bit rate for the downlink direction (i.e., DL direction) of the AI / ML model, the packet delay for the uplink direction (i.e., UL direction) and the packet delay for the downlink direction (i.e., DL direction) of the AI / ML model, the number of abnormal releases of the Quality of Service Flow during the AI / ML model transmission period (QoS Sustainability), the number of packet transmissions of the AI / ML model, and the number of packet retransmissions of the AI / ML model (i.e., packet retransmissions). The data includes: retransmission, data acquisition time (i.e., timestamp), AI / ML model transmission duration (i.e., AI / ML model transmission time period), AI / ML model transmission start timestamp, AI / ML model transmission end timestamp, AI / ML transmission model size, network slice of PDU session used for transmitting AI / ML model QoS flow, data network of PDU session used for transmitting AI / ML model QoS flow, and service process used for the AF (i.e., IP filter information).

[0129] If the AI / ML model performs federated learning, the parameters requested in the data acquisition request shall also include at least one of the following: the identifier of the federated learning group to be analyzed (i.e., Federated Learning(FL) group ID), the identifier of the UE or UE(s) participating in the federated learning (i.e., Federated Learning(FL) UE ID or UE group ID), and the identifier of the application participating in the federated learning (i.e., Federated Learning(FL) Application ID).

[0130] Step 103: Based on the analysis information, perform information processing on the AI / ML model transmission.

[0131] The analytical information is used for negotiating fees / sponsorships, PDU session fee statistics, and making session management strategy decisions, i.e., the analytical information is used to support fee decisions and SM strategies.

[0132] The analysis information includes at least one of the following: network slices of PDU sessions used to transmit AI / ML model quality of service flows, identifiers of applications using AI / ML models, region information using AI / ML models, validity period of analysis results, and user plane function UPF (i.e., UPF) providing AI / ML model transmission. The information includes: the data network name of the PDU session used to transmit the AI / ML model service quality flow; the size of the AI / ML transmission model; the duration of the AI / ML model transmission; the start timestamp of the AI / ML model transmission; the end timestamp of the AI / ML model transmission; and the service quality requirements. The service quality requirements may include the service quality flow identifier (QFI) for transmitting the AI / ML model; the uplink and downlink bit rates for transmitting the AI / ML model; the uplink and downlink packet delays for the AI / ML model; the number of abnormal releases of the service quality flow during the AI / ML model transmission period; the number of times the reporting threshold for abnormal releases of the service quality flow during the AI / ML model transmission period was reached; the number of AI / ML model packet transmissions; and the number of AI / ML model packet retransmissions.

[0133] If the AI / ML model performs federated learning, the analysis information also includes at least one of the following: an identifier for indicating the federated learning group to be analyzed, an identifier of the UE participating in the federated learning or an identifier of the UE(s) to indicate the individual application providing the AI / ML model or participating in the federated learning (i.e., the Application Server Instance Address).

[0134] In this embodiment, a request for analysis information on the transmission status of AI / ML models in the subscription network is sent to NWDAF. The analysis information determined by NWDAF based on the collected 5GC NF(s) data is received. Then, information processing is performed on the AI / ML model transmission based on the analysis information, thereby realizing operations such as fee negotiation, fee statistics, and strategy decision-making in the AI / ML model transmission.

[0135] Optionally, the information processing includes fee / sponsorship negotiation; the information processing of the AI / ML model transmission based on the analysis information can be achieved through the following steps:

[0136] Step a1: Receive a second message sent directly or through the Network Capability Opening Function (NEF) from the Application Service Provider (ASP) or Application Function (AF). The second message is used to request negotiation on the charging information for AI / ML model transmission. The parameters requested in the second message include at least one of the following: Federated Learning Group Identifier, Application Identifier participating in Federated Learning, Model Size for AI / ML Transmission, Regional Information for Using the AI / ML Model, Time Period for AI / ML Model Transmission, Quality of Service Parameter Set, Transaction Reference Identifier, and Sponsorship Status.

[0137] Step a2: Based on the analysis information and the second message, determine whether to accept the ASP / AF's fee / sponsorship request, and send a first result directly or through NEF to the ASP / AF; the first result includes accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission or not accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission.

[0138] In this embodiment, if the PCF is in the trusted area, it directly receives the request from the ASP / AF to negotiate the charging information for AI / ML model transmission. If the PCF is not in the trusted area, it receives the request from the ASP / AF to negotiate the charging information for AI / ML model transmission through the NEF. Based on the parameters of the request and the parameters in the received analysis information, such as the federated learning group identifier, the application identifier participating in federated learning, the size of the AI / ML model being transmitted, the area information of the AI / ML model being used, the time period of the AI / ML model transmission, the service quality parameter set (including the service quality requirements used to indicate the service quality flow of the transmitted AI / ML model and / or the specific service quality requirements used to indicate the transmitted AI / ML model), the transaction reference identifier, and the sponsorship status, it determines whether to accept the charging / sponsorship request from the ASP / AF, and feeds back to the ASP / AF whether it accepts the ASP / AF as the charging party / sponsor for the AI / ML model transmission or does not accept the ASP / AF as the charging party / sponsor for the AI / ML model transmission.

[0139] For example, in Implementation Example 1 (PCF subscribes to the NWDAF for analytics information transmitted by AI / ML models in order to negotiate fees with the ASP / AF).

[0140] See Figure 4 As shown, Figure 4 This is a schematic diagram of the signaling flow of the information processing method based on model transmission state analysis provided in the embodiments of this application. Figure 4 This is a signaling interaction diagram between ASP / AF and NWDAF, NEF, and PCF in the model-based transport state analysis-based information processing method. The model-based transport state analysis-based information processing method provided in this embodiment includes the following steps: (wherein, the AF is in an untrusted area as an example.)

[0141] Step 4011: PCF sends an AI / ML model open transport state subscription request (Nnwdaf_AnalyticsSubscription_Subscribe) to NWDAF.

[0142] Among them, the AI / ML model transmission status subscription Nnwdaf_AnalyticsSubscription_Subscribe request can be used as the first message.

[0143] In this embodiment, the request may carry a network data analysis identifier (i.e., Analytics ID), the identifier of a user equipment (UE) or a group of UEs receiving the AI / ML model, or any UE that meets the analysis conditions (i.e., Target of Analytics Reporting), the identifier of the application using the AI / ML model (i.e., Application ID), the area of ​​AI / ML model transmission (i.e., Area of ​​Interest (AoI)), the network slice (i.e., S-NSSAI) indicating the Protocol Data Unit (PDU) session transmitting the AI / ML model Quality of Service (QoS) flow, the data network (i.e., DNN) indicating the PDU session transmitting the AI / ML model QoS flow, the time period of AI / ML model transmission (i.e., Model transmission duration), the start timestamp of AI / ML model transmission (i.e., Model transmission start), the end timestamp of AI / ML model transmission (i.e., Model transmission stop), the size of the AI / ML transmission model (i.e., Model size), QoS requirements (including QoS requirements used to indicate the QoS flow of the AI / ML model transmission (i.e., 5QI (5G QoS Identifier)) and / or QoS characteristics used to indicate specific QoS requirements for the transmission of the AI / ML model (i.e., QoS Characteristics)), and the identifier of the federated learning group used to indicate the analysis (i.e., Federated). The system requests analysis information on the transmission status of AI / ML models in the network, including the learning(FL)group ID, the UE identifier or UE(s) identifier participating in federated learning (i.e., Federated Learning(FL)UE ID or UE group ID), the application identifier participating in federated learning (i.e., Federated Learning(FL)Application ID), transaction reference identifier, sponsorship status, etc.

[0144] Step 4012: NWDAF calls Nnwdaf_AnalyticsSubscription_Notify (i.e., analysis subscription notification) to send analysis information about the AI / ML model transmission status to PCF.

[0145] Specifically, NWDAF first calls Nnf_EventExposure_Subscribe (i.e., event open subscription) to send a data collection request to 5GCNF(s) to analyze the transmission status of AI / ML models in the network. NWDAF can send the data collection request to the 5GC NF(s) by calling Nnf_EventExposure_Subscribe. Then, the 5GC NF(s) calls Nnf_EventExposure_Notify (i.e., event open notification) to send the required data back to NWDAF.

[0146] Step 4013: ASP / AF sends an Nnef_ChargeableParty_Create request message to NEF. This message carries request information such as Federated Learning (FL) group ID, Application ID, model size, Validity area, Model transmission duration, QoS Parameter Sets, TransactionReference ID, and Sponsoring Status, requesting negotiation on charging information for AI / ML model transmission.

[0147] Step 4014: NEF interacts with PCF by triggering the Npcf_Policy_Create request (i.e., policy creation request) message, carrying the same information as the request in step 4013, that is, carrying the same request information as in step 4013.

[0148] The Npcf_Policy_Create request message can be used as a second message.

[0149] Step 4015: PCF sends a message to NEF indicating whether to accept the request by triggering the Npcf_Policy_Create response (i.e., the policy creation response).

[0150] In this embodiment, PCF determines whether to accept the ASP / AF's charging / sponsorship request based on the analysis information of the AI / ML model transmission status and the request information in step 4014, thereby setting whether ASP / AF can become the charging party / sponsor of AI / ML model transmission.

[0151] Step 4016: NEF sends an Nnef_ChargeableParty_Create response message to AF.

[0152] The Nnef_ChargeableParty_Create response message can be used as the first result.

[0153] Specifically, the ASP / AF requests the PCF to become the fee-payer / sponsor for AI / ML model transmission. The PCF determines whether to accept the request. If accepted, the ASP / AF can become the fee-payer / sponsor; otherwise, it cannot.

[0154] Optionally, the information processing further includes PDU session charge statistics; the second message is also used to request negotiation of PDU session charge statistics information for AI / ML model transmission;

[0155] The PDU session used for AI / ML model transmission is created by the user equipment (UE). The parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model (i.e., effective area information of the AI / ML model), time period of AI / ML model transmission, and quality of service parameter set.

[0156] Accordingly, the parameters requested in the first message include the parameters in the PDU session.

[0157] In this embodiment, if the PCF is in the trusted area, it directly receives the request from the ASP / AF for PDU session charge statistics transmitted by the negotiated AI / ML model; if the PCF is not in the trusted area, it receives the request from the ASP / AF for PDU session charge statistics transmitted by the negotiated AI / ML model through the NEF.

[0158] The user equipment (UE) creates a PDU session for AI / ML model transmission. The parameters requested in the first message include parameters from the PDU session, such as: UE identifier or UE(s) (i.e., user equipment group) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set.

[0159] Optionally, when the information processing includes PDU session charge statistics, the step of processing the AI / ML model transmission based on the analysis information can be achieved through the following steps:

[0160] Step b1: Based on the analysis information and the parameters requested in the second message, determine the billing rules for AI / ML model transmission in the PCC rules.

[0161] Step b2: Send a second result to the Session Management Function (SMF), the second result including the determined PCC rules.

[0162] Step b3: Receive the information from the User Plane Function (UPF) report and session / user subscription information sent by the SMF, and charge the UE's AI / ML model transmission PDU session based on the information from the UPF report, session / user subscription information, and other billing function entities.

[0163] The information reported by the UPF is determined by the UPF based on the usage reporting rules for data packets used to transmit AI / ML models. The usage reporting rules are determined by the SMF based on the received and determined PCC rules.

[0164] In this embodiment, billing rules, such as charging models and rates, are defined. The PCF obtains the corresponding parameters from the received analysis information based on the parameters requested in the second message. Then, based on these parameters, it determines (or updates) the billing rules for AI / ML model transmission in the PCC rules and sends the determined (or updated) PCC rules to the SMF. The SMF then determines the information reported by the UPF and the session / user subscription information based on the determined (or updated) PCC rules and sends it to the PCF, thus realizing PDU session charging statistics.

[0165] Optionally, the billing rules include at least one of the following:

[0166] No payment required;

[0167] Different rates apply depending on the size of the AI / ML model being transmitted; the larger the AI / ML model being transmitted, the higher the rate for transmitting the AI / ML model.

[0168] Different rates are applied based on the length of the AI / ML model transmission period; the longer the AI / ML model transmission time, the higher the rate applied to AI / ML model transmission.

[0169] Different rates are applied based on a weighted average of the size of the AI / ML model being transmitted, the region where the AI / ML model is used, and the time period for which the AI / ML model is transmitted; where a higher weighted average of these factors results in a higher rate for AI / ML model transmission.

[0170] Different rates are applied based on the set of service quality parameters; among them, the higher the service quality stream bit rate for AI / ML model transmission, the higher the rate applied to AI / ML model transmission.

[0171] Whether or not a fee is charged and the corresponding rate is determined based on whether or not the Federal Learning Group logo is present.

[0172] Specifically, the billing rules include one or more of the following information:

[0173] a. Volume-based charging rates (different rates are applied based on the model size; larger AI / ML models are charged at higher rates).

[0174] b. Time-based charging rates (different rates are applied based on the length of the model transmission duration; higher rates are applied for longer AI / ML model transmission times);

[0175] c. Volume and time based charging rates (different rates are applied based on the weighted values ​​of model size, validity area, and model transmission duration; the higher the weighted value of these three parameters, the higher the rate applied to AI / ML model transmission).

[0176] d. QoS-based charging rates (applying different rates based on QoS Parameter Sets; for example, a higher QoS flow bit rate for AI / ML model transmission results in a higher rate).

[0177] e. Event-based charging, rates (setting whether to charge based on the presence or absence of a Federated Learning (FL) group ID, and determining the corresponding rate);

[0178] f. No charging.

[0179] For example, in Embodiment 2 (PCF subscribes to the NWDAF for analytics information transmitted by AI / ML models to collect PDU session charges).

[0180] See Figure 5 As shown, Figure 5A signaling flow diagram of an information processing method based on model transmission state analysis provided in another embodiment of this application. Figure 5 This is a signaling interaction diagram between ASP / AF and NWDAF, NEF, UPF, SMF, and PCF in the model-based transport state analysis-based information processing method. The model-based transport state analysis-based information processing method provided in this embodiment includes the following steps: (wherein, ASP / AF is used in an untrusted area as an example.)

[0181] Step 5010: AI / ML Model Transfer PDU session establishment.

[0182] Specifically, when a UE creates an AI / ML model transmission PDU session, the information carried, such as UE ID or UE group ID (i.e., UE identifier or UE(s) identifier), Analytics ID (i.e., network data analysis identifier), Application ID, AoI (Area of ​​Interest), Model transmission duration, and QoS Parameter Sets, indicates that the PDU session will transmit AI / ML models.

[0183] Step 5011: If PCF has not yet subscribed to the analysis information on the transmission status of AI / ML models in the network from NWDAF, then send an Nnwdaf_AnalyticsSubscription_Subscribe request to NWDAF.

[0184] The request carries information such as UE ID or UE group ID, Analytics ID, Application ID, AoI (Area of ​​Interest), Model transmission duration, and QoS Parameter Sets, and subscribes to analysis information on the transmission status of AI / ML models in the network.

[0185] Step 5012: NWDAF sends Nnwdaf_AnalyticsSubscription_Notify to PCF, which means that NWDAF provides PCF with analysis information on the transmission status of AI / ML models in the network.

[0186] Specifically, when the NWDAF receives a request carrying this information, it collects information from some network elements AMF, UPF, and AF, analyzes and outputs analysis information on the model transmission status.

[0187] Step 5013: ASP / AF sends an Nnef_ChargeableParty_Create request message to NEF. This message carries the Federated Learning (FL) group ID, Application ID, model size, Validityarea, Model transmission duration, QoS Parameter Sets, Transaction Reference ID, Sponsoring Status, and PDU session charge statistics for the requested AI / ML model transmission.

[0188] Step 5014: NEF sends an Npcf_Policy_Create request message to PCF. This message carries the same information as the request made in step 5013.

[0189] Step 5015: PCF sends an Npcf_Policy_Create response to NEF.

[0190] Specifically, PCF determines (or updates) the billing rules (e.g., charging mode, rate) for AI / ML model transmission in the PCC rules based on the analysis information of AI / ML model transmission status and the request information in step 5014.

[0191] Step 5016: The PCF sends Npcf_SMPolicyControl_UpdateNotify (i.e., Session Policy Control Update Notification) to the SMF. This notification carries the PCC rules determined (or updated) in step 5015.

[0192] Step 5017: The SMF charges the UE's AI / ML Model Transfer PDU session according to the received PCC rules.

[0193] Specifically, the SMF, based on the received PCC rules, obtains the Usage Reporting Rule (URR) for data packets transmitting AI / ML models and sends it to the UPF, which manages the user plane of the AI / ML model transmission PDU session. The UPF then uses this URR to perform statistical analysis on the data packets transmitting AI / ML models (e.g., analyzing model size, QoS flow, etc.) and reports it to the SMF. The SMF then reports the information from the UPF along with session / user subscription information to the PCF and other charging entities, thereby enabling the execution of charges for the UE's AI / ML model transmission PDU session.

[0194] 5018. NEF sends an Nnef_ChargeableParty_Create response message to AF.

[0195] Optionally, the parameters requested in the first message include at least one of the following: the identifier of the application using the AI / ML model, the network slice of the PDU session used to transmit the AI / ML model service quality flow, the data network name of the PDU session used to transmit the AI / ML model service quality flow, and the service quality parameter set; the analysis information includes the analysis results corresponding to the parameters requested in the first message; wherein, the information processing includes session management policy decision processing.

[0196] Accordingly, the information processing of the AI / ML model transmission based on the analysis information can be achieved through the following steps:

[0197] Step c1: Based on the analysis information, determine the authorized service quality parameters for AI / ML model transmission in the PCC rules.

[0198] Step c2: Take the authorized service quality parameters transmitted by the AI / ML model in the determined PCC rule as the latest session management policy information, and send a third message carrying the latest session management policy information to the session management function SMF. The third message is used to request the SMF to update the session management policy.

[0199] Step c3: Receive the second result sent by the SMF. The second result is determined by the SMF based on the latest session management policy information. The second result includes updating the session management policy or not updating the session management policy.

[0200] In this embodiment, the PCF determines the authorized quality of service (QoS) parameters for AI / ML model transmission in the PCC rules based on the application identifier using the AI / ML model, the network slice of the PDU session used to transmit the AI / ML model QoS flow, the data network name of the PDU session used to transmit the AI / ML model QoS flow, and the QoS parameter set. Then, based on these authorized QoS parameters, it sends a request to the SMF to update the session management policy, carrying the latest session management policy information. The SMF determines whether to update the session management policy based on the received latest session management policy information and sends the result (updated or not updated) to the PCF, thus enabling the PCF to perform session management policy decision-making for AI / ML model transmission.

[0201] Optionally, determining the authorized service quality parameters for AI / ML model transmission in the PCC rules based on the analysis information includes:

[0202] If the data rate of the AI / ML model transmission in the analysis information is detected to be too low, then the priority of the 5G QoS Identifier in the authorized QoS parameters of the AI / ML model transmission, or the Reflective QoS Control, the maximum bit rate (UL-maximum bitrate) in the uplink direction of the transmitted AI / ML model, the maximum bit rate (DL-maximum bitrate) in the downlink direction of the transmitted AI / ML model, the minimum bit rate (UL-guaranteed bitrate) in the uplink direction of the transmitted AI / ML model, and the minimum bit rate (DL-guaranteed bitrate) in the downlink direction of the transmitted AI / ML model are adjusted.

[0203] For example, in Implementation Example 3 (PCF subscribes to the analysis information transmitted by the AI / ML model from NWDAF to make SM strategy decisions.)

[0204] See Figure 6 As shown, Figure 6 This is a signaling flow diagram of an information processing method based on model transmission state analysis provided in another embodiment of this application. Figure 6 This is a signaling interaction diagram between NWDAF, SMF, and PCF in the model-based transport state analysis-based information processing method. The model-based transport state analysis-based information processing method provided in this embodiment includes the following steps:

[0205] Step 6011: PCF sends an Nnwdaf_AnalyticsSubscription_Subscribe request to NWDAF. This request carries the Application ID (identifier of the application using the AI / ML model), S-NSSAI (network slice of the PDU session used to transmit the AI / ML model QoS flow), DNN (data network name), QoS Parameter Sets, and other information to subscribe to analysis information on the transmission status of AI / ML models in the network.

[0206] 6012. NWDAF calls Nnwdaf_AnalyticsSubscription_Notify to send analysis information on the AI / ML model transmission status to PCF. In other words, NWDAF provides PCF with analysis information on the transmission status of AI / ML models in the network.

[0207] 6013. Based on the analysis information obtained in step 6012, PCF makes a policy decision.

[0208] Specifically, based on the analysis information obtained in step 6012, such as the analysis results of S-NSSAI, the data rate of AI / ML model transmission in DNN, and QoS Parameter, the PCF determines and / or modifies the authorized QoS (Quality of Service) parameters of AI / ML model transmission in the PCC rules (e.g., if the data rate of AI / ML model transmission is detected to be too low, the priority of 5QI in the authorized QoS parameters of this model transmission or the Reflective QoS Control, UL-maximum bitrate, DL-maximum bitrate, UL-guaranteed bitrate, DL-guaranteed bitrate, etc. in its authorized QoS parameters).

[0209] 6014. The PCF sends an Npcf_SMPolicyControl_UpdateNotify request to the SMF. This request contains the latest SM policy information regarding AI / ML model transmission from step 6013.

[0210] 6015. The SMF sends an Npcf_SMPolicyControl_UpdateNotify response to the PCF. That is, the SMF acknowledges the PCF request with the Npcf_SMPolicyControl_UpdateNotify response.

[0211] In this embodiment, within the 5GC system, the PCF can obtain the analysis results of AI / ML model transmission from the NWDAF. Based on this analysis information, the PCF can make corresponding decisions (including charging decisions and SM strategies).

[0212] Figure 7 A flowchart illustrating an information processing method based on model transmission state analysis is provided in another embodiment of this application, as shown below. Figure 7 As shown, the execution entity of the information processing method based on model transfer state analysis provided in this embodiment is SMF. Therefore, the information processing method based on model transfer state analysis provided in this application embodiment includes the following steps:

[0213] Step 701: Send a first message to the Network Data Analysis Function (NWDAF). The first message is used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network.

[0214] Step 702: Receive the analysis information sent by NWDAF.

[0215] Step 703: Based on the analysis information, perform information processing on the AI / ML model transmission.

[0216] The analytics information is used to support session management.

[0217] In this embodiment, the first message is used to request analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network. This analysis information is determined by the NWDAF based on AI / ML model transmission status data sent by other network functions (5GC NF(s)) of the 5G core network. This AI / ML model transmission status data is collected by the NWDAF by sending a data collection request to the 5GC NF(s) according to the parameters requested in the received first message.

[0218] The parameters requested in the first message, the parameters requested in the data acquisition request, and the analysis information can refer to the parameters requested in the first message, the parameters requested in the data acquisition request, and the analysis information in the above embodiment of the information processing method based on model transmission state analysis executed by PCF as the execution subject.

[0219] In this embodiment, by sending a request to NWDAF to subscribe to the analysis information of the AI / ML model transmission status in the network, the system receives the analysis information determined by NWDAF based on the collected 5GC NF(s) data, and then processes the AI / ML model transmission based on the analysis information, thereby realizing operations such as session management in the AI / ML model transmission.

[0220] Optionally, the parameters requested in the first message include parameters from a PDU session for AI / ML model transmission, wherein the PDU session for AI / ML model transmission is created by a user equipment (UE); the parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set; the information processing includes session management.

[0221] Accordingly, the information processing of the AI / ML model transmission based on the analysis information can be achieved through the following steps:

[0222] Step d1: Based on the location of each UE, match the information of the optimal AI / ML model transmission service experience from the analysis information;

[0223] Step d2: Obtain the information of the service session anchor UPF from the information of the service experience transmitted by the optimal AI / ML model, and determine the service session anchor UPF as the new PDU session anchor UPF;

[0224] Step d3: The selected new PDU session anchor point UPF will be used to provide the optimal path for AI / ML model transmission for each UE.

[0225] In this embodiment, the SMF selects the optimal AI / ML model transmission service experience information based on the location and analysis information of each UE, obtains the service session anchor point (UPF) information from it, uses it as the new PDU session anchor point (UPF), and uses the selected new PDU session anchor point (UPF) to provide the optimal path for AI / ML model transmission for each UE.

[0226] Optionally, determining the service session anchor point UPF as the new PDU session anchor point UPF can be achieved through the following steps:

[0227] Step e1: In AI / ML model transmission, determine the optimal path for AI / ML model transmission based on the location of each UE and the information on the best AI / ML model transmission service experience.

[0228] Step e2: Determine the service session anchor point UPF corresponding to the optimal path as the new PDU session anchor point UPF.

[0229] In this embodiment, the SMF determines the optimal path for AI / ML model transmission based on the location of each UE and the information on the best AI / ML model transmission service experience, and selects the service session anchor point UPF corresponding to the optimal path as the new PDU session anchor point UPF.

[0230] Optionally, the method may further include:

[0231] Step f1: Before changing the PDU session anchor point UPF, send a fourth message to the application function AF. The fourth message is used to notify the user plane anchor point of the change and the target data network access identifier.

[0232] Step f2: Receive the confirmation result sent by AF.

[0233] In this embodiment, before changing the PDU session anchor point UPF, a notification of the user plane anchor point change and the target data network access identifier is sent to the AF. The AF confirms based on the notification and sends the confirmation result to the SMF. After receiving the confirmation, the SMF will use the selected new PDU session anchor point UPF to provide the UE with the optimal path for AI / ML model transmission.

[0234] For example, in Implementation Example 4 (SMF subscribes to the NWDAF for analytics information transmitted from the AI / ML model to select the best UPF).

[0235] See Figure 8 As shown, Figure 8 A signaling flow diagram of an information processing method based on model transmission state analysis provided in another embodiment of this application. Figure 8 This is a signaling interaction diagram between the SMF and the UE, AF, Newanchor UPF (i.e., the new PDU session anchor point UPF), NWDAF, UPF, and PCF in the model-based transmission state analysis-based information processing method. The model-based transmission state analysis-based information processing method provided in this embodiment includes the following steps:

[0236] Step 8010: AI / ML Model Transfer PDU session establishment.

[0237] Specifically, when a UE creates an AI / ML model transmission PDU session, the parameters it carries, such as UE ID or UE group ID, Application ID, AoI (Area of ​​Interest), Model transmission duration, and QoS Parameter Sets, indicate that the PDU session will transmit AI / ML models.

[0238] Step 8011: If the SMF has not yet subscribed to the analysis information on the transmission status of AI / ML models in the network from the NWDAF, then send an Nnwdaf_AnalyticsSubscription_Subscribe request to the NWDAF.

[0239] The request carries information such as UE ID or UE group ID, Analytics ID, Application ID, AoI (Area of ​​Interest), Model transmission duration, and QoS Parameter Sets, and subscribes to analysis information on the transmission status of AI / ML models in the network.

[0240] Step 8012: NWDAF sends Nnwdaf_AnalyticsSubscription_Notify to SMF, which means that NWDAF provides SMF with analysis information on the transmission status of AI / ML models in the network.

[0241] Step 8013: The SMF determines the UPF re-allocation (i.e., UPF reallocation) based on the location of each UE and the analysis information.

[0242] Specifically, based on the UE's location, the SMF will further match the information of the optimal AI / ML model transmission service experience from the analysis results of the AI / ML model transmission status, and obtain the serving anchor UPF information from it. The SMF will comprehensively consider the UE's location and the model transmission service experience in the AI / ML model transmission to select the optimal path and determine the serving anchor UPF (i.e., the service session anchor UPF) as the new anchor UPF (i.e., the new PDU session anchor UPF).

[0243] Step 8014: SMF sends Nsmf_EventExpose_Notification (i.e., event exposure notification) to AF.

[0244] Specifically, before changing the anchor UPF (i.e., PDU session anchor UPF), the SMF sends a notification to the AF that the user plane anchor has changed and the target DNAI.

[0245] Step 8015: AF sends Nsmf_AppRelocationInfo (i.e., application relocation information) to SMF. This means AF sends an acknowledgment to SMF.

[0246] 8016. After receiving the confirmation, the SMF will use the selected new anchor UPF (i.e., the new PDU session anchor UPF) to provide the UE with the optimal path for AI / ML model transmission.

[0247] It should be noted that steps 8014 and 8015 can also be executed after step 8016. No specific restrictions are made here.

[0248] Optionally, the information processing of the AI / ML model transmission based on the analysis information can also be achieved through the following steps:

[0249] Step g1: Based on the location of each UE, match the information of the optimal AI / ML model transmission service experience from the analysis information;

[0250] Step g2: Obtain the information of the service session anchor point UPF and the analysis results of the service quality parameter set from the information of the optimal AI / ML model transmitting service experience;

[0251] Step g3: Based on the information of the service session anchor UPF and the analysis results of the service quality parameter set, determine to establish a new PDU session anchor UPF to transmit the AI / ML model. The new PDU session anchor UPF is the second PDU session anchor UPF, where the first PDU session anchor UPF is the current PDU session anchor for transmitting the AI / ML model.

[0252] Step g4: Select a User Plane Function (UPF) as the branch point (BP) or uplink classifier (ULCL) of the PDU session;

[0253] Step g5: Provide the BP or UL CL with a traffic filter corresponding to the first PDU session anchor UPF and the second PDU session anchor UPF, and instruct the BP or UL CL to forward the QoS stream / data packet transmitted by the uplink AI / ML model to the second PDU session anchor UPF;

[0254] Step g6: Use the created second PDU session anchor point UPF to provide each UE with the optimal path for information related to the quality of service flow in AI / ML model transmission.

[0255] In this embodiment, the SMF selects the optimal AI / ML model transmission service experience information based on the location of each UE and analysis information, and obtains the service session anchor point UPF information and service quality parameter set analysis results from it to establish a new PDU session anchor point UPF for transmitting the AI / ML model. Then, the SMF determines and selects a user plane function UPF as the branch point BP or UL CL of the PDU session, and provides it with traffic filters corresponding to the first PDU session anchor point UPF (i.e., the currently used PDU session anchor point UPF) and the second PDU session anchor point UPF (i.e., the new PDU session anchor point UPF), and instructs the BP or UL CL to forward the service quality flow / data packet of the uplink AI / ML model transmission to the new PDU session anchor point UPF, and uses the newly created PDU session anchor point UPF to provide each UE with the optimal path for service quality flow related information in AI / ML model transmission.

[0256] Optionally, determining the establishment of a new PDU session anchor for transmitting AI / ML models based on the information of the service session anchor UPF and the analysis results of the service quality parameter set can be achieved through the following steps:

[0257] Step h1: Based on the information of the service session anchor point UPF and the analysis results of the service quality parameter set, determine the optimal path for AI / ML model transmission;

[0258] Step h2: Based on the optimal path, determine and establish a new PDU session anchor point UPF to transmit the AI / ML model.

[0259] In this embodiment, the SMF selects the optimal path for AI / ML model transmission based on the information of the service session anchor point UPF and the analysis results of the service quality parameter set, and establishes a new PDU session anchor point UPF based on the optimal path to transmit the AI / ML model.

[0260] Optionally, the method may further include:

[0261] Step i1: Before creating a new PDU session anchor point UPF to transmit the AI / ML model, send a fifth message to the application function AF. The fifth message is used to notify the user plane anchor point of the change and the target data network access identifier.

[0262] Step i2: Receive the confirmation result sent by AF.

[0263] In this embodiment, before creating a new PDU session anchor point UPF to transmit the AI / ML model, a notification of user plane anchor point change and target data network access identifier is sent to the AF. The AF confirms based on this notification and sends the confirmation result to the SMF. After receiving the confirmation, the SMF will use the new PDU session anchor point UPF to provide the UE with the optimal path for QoS flow-related information in the AI / ML model transmission.

[0264] For example, in Example 5 (SMF subscribes to the NWDAF for analysis information transmitted by the AI / ML model to set BP or ULCL.)

[0265] See Figure 9 As shown, Figure 9 This is a schematic diagram of the signaling flow of an information processing method based on model transmission state analysis provided in another embodiment of this application. Figure 9 This is a signaling interaction diagram between SMF and UE, AF, UPF (where UPF includes Branching Point or UL CL, PSA1 (current or initial PDU session anchor point UPF), and PSA1 (new PDU session anchor point UPF)), NWDAF, UPF, and PCF in the model-based transmission state analysis-based information processing method. The model-based transmission state analysis-based information processing method provided in this embodiment includes the following steps:

[0266] Step 9010: AI / ML Model Transfer PDU session establishment.

[0267] Specifically, when a UE creates an AI / ML model transmission PDU session, the parameters it carries, such as UE ID or UE group ID, Application ID, AoI (Area of ​​Interest), Model transmission duration, and QoS Parameter Sets, indicate that the PDU session will transmit AI / ML models.

[0268] Step 9011: If the SMF has not yet subscribed to the analysis information on the transmission status of AI / ML models in the network from the NWDAF, then send an Nnwdaf_AnalyticsSubscription_Subscribe request to the NWDAF.

[0269] The request carries information such as UE ID or UE group ID, Analytics ID, Application ID, AoI (Area of ​​Interest), Model transmission duration, and QoS Parameter Sets, and subscribes to analysis information on the transmission status of AI / ML models in the network.

[0270] Step 9012: NWDAF sends Nnwdaf_AnalyticsSubscription_Notify to SMF, which means that NWDAF provides SMF with analysis information on the transmission status of AI / ML models in the network.

[0271] Step 9013: Based on the location of each UE and the analysis information, the SMF determines to select a UPF as the branching point (BP) or UL CL of the PDU session.

[0272] Specifically, based on the UE's location, the SMF will further analyze the AI / ML model transmission status to identify the optimal AI / ML model transmission service experience. This includes obtaining serving anchor UPF information and analyzing QoS ParameterSets. The SMF will comprehensively consider the UE's location and the model transmission service experience to select the optimal path during AI / ML model transmission. Based on this, a new PDU session anchor, PSA2, will be established to transmit the AI / ML model.

[0273] The SMF selects a UPF as the branch point BP (in the case of IPv6 multi-homing) or UL CL (Uplink Classifier) ​​for the PDU session. The SMF provides the UPF (BP / UL CL) with traffic filters corresponding to PSA1 and PSA2, instructing the UPF to forward the QoS flow / packets transmitted by the uplink AI / ML model to PSA2.

[0274] Step 9014: SMF sends Nsmf_EventExpose_Notification to AF.

[0275] Specifically, before deciding to establish PSA2, SMF sends a notification to AF, in accordance with existing technology, regarding changes to the user plane anchor point and the target DNAI.

[0276] Step 9015: AF sends Nsmf_AppRelocationInfo to SMF. That is, AF sends an acknowledgment to SMF.

[0277] Step 9016: After receiving the confirmation, the SMF will use the newly added PSA2 to provide the UE with the optimal path for QoS flow-related information in AI / ML model transmission (i.e., The PDU Session Anchor 2 is used to provide model transmission service for UE).

[0278] It should be noted that steps 9014 and 9015 can also be executed after step 9016. No specific restrictions are made here.

[0279] In this embodiment, in the 5GC system, the SMF can obtain the analysis results of AI / ML model transmission from the NWDAF. Based on the UE's location and the analysis results, the SMF can create the optimal path for the best service experience of AI / ML model transmission, realize session management, and thus ensure the service experience and performance of AI / ML model transmission.

[0280] Figure 10 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application is shown below. Figure 10 As shown, the information processing device based on model transmission state analysis provided in this embodiment is applied to PCF. The information processing device based on model transmission state analysis provided in this embodiment includes: a transceiver 1000, used to receive and send data under the control of a processor 1010.

[0281] Among them, Figure 10 In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 1010) and memory (memory 1020). The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 1000 can be multiple elements, including transmitters and receivers, providing units for communicating with various other devices over transmission media, including wireless channels, wired channels, optical fibers, etc. The processor 1010 is responsible for managing the bus architecture and general processing, and the memory 1020 can store data used by the processor 1010 during operation.

[0282] The processor 1010 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a complex programmable logic device (CPLD). The processor can also adopt a multi-core architecture.

[0283] In this embodiment, the memory 1020 is used to store computer programs; the transceiver 1000 is used to send and receive data under the control of the processor 1010; the processor 1010 is used to read the computer programs in the memory and perform the following operations:

[0284] Send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0285] Receive analysis information sent by NWDAF;

[0286] Based on the analysis information, information processing is performed on the AI / ML model transmission.

[0287] Optionally, the information processing includes fee / sponsorship negotiation;

[0288] The step of processing the AI / ML model transmission based on the analysis information includes:

[0289] The system receives a second message, either directly or through the Network Capability Opening Function (NEF), from an Application Service Provider (ASP) or Application Function (AF). This second message requests negotiation on pricing information for AI / ML model transmission. The parameters requested in the second message include at least one of the following: Federated Learning Group Identifier, Identifier of the application using the AI / ML model, Model Size for AI / ML Transmission, Region Information for AI / ML Model Use, Time Period for AI / ML Model Transmission, Quality of Service (QoS) Parameter Set, Transaction Reference Identifier, and Sponsorship Status.

[0290] Based on the analysis information and the second message, determine whether to accept the ASP / AF's fee / sponsorship request, and send a first result directly or through NEF to the ASP / AF; the first result includes accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission or not accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission.

[0291] Optionally, the information processing further includes PDU session charge statistics; the second message is also used to request negotiation of PDU session charge statistics information for AI / ML model transmission;

[0292] The PDU session used for AI / ML model transmission is created by the user equipment (UE). The parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set.

[0293] Accordingly, the parameters requested in the first message include the parameters in the PDU session.

[0294] Optionally, when the processor 1010 processes the AI / ML model transmission based on the analysis information, it further includes:

[0295] Based on the analysis information and the parameters requested in the second message, determine the billing rules for AI / ML model transmission in the PCC rules;

[0296] Send a second result to the Session Management Function (SMF), the second result including the determined PCC rules;

[0297] The system receives information from the User Plane Function (UPF) report and session / user subscription information sent by the SMF, and charges the UE's AI / ML model transmission PDU session based on the information from the UPF report, session / user subscription information, and other charging function entities.

[0298] The information reported by the UPF is determined by the UPF based on the usage reporting rules for data packets used to transmit AI / ML models. The usage reporting rules are determined by the SMF based on the received and determined PCC rules.

[0299] Optionally, the billing rules include at least one of the following:

[0300] No payment required;

[0301] Different rates apply depending on the size of the AI / ML model being transmitted; the larger the AI / ML model being transmitted, the higher the rate for transmitting the AI / ML model.

[0302] Different rates are applied based on the length of the AI / ML model transmission period; the longer the AI / ML model transmission time, the higher the rate applied to AI / ML model transmission.

[0303] Different rates are applied based on a weighted average of the size of the AI / ML model being transmitted, the region where the AI / ML model is used, and the time period during which the AI / ML model is transmitted; where a higher weighted average of the size of the AI / ML model being transmitted, the effective region where the AI / ML model is used, and the time period during which the AI / ML model is transmitted results in a higher rate for AI / ML model transmission.

[0304] Different rates are applied based on the set of service quality parameters; among them, the higher the service quality stream bit rate for AI / ML model transmission, the higher the rate applied to AI / ML model transmission.

[0305] Whether or not a fee is charged and the corresponding rate is determined based on whether or not the Federal Learning Group logo is present.

[0306] Optionally, the parameters requested in the first message include at least one of the following: the identifier of the application using the AI / ML model, the network slice of the PDU session used to transmit the AI / ML model service quality flow, the data network name of the PDU session used to transmit the AI / ML model service quality flow, and the service quality parameter set; the analysis information includes the analysis results corresponding to the parameters requested in the first message; wherein, the information processing includes session management policy decision processing;

[0307] Accordingly, the processor 1010 is used to process the AI / ML model transmission based on the analysis information, specifically including:

[0308] Based on the analysis information, determine the authorized service quality parameters for AI / ML model transmission in the PCC rules;

[0309] The authorized service quality parameters transmitted by the AI / ML model in the determined PCC rules are used as the latest session management policy information, and a third message carrying the latest session management policy information is sent to the session management function SMF. The third message is used to request the SMF to update the session management policy.

[0310] The system receives a second result sent by the SMF, which is determined by the SMF based on the latest session management policy information. The second result may include updating the session management policy or not updating the session management policy.

[0311] Optionally, the processor 1010, when determining the authorized service quality parameters for AI / ML model transmission in the PCC rules based on the analysis information, specifically includes:

[0312] If the data rate of the AI / ML model transmission in the analysis information is detected to be too low, then the priority of the 5G service quality identifier in the authorized service quality parameters of the AI / ML model transmission, or the reflective service quality control, the maximum bit rate of the uplink direction of the transmitted AI / ML model, the maximum bit rate of the transmitted AI / ML model in the downlink direction, the minimum bit rate of the transmitted AI / ML model in the uplink direction, and the minimum bit rate of the transmitted AI / ML model in the downlink direction in the authorized service quality parameters are adjusted.

[0313] It should be noted that the information processing device based on model transmission state analysis provided in this application can achieve... Figures 3-6 All method steps implemented in the method embodiment shown are capable of achieving the same technical effect. Therefore, the parts that are the same as those in the method embodiment and their beneficial effects will not be described in detail here.

[0314] Figure 11 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application is shown below. Figure 11 As shown, the information processing device based on model transmission state analysis provided in this embodiment is applied to PCF. Therefore, the information processing device 1100 based on model transmission state analysis provided in this embodiment includes:

[0315] The sending unit 1101 is used to send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0316] The receiving unit 1102 is used to receive the analysis information sent by NWDAF;

[0317] The processing unit 1103 is used to process the AI / ML model transmission based on the analysis information.

[0318] Optionally, the information processing includes fee / sponsorship negotiation;

[0319] The processing unit 1103 is specifically used for:

[0320] The system receives a second message, either directly or through the Network Capability Opening Function (NEF), from an Application Service Provider (ASP) or Application Function (AF). This second message requests negotiation on pricing information for AI / ML model transmission. The parameters requested in the second message include at least one of the following: Federated Learning Group Identifier, Identifier of the application using the AI / ML model, Model Size for AI / ML Transmission, Region Information for AI / ML Model Use, Time Period for AI / ML Model Transmission, Quality of Service (QoS) Parameter Set, Transaction Reference Identifier, and Sponsorship Status.

[0321] Based on the analysis information and the second message, determine whether to accept the ASP / AF's fee / sponsorship request, and send a first result directly or through NEF to the ASP / AF; the first result includes accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission or not accepting the ASP / AF's fee / sponsorship request for AI / ML model transmission.

[0322] Optionally, the information processing further includes PDU session charge statistics; the second message is also used to request negotiation of PDU session charge statistics information for AI / ML model transmission;

[0323] The PDU session used for AI / ML model transmission is created by the user equipment (UE). The parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set.

[0324] Accordingly, the parameters requested in the first message include the parameters in the PDU session.

[0325] Optionally, the processing unit 1103 is further specifically used for:

[0326] Based on the analysis information and the parameters requested in the second message, determine the billing rules for AI / ML model transmission in the PCC rules;

[0327] Send a second result to the Session Management Function (SMF), the second result including the determined PCC rules;

[0328] The system receives information from the User Plane Function (UPF) report and session / user subscription information sent by the SMF, and charges the UE's AI / ML model transmission PDU session based on the information from the UPF report, session / user subscription information, and other charging function entities.

[0329] The information reported by the UPF is determined by the UPF based on the usage reporting rules for data packets used to transmit AI / ML models. The usage reporting rules are determined by the SMF based on the received and determined PCC rules.

[0330] Optionally, the billing rules include at least one of the following:

[0331] No payment required;

[0332] Different rates apply depending on the size of the AI / ML model being transmitted; the larger the AI / ML model being transmitted, the higher the rate for transmitting the AI / ML model.

[0333] Different rates are applied based on the length of the AI / ML model transmission period; the longer the AI / ML model transmission time, the higher the rate applied to AI / ML model transmission.

[0334] Different rates are applied based on a weighted average of the size of the AI / ML model being transmitted, the region where the AI / ML model is used, and the time period during which the AI / ML model is transmitted; where a higher weighted average of the size of the AI / ML model being transmitted, the effective region where the AI / ML model is used, and the time period during which the AI / ML model is transmitted results in a higher rate for AI / ML model transmission.

[0335] Different rates are applied based on the set of service quality parameters; among them, the higher the service quality stream bit rate for AI / ML model transmission, the higher the rate applied to AI / ML model transmission.

[0336] Whether or not a fee is charged and the corresponding rate is determined based on whether or not the Federal Learning Group logo is present.

[0337] Optionally, the parameters requested in the first message include at least one of the following: the identifier of the application using the AI / ML model, the network slice of the PDU session used to transmit the AI / ML model service quality flow, the data network name of the PDU session used to transmit the AI / ML model service quality flow, and the service quality parameter set; the analysis information includes the analysis results corresponding to the parameters requested in the first message; wherein, the information processing includes session management policy decision processing;

[0338] Accordingly, the processing unit 1103 is specifically used for:

[0339] Based on the analysis information, determine the authorized service quality parameters for AI / ML model transmission in the PCC rules;

[0340] The authorized service quality parameters transmitted by the AI / ML model in the determined PCC rules are used as the latest session management policy information, and a third message carrying the latest session management policy information is sent to the session management function SMF. The third message is used to request the SMF to update the session management policy.

[0341] The system receives a second result sent by the SMF, which is determined by the SMF based on the latest session management policy information. The second result may include updating the session management policy or not updating the session management policy.

[0342] Optionally, the processing unit is specifically used for:

[0343] When the data rate of the AI / ML model transmission in the analysis information is detected to be too low, the priority of the 5G service quality identifier in the authorized service quality parameters of the AI / ML model transmission, or the reflective service quality control, the maximum bit rate of the uplink direction of the transmitted AI / ML model, the maximum bit rate of the transmitted AI / ML model in the downlink direction, the minimum bit rate of the transmitted AI / ML model in the uplink direction, and the minimum bit rate of the transmitted AI / ML model in the downlink direction are adjusted.

[0344] It should be noted that the information processing device based on model transmission state analysis provided in this application can achieve... Figures 3-6 All method steps implemented in the method embodiment can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiment and their beneficial effects will not be described in detail here.

[0345] Figure 12 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application is shown below. Figure 12 As shown, the information processing device based on model transmission state analysis provided in this embodiment is applied to SMF. The information processing device based on model transmission state analysis provided in this embodiment includes: a transceiver 1200, used to receive and send data under the control of a processor 1210.

[0346] Among them, Figure 12 In this context, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 1210) and memory (memory 1220). The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 1200 may be multiple elements, including transmitters and receivers, providing units for communicating with various other devices over transmission media, including wireless channels, wired channels, optical fibers, etc. The processor 1210 is responsible for managing the bus architecture and general processing, and the memory 1220 may store data used by the processor 1210 during operation.

[0347] The processor 1210 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a complex programmable logic device (CPLD). The processor can also adopt a multi-core architecture.

[0348] In this embodiment, the memory 1220 is used to store computer programs; the transceiver 1200 is used to send and receive data under the control of the processor; and the processor 1210 is used to read the computer programs from the memory and perform the following operations:

[0349] Send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0350] Receive analysis information sent by NWDAF;

[0351] Based on the analysis information, information processing is performed on the AI / ML model transmission.

[0352] Optionally, the parameters requested in the first message include parameters from a PDU session for AI / ML model transmission, wherein the PDU session for AI / ML model transmission is created by a user equipment (UE); the parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, valid area information for using the AI / ML model, time period for AI / ML model transmission, and quality of service parameter set; the information processing includes session management;

[0353] Processor 1210, used for processing information on AI / ML model transmission based on the analysis information, specifically includes:

[0354] Based on the location of each UE, information on the optimal AI / ML model transmission service experience is matched from the analysis information;

[0355] The information of the service session anchor UPF is obtained from the information of the service experience transmitted by the optimal AI / ML model, and the service session anchor UPF is determined to be the new PDU session anchor UPF.

[0356] The selected new PDU session anchor point UPF will be used to provide the optimal path for AI / ML model transmission for each UE.

[0357] Optionally, the processor 1210 is configured to determine, when using the service session anchor point UPF as a new PDU session anchor point UPF, specifically include:

[0358] In AI / ML model transmission, the optimal path for AI / ML model transmission is determined based on the location of each UE and the information on the best AI / ML model transmission service experience.

[0359] The service session anchor point UPF corresponding to the optimal path is determined as the new PDU session anchor point UPF.

[0360] Optionally, the processor 1210 also includes:

[0361] Before changing the PDU session anchor point UPF, a fourth message is sent to the application function AF. The fourth message is used to notify the user plane anchor point of the change and the target data network access identifier.

[0362] Receive the confirmation result sent by AF.

[0363] Optionally, when the processor 1210 processes the AI / ML model transmission based on the analysis information, it further includes:

[0364] Based on the location of each UE, information on the optimal AI / ML model transmission service experience is matched from the analysis information;

[0365] The information of the service session anchor point UPF and the analysis results of the service quality parameter set are obtained from the information of the service experience transmitted by the optimal AI / ML model.

[0366] Based on the information of the service session anchor UPF and the analysis results of the service quality parameter set, a new PDU session anchor UPF is determined to be established to transmit the AI / ML model. The new PDU session anchor UPF is the second PDU session anchor UPF, wherein the first PDU session anchor UPF is the PDU session anchor currently transmitting the AI / ML model.

[0367] Select either a User Plane Function (UPF) as the branch point (BP) or the uplink classifier (UL CL) of the PDU session;

[0368] Provide the BP or UL CL with a traffic filter corresponding to the first PDU session anchor UPF and the second PDU session anchor UPF, and instruct the BP or UL CL to forward the quality of service stream / data packet transmitted by the uplink AI / ML model to the second PDU session anchor UPF;

[0369] The created second PDU session anchor point UPF will be used to provide each UE with the optimal path for information related to the quality of service flow in the AI / ML model transmission.

[0370] Optionally, the processor 1210, when determining the establishment of a new PDU session anchor for transmitting AI / ML models based on the information of the service session anchor UPF and the analysis results of the service quality parameter set, specifically includes:

[0371] Based on the information from the service session anchor point UPF and the analysis results of the service quality parameter set, the optimal path for AI / ML model transmission is determined.

[0372] Based on the optimal path, a new PDU session anchor point UPF is determined to be established to transmit the AI / ML model.

[0373] Optionally, the processor 1210 also includes:

[0374] Before creating a new PDU session anchor point UPF to transfer the AI / ML model, the fourth message is sent to the application function AF.

[0375] Receive the confirmation result sent by AF.

[0376] It should be noted that the information processing device based on model transmission state analysis provided in this application can achieve... Figures 7-9 All method steps implemented in the method embodiment shown are capable of achieving the same technical effect. Therefore, the parts that are the same as those in the method embodiment and their beneficial effects will not be described in detail here.

[0377] Figure 13 A schematic diagram of the structure of an information processing device based on model transmission state analysis provided in another embodiment of this application is shown below. Figure 13 As shown, the information processing device based on model transfer state analysis provided in this embodiment is applied to SMF. Therefore, the information processing device 1300 based on model transfer state analysis provided in this embodiment includes:

[0378] The sending unit 1301 is used to send a first message to the Network Data Analysis Function (NWDAF), the first message being used to request subscription to analysis information on the transmission status of artificial intelligence / machine learning (AI / ML) models in the network;

[0379] The receiving unit 1302 is used to receive the analysis information sent by NWDAF;

[0380] The processing unit 1303 is used to process the AI / ML model transmission based on the analysis information.

[0381] Optionally, the parameters requested in the first message include parameters from a PDU session for AI / ML model transmission, wherein the PDU session for AI / ML model transmission is created by a user equipment (UE); the parameters in the PDU session include at least one of the following: UE identifier or user equipment group (UE(s)) identifier, network data analysis identifier, identifier of the application using the AI / ML model, area information of the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set; the information processing includes session management;

[0382] The processing unit is specifically used for:

[0383] Based on the location of each UE, information on the optimal AI / ML model transmission service experience is matched from the analysis information;

[0384] The information of the service session anchor UPF is obtained from the information of the service experience transmitted by the optimal AI / ML model, and the service session anchor UPF is determined to be the new PDU session anchor UPF.

[0385] The selected new PDU session anchor point UPF will be used to provide the optimal path for AI / ML model transmission for each UE.

[0386] Optionally, the processing unit is specifically used for:

[0387] In AI / ML model transmission, the optimal path for AI / ML model transmission is determined based on the location of each UE and the information on the best AI / ML model transmission service experience.

[0388] The service session anchor point UPF corresponding to the optimal path is determined as the new PDU session anchor point UPF.

[0389] Optionally, the device further includes: a notification unit; the notification unit is used for:

[0390] Before changing the PDU session anchor point UPF, a fourth message is sent to the application function AF. The fourth message is used to notify the user plane anchor point of the change and the target data network access identifier.

[0391] Receive the confirmation result sent by AF.

[0392] Optionally, the processing unit is also used for:

[0393] Based on the location of each UE, information on the optimal AI / ML model transmission service experience is matched from the analysis information;

[0394] The information of the service session anchor point UPF and the analysis results of the service quality parameter set are obtained from the information of the service experience transmitted by the optimal AI / ML model.

[0395] Based on the information of the service session anchor UPF and the analysis results of the service quality parameter set, a new PDU session anchor UPF is determined to be established to transmit the AI / ML model. The new PDU session anchor UPF is the second PDU session anchor UPF, wherein the first PDU session anchor UPF is the PDU session anchor currently transmitting the AI / ML model.

[0396] Select either a User Plane Function (UPF) as the branch point (BP) or the uplink classifier (UL CL) of the PDU session;

[0397] Provide the BP or UL CL with a traffic filter corresponding to the first PDU session anchor UPF and the second PDU session anchor UPF, and instruct the BP or UL CL to forward the quality of service stream / data packet transmitted by the uplink AI / ML model to the second PDU session anchor UPF;

[0398] The created second PDU session anchor point UPF will be used to provide each UE with the optimal path for information related to the quality of service flow in the AI / ML model transmission.

[0399] Optionally, the processing unit is also specifically used for:

[0400] Based on the information from the service session anchor point UPF and the analysis results of the service quality parameter set, the optimal path for AI / ML model transmission is determined.

[0401] Based on the optimal path, a new PDU session anchor point UPF is determined to be established to transmit the AI / ML model.

[0402] Optionally, the device further includes: a notification unit; the notification unit is used for:

[0403] Before creating a new PDU session anchor point UPF to transfer the AI / ML model, the fourth message is sent to the application function AF.

[0404] Receive the confirmation result sent by AF.

[0405] It should be noted that the information processing device based on model transmission state analysis provided in this application can achieve... Figures 7-9 All method steps implemented in the method embodiment can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiment and their beneficial effects will not be described in detail here.

[0406] It should be noted that the division of units in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units.

[0407] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0408] This application also provides a processor-readable storage medium. The processor-readable storage medium stores a computer program that causes the processor to execute any of the above-described method embodiments.

[0409] The processor-readable storage medium can be any available medium or data storage device that the processor can access, including but not limited to magnetic storage (e.g., floppy disks, hard disks, magnetic tapes, magneto-optical disks (MOs)), optical storage (e.g., CDs, DVDs, BDs, HVDs), and semiconductor storage (e.g., ROMs, EPROMs, EEPROMs, non-volatile memory (NAND flash), solid-state drives (SSDs)).

[0410] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0411] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-executable instructions. These computer-executable instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0412] These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the processor-readable memory produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0413] These processors can execute instructions that can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0414] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. An information processing method based on model-transfer state analysis, applied to policy control function (PCF), characterized in that, The method comprises: sending a first message to a network data analysis function (NWDAF), the first message being used for requesting subscription of analysis information of an artificial intelligence / machine learning (AI / ML) model transmission state in a network; receiving analysis information sent by the NWDAF; wherein the analysis information is used for charging / sponsorship negotiation, PDU session charging statistics, and session management policy decision-making; performing information processing on the AI / ML model transmission according to the analysis information.

2. The method of claim 1, wherein, The information processing comprises charging / sponsorship negotiation; and the information processing on the AI / ML model transmission according to the analysis information comprises: receiving a second message sent by an application service provider (ASP) or an application function (AF) directly or through a network capability exposure function (NEF), the second message being used for requesting negotiation of charging information of the AI / ML model transmission; wherein parameters requested in the second message comprise at least one of the following: a federated learning group identifier, an identifier of an application using the AI / ML model, a model size of the AI / ML transmission, regional information using the AI / ML model, a time period of the AI / ML model transmission, a quality of service parameter set, a transaction reference identifier, and a sponsorship state; determining whether to accept the charging / sponsorship request of the ASP / AF according to the analysis information and the second message, and sending a first result to the ASP / AF directly or through the NEF; the first result comprises acceptance of the ASP / AF as a charging party / sponsor of the AI / ML model transmission or non-acceptance of the ASP / AF as the charging party / sponsor of the AI / ML model transmission.

3. The method of claim 2, wherein, The information processing further comprises PDU session charging statistics; and the second message is further used for requesting negotiation of PDU session charging statistical information of the AI / ML model transmission. The PDU session used for the AI / ML model transmission is created by a user equipment (UE); and parameters in the PDU session comprise at least one of the following: an identifier of the UE or an identifier of a user equipment group (UE(s)), a network data analysis identifier, an identifier of an application using the AI / ML model, regional information using the AI / ML model, a time period of the AI / ML model transmission, and a quality of service parameter set. Correspondingly, parameters requested in the first message comprise the parameters in the PDU session.

4. The method of claim 3, wherein, The information processing on the AI / ML model transmission according to the analysis information further comprises: determining charging rules of the AI / ML model transmission in a PCC rule according to the analysis information and the parameters requested in the second message; sending a second result to a session management function (SMF), the second result comprising the determined PCC rule; receiving information reported by a user plane function (UPF) and session / user subscription information sent by the SMF, and performing charging on a PDU session of the AI / ML model transmission of the UE according to the information reported by the UPF, the session / user subscription information, and other charging function entities. The information reported by the UPF is determined by the UPF according to a usage reporting rule for the data packet of the AI / ML model, and the usage reporting rule is determined by the SMF according to the received determined PCC rule.

5. The method of claim 4, wherein, The charging rule includes at least one of the following: No charge; Different rates are applied according to the size of the AI / ML model to be transmitted; wherein the larger the AI / ML model to be transmitted, the higher the rate applied to the AI / ML model transmission; Different rates are applied according to the length of the time period of AI / ML model transmission; wherein the longer the AI / ML model transmission time, the higher the rate applied to the AI / ML model transmission; Different rates are applied according to the weighted values of the size of the AI / ML model to be transmitted, the area information of the AI / ML model to be used, and the time period of the AI / ML model to be transmitted; wherein the higher the weighted values of the size of the AI / ML model to be transmitted, the area information of the AI / ML model to be used, and the time period of the AI / ML model to be transmitted, the higher the rate applied to the AI / ML model transmission; Different rates are applied according to the service quality parameter set; wherein the higher the service quality flow bit rate of the AI / ML model transmission, the higher the rate applied to the AI / ML model transmission. Whether to charge and determine the corresponding rate according to whether there is a federated learning group identifier.

6. The method of claim 1, wherein, The parameters requested in the first message include at least one of the following: the identifier of the application using the AI / ML model, the network slice of the PDU session for transmitting the AI / ML model service quality flow, the data network name of the PDU session for transmitting the AI / ML model service quality flow, and the service quality parameter set; the analysis information includes at least one of the analysis results corresponding to the parameters requested in the first message; wherein the information processing includes session management policy decision processing; Correspondingly, the information processing of the AI / ML model transmission according to the analysis information includes: Determining the authorized service quality parameters of the AI / ML model transmission in the PCC rule according to the analysis information; Taking the authorized service quality parameters of the AI / ML model transmission in the determined PCC rule as the latest session management policy information, and sending a third message carrying the latest session management policy information to the session management function SMF, wherein the third message is used to request the SMF to update the session management policy; Receiving the second result sent by the SMF, wherein the second result is determined by the SMF according to the latest session management policy information, and the second result includes updating the session management policy or not updating the session management policy.

7. The method of claim 6, wherein, The determination of the authorized service quality parameters of the AI / ML model transmission in the PCC rule according to the analysis information includes: adjusting a priority of a 5G quality of service identifier in a granted quality of service parameter or reflective quality of service control in the granted quality of service parameter for AI / ML model transmission if a data rate of AI / ML model transmission in the analysis information is detected to be too low, an uplink direction maximum bit rate for AI / ML model transmission, a downlink direction maximum bit rate for AI / ML model transmission, an uplink direction minimum bit rate for AI / ML model transmission, and a downlink direction minimum bit rate for AI / ML model transmission.

8. An information processing method based on model-driven transport state analysis, applied to the Session Management Function (SMF), characterized in that, The method comprises: sending a first message to a network data analytics function (NWDAF), the first message being used to request subscription of analysis information of artificial intelligence / machine learning (AI / ML) model transmission status in a network; receiving the analysis information sent by the NWDAF; wherein the analysis information is used for negotiation of charging / sponsorship, PDU session charging statistics, and session management policy decision making; performing information processing on AI / ML model transmission according to the analysis information.

9. The method of claim 8, wherein, The parameters requested in the first message include parameters in a PDU session for AI / ML model transmission, wherein the PDU session for AI / ML model transmission is created by a user equipment (UE); the parameters in the PDU session include at least one of the following: UE identity or user equipment group (UE(s)) identity, network data analysis identity, identity of an application using the AI / ML model, area information using the AI / ML model, time period of AI / ML model transmission, and quality of service parameter set; and the information processing includes session management. The information processing on AI / ML model transmission according to the analysis information comprises: matching information of optimal AI / ML model transmission service experience from the analysis information according to positions of respective UEs; obtaining information of a service session anchor point (UPF) from the information of optimal AI / ML model transmission service experience, and determining the service session anchor point (UPF) as a new PDU session anchor point (UPF); providing an optimal path for AI / ML model transmission for respective UEs using the selected new PDU session anchor point (UPF).

10. The method of claim 9, wherein, The determination of the service session anchor point (UPF) as the new PDU session anchor point (UPF) comprises: determining an optimal path for AI / ML model transmission according to positions of respective UEs and information of optimal AI / ML model transmission service experience in AI / ML model transmission; determining a service session anchor point (UPF) corresponding to the optimal path as a new PDU session anchor point (UPF).

11. The method of claim 9, wherein, The method further comprises: sending a fourth message to an application function (AF) before changing the PDU session anchor point (UPF), the fourth message being used to notify user plane anchor point change and target data network access identity; receiving a confirmation result sent by the AF.

12. The method of claim 9, wherein, The information processing on AI / ML model transmission according to the analysis information further comprises: matching information of optimal AI / ML model transmission service experience from the analysis information according to positions of respective UEs; obtaining information of a service session anchor UPF and analysis results of a service quality parameter set from information of a service experience transmitted by the optimal AI / ML model; determining, according to the information of the service session anchor UPF and the analysis results of the service quality parameter set, to establish a new PDU session anchor UPF for transmitting the AI / ML model, the new PDU session anchor UPF being a second PDU session anchor UPF, wherein the first PDU session anchor UPF is a PDU session anchor currently transmitting the AI / ML model; determining to select a user plane function UPF as a branch point BP or an uplink classifier UL CL of the PDU session; providing, to the BP or the UL CL, traffic filters corresponding to the first PDU session anchor UPF and the second PDU session anchor UPF, and instructing the BP or the UL CL to forward a service quality flow / packet of uplink AI / ML model transmission to the second PDU session anchor UPF; providing, for each UE, an optimal path associated with information of a service quality flow in AI / ML model transmission by using the created second PDU session anchor UPF.

13. The method of claim 12, wherein, The determining, according to the information of the service session anchor UPF and the analysis results of the service quality parameter set, to establish a new PDU session anchor for transmitting the AI / ML model includes: determining, according to the information of the service session anchor UPF and the analysis results of the service quality parameter set, an optimal path of AI / ML model transmission; determining, according to the optimal path, to establish a new PDU session anchor UPF for transmitting the AI / ML model.

14. The method of claim 12, wherein, The method further includes: sending, to an application function AF, a fifth message for notifying a user plane anchor point change and a target data network access identifier before the new PDU session anchor UPF is created for transmitting the AI / ML model; receiving a confirmation result sent by the AF. 15.A decision and session management apparatus based on model transmission state analysis, applied to a policy control function (PCF), characterized in that, The apparatus includes a memory, a transceiver, and a processor: the memory is configured to store a computer program; the transceiver is configured to transceive data under control of the processor; and the processor is configured to read the computer program in the memory and perform the following operations: sending, to a network data analysis function NWDAF, a first message for requesting to subscribe to analysis information of a state of artificial intelligence / machine learning AI / ML model transmission in a network; receiving the analysis information sent by the NWDAF; wherein the analysis information is used for negotiating charging / sponsorship, PDU session charging statistics, and making session management policy decisions; performing information processing on the AI / ML model transmission according to the analysis information.

16. A decision and session management apparatus based on model transmission state analysis, applied to a session management function (SMF), characterized in that, The apparatus includes a memory, a transceiver, and a processor: the memory is configured to store a computer program; the transceiver is configured to transceive data under control of the processor; and the processor is configured to read the computer program in the memory and perform the following operations: sending, to a network data analysis function NWDAF, a first message for requesting to subscribe to analysis information of a state of artificial intelligence / machine learning AI / ML model transmission in a network; Receive the analysis information sent by the NWDAF; wherein the analysis information is used for negotiating charging / sponsorship, PDU session charging statistics, making session management policy decisions; According to the analysis information, the AI / ML model transmission is information processed.

17. An information processing device based on model-transfer state analysis, applied to policy control function (PCF), characterized in that, The device comprises: A sending unit configured to send a first message to a network data analysis function (NWDAF), wherein the first message is used to request to subscribe to analysis information of an artificial intelligence / machine learning (AI / ML) model transmission state in a network; A receiving unit configured to receive the analysis information sent by the NWDAF; wherein the analysis information is used for negotiating charging / sponsorship, PDU session charging statistics, making session management policy decisions; A processing unit configured to process the AI / ML model transmission according to the analysis information.

18. A decision and session management apparatus based on model transmission state analysis, applied to a session management function (SMF), comprising: The device comprises: A sending unit configured to send a first message to a network data analysis function (NWDAF), wherein the first message is used to request to subscribe to analysis information of an artificial intelligence / machine learning (AI / ML) model transmission state in a network; A receiving unit configured to receive the analysis information sent by the NWDAF; wherein the analysis information is used for negotiating charging / sponsorship, PDU session charging statistics, making session management policy decisions; A processing unit configured to process the AI / ML model transmission according to the analysis information.

19. A processor-readable storage medium, comprising: The processor readable storage medium stores a computer program, and the computer program is used to make the processor execute the method in any one of claims 1 to 14.

20. A computer program product comprising a computer program, characterized in that, The computer program is used to implement the method in any one of claims 1 to 14.

Images