Quality of service policy determination method, device, equipment, medium and program product

Abstract

The invention provides a service quality strategy determination method and device, equipment, a medium and a program product, relates to the technical field of communication, and is used for ensuring the service quality of an immersive communication service. The method comprises the following steps: acquiring network key parameters, application server information and user subscription information through an AI module in a PCF; through an AI module, based on the network key parameters, whether network congestion occurs at present is judged or whether network congestion occurs within a preset duration after the current moment is predicted; through an AI module, based on application server information and user subscription information, determining a target service flow and a target user which need to be subjected to QoS (Quality of Service) guarantee; and generating a target QoS strategy for the target service flow and the target user through the AI module under the condition that the network congestion is currently or predicted to occur, and issuing the target QoS strategy to the SMF. The method and device are applied to the process of guaranteeing the QoS of the communication service.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to methods, apparatus, equipment, media, and program products for determining quality of service strategies. Background Technology

[0002] With the development of fifth-generation mobile communication enhancement technologies and future sixth-generation mobile communication systems, immersive communication services, such as extended reality (XR) and holographic communication, are constantly emerging. These new modal media have dynamic characteristics such as high diversity, heterogeneous data stream fusion, and bursty traffic, which place more stringent requirements on the network's Quality of Service (QoS).

[0003] Currently, the immersive communication scenarios and performance requirements of 6G mobile communications are still in the standardization process, and a complete quality of service (QoS) assurance mechanism has not yet been established. Because these services are extremely sensitive to latency and bandwidth, user experience will significantly deteriorate once network congestion occurs. Therefore, their QoS requirements for the mobile core network are far higher than those for services in 5G mobile communication systems. However, there is still a lack of effective control mechanisms to maintain QoS for the high-dynamic load scenarios of 6G mobile immersive services.

[0004] Therefore, ensuring the quality of service for services (such as immersive communication services) under high dynamic load environments is a problem that needs to be solved. Summary of the Invention

[0005] This application provides a method, apparatus, device, medium, and program product for determining service quality strategies to ensure the service quality of communication services.

[0006] In a first aspect, this application provides a method for determining a quality of service (QoS) policy, comprising: acquiring key network parameters, application server information, and user subscription information through the Artificial Intelligence (AI) module in the Policy Control Function (PCF); determining, based on the key network parameters, whether network congestion is currently occurring or whether network congestion is predicted to occur within a preset time period after the current moment through the AI ​​module; determining, based on the application server information and user subscription information, the target service flow and target users that require QoS assurance through the AI ​​module; and generating a target QoS policy for the target service flow and target users in the event of current or predicted network congestion, and distributing the target QoS policy to the Session Management Function (SMF).

[0007] The technical solution provided in this application brings at least the following beneficial effects: Deploying an AI module in the PCF (Power Processing Function) allows for the acquisition and comprehensive analysis of key network parameters from network data analysis functions, application server information from network open functions, and user subscription information from unified data storage. This enables multidimensional and refined decision-making. Based on this real-time and comprehensive data, the AI ​​module can not only determine whether network congestion is currently occurring, but also predict whether network congestion will occur within a preset timeframe. When congestion is identified or predicted, the AI ​​module further accurately identifies the target business flows (such as XR and holographic communication flows) and target users (such as high-value users) that require protection based on application server information and user subscription information, ensuring the precise targeting of policy adjustments. Finally, the AI ​​module generates and distributes dynamic measurement policies to the session management function for the target business flows and target users, achieving dynamic and automated protection of critical business resources. In this way, service quality assurance is transformed into a new dynamic, accurate and predictive model, which solves the problems of delayed response and rigid strategies of traditional solutions in high dynamic load environments, improves the ability to stably guarantee the service quality of critical services such as immersive communication, and ensures the service quality of communication services.

[0008] One possible implementation method for obtaining key network parameters, application server information, and user subscription information includes: obtaining key network parameters from the Network Data Analytics Function (NWDAF); obtaining application server information from the Network Exposure Function (NEF); and obtaining user subscription information from the Unified Data Repository (UDR).

[0009] Another possible implementation, based on key network parameters, determines whether network congestion is currently occurring or predicts whether network congestion will occur within a preset time period after the current moment. This includes: if the key network parameters meet the congestion judgment conditions, then determining that network congestion is currently occurring; or, based on historical key network parameter data, predicting key network parameters within a preset time period, and if the predicted key network parameters meet the congestion judgment conditions, then predicting that network congestion will occur within the preset time period; wherein the key network parameters include at least one of the following: User Plane Function (UPF) load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency.

[0010] Another possible implementation involves determining the target service flow and target user that require QoS assurance based on application server information and user subscription information. This includes: matching the service flow in the network with the application server information to determine the service flow belonging to the target communication service type, which is then used as the target service flow; and determining at least one of the users who have subscribed to the target communication service and users belonging to the high-priority subscription level from the user subscription information, which is then used as the target user.

[0011] Another possible implementation is that the target QoS policy includes at least one of the following: increasing the Allocation and Retention Priority (ARP) of the target user; increasing the Session-Aggregate Maximum Bit Rate (Session-AMBR) or User Equipment-AMBR (UE-AMBR) related to the target service flow; updating the packet detection rules of the uplink (UL) or downlink (DL) to reserve bandwidth resources for the target service flow; and routing the new sessions of the target service flow to the UPF with a load less than a preset load threshold.

[0012] Another possible implementation method described above includes: in the event that the generation or distribution of the target QoS policy fails or times out, executing the default QoS policy based on the user's subscription information.

[0013] Secondly, this application provides a service quality policy determination apparatus, comprising: an acquisition module, a determination module, a processing module, and a sending module; the acquisition module is used to acquire key network parameters, application server information, and user subscription information through the AI ​​module in the PCF; the determination module is used to determine, based on the key network parameters, whether network congestion is currently occurring or whether network congestion is predicted to occur within a preset time period after the current moment, based on the application server information and user subscription information, based on the AI ​​module; the determination module is also used to determine the target service flow and target user that need QoS protection, based on the application server information and user subscription information, based on the AI ​​module; the processing module is used to generate a target QoS policy for the target service flow and target user, based on the AI ​​module, in the case of current or predicted network congestion; and the sending module is used to send the target QoS policy to the SMF.

[0014] One possible implementation is that the aforementioned acquisition module is specifically used to: acquire key network parameters from NWDAF; acquire application server information from NEF; and acquire user subscription information from UDR.

[0015] Another possible implementation is that the aforementioned determining module is specifically used to: determine that network congestion has occurred if the network key parameters meet the congestion judgment conditions; or, based on historical network key parameter data, predict the network key parameters within a preset time period, and if the predicted network key parameters meet the congestion judgment conditions, predict that network congestion will occur within the preset time period; wherein the aforementioned network key parameters include at least one of the following: UPF load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency.

[0016] Another possible implementation is that the aforementioned determining module is specifically used to: match the service flows in the network with the application server information to determine the service flows that belong to the target communication service type as the target service flows; and determine, from the user subscription information, at least one of the users who have subscribed to the target communication service and the users who belong to the high priority subscription level as the target users.

[0017] Another possible implementation is that the target QoS policy includes at least one of the following: increasing the ARP of the target user; adding a Session-AMBR or UE-AMBR related to the target service flow; updating the packet detection rules of the UL or DL ​​to reserve bandwidth resources for the target service flow; and routing the new session of the target service flow to the UPF with a load less than a preset load threshold.

[0018] In another possible implementation, the aforementioned processing module is also used to execute a default QoS policy based on the user's subscription information in the event that the generation or distribution of the target QoS policy fails or times out.

[0019] Thirdly, this application provides an electronic device comprising: a processor and a memory; the memory storing processor-executable instructions; when the processor is configured to execute the instructions, causing the electronic device to implement the method of the first aspect described above.

[0020] Fourthly, this application provides a computer-readable storage medium comprising: computer software instructions; which, when executed in an electronic device, cause the electronic device to implement the method described in the first aspect.

[0021] Fifthly, this application provides a computer program product comprising a computer program; when the computer program is run in an electronic device, the electronic device performs the method described in the first aspect.

[0022] The beneficial effects of the second to fifth aspects mentioned above are described in the corresponding description of the first aspect and will not be repeated here. Attached Figure Description

[0023] Figure 1This is a schematic diagram illustrating the application environment of a service quality strategy determination method provided in an embodiment of this application. Figure 2 A flowchart illustrating a service quality strategy determination method provided in an embodiment of this application; Figure 3 A flowchart illustrating another service quality strategy determination method provided in an embodiment of this application; Figure 4 A schematic diagram illustrating the overall implementation flow of a service quality strategy determination method provided in an embodiment of this application; Figure 5 This application provides an architectural diagram of a service quality strategy determination system. Figure 6 A schematic diagram of the composition of a service quality strategy determination device provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] The following is a detailed description, with reference to the accompanying drawings, of the service quality strategy determination method, apparatus, equipment, media, and program products provided in this application.

[0025] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0026] The terms "first" and "second," etc., used in the specification and drawings of this application are used to distinguish different objects or to distinguish different treatments of the same object, rather than to describe a specific order of objects.

[0027] Furthermore, the terms "comprising" and "having," and any variations thereof, used in the description of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0028] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0029] To facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish the same or similar items with essentially the same function and effect. Those skilled in the art can understand that the terms "first" and "second" are not intended to limit the quantity or execution order.

[0030] In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0031] This application provides a service quality policy determination method that can be applied to the core network systems of 5G-Advanced / 5G-A and 6G technologies. It is used to achieve dynamic QoS assurance for immersive communication services such as XR and holographic communication. This solution, through an AI module deployed in the PCF, collaborates with NWDAF, NEF, and UDR to acquire multi-dimensional network status, service characteristics, and user information in real time. This enables accurate judgment and prediction of network congestion status, and dynamically generates and distributes optimal QoS policies for identified high-value service flows and users. These policies are executed via SMF and UPF through standard interfaces, thereby providing intelligent, adaptive, and highly reliable end-to-end service quality assurance for critical services such as immersive communication in a 6G heterogeneous and highly dynamic network environment.

[0032] Currently, 5G-Advanced (5G-A), including audio, video, and real-time captioning, possesses highly mature media features and diverse QoS guarantee mechanisms. New modal media exhibit high diversity (e.g., multimodal fusion, heterogeneous data streams) and variability (e.g., cross-modal collaboration, burst traffic characteristics). These new modal media include tokens, tensors, intents, semantics, inference, AI models, and model training data processed by AI / machine learning (ML). Research on their streaming characteristics is relatively limited, and continuous research on typical QoS guarantee technologies is needed.

[0033] In today's rapidly evolving communication technology environment, the quality of service (QoS) control requirements for immersive communication services using 6G (6th Generation Mobile Communication Technology) and new media have far exceeded the scope of XR or single-media services. Emerging 6G services, including artificial intelligence communication and computing, which provide or enhance immersive experiences and their unique attributes, also require precise QoS management to achieve optimal performance.

[0034] Current communication standards organizations are still aligning with key aspects of 6G immersive communication scenarios and requirements, and there are no standardized solutions yet.

[0035] 6G immersive communication has extremely high requirements for latency and bandwidth, such as Virtual Reality (VR), Augmented Reality (AR), or holographic calls. Once the network becomes congested, the user experience will deteriorate sharply. Therefore, 6G immersive communication places higher demands on the QoS of the mobile core network than 5G VoNR services. Currently, 6G immersive communication is still in the early stages of research, and there are no specific and clearly feasible methods to improve 6G network QoS guarantees.

[0036] Therefore, ensuring the quality of service for services (such as immersive communication services) under high dynamic load environments is a problem that needs to be solved.

[0037] To address the aforementioned technical issues, this application provides a method, apparatus, device, medium, and program product for determining Quality of Service (QoS) policies. By introducing and deploying an AI module within the Policy Control Function (PCF), the AI ​​module proactively acquires real-time key network parameters (such as bandwidth utilization and latency) from the Network Data Analysis Function (NWDAF), achieving accurate and real-time identification of network congestion. Furthermore, it utilizes predictive models to forecast congestion trends over a preset time period. This enables the system to shift from passive, reactive safeguards to proactive, predictive interventions, improving the timeliness and foresight of QoS guarantees. Secondly, by integrating application server information obtained from the Network Open Function (NEF) and user subscription information obtained from the Unified Data Storage (UDR), the AI ​​module can accurately identify specific service flows requiring priority (such as immersive communication services) and high-value users, thereby achieving precise and differentiated policy control and preventing disordered allocation of network resources during congestion. Ultimately, when congestion is detected or is about to occur, the AI ​​module can automatically generate and distribute dynamic QoS policies (such as adjusting priorities and guaranteeing bandwidth) for the aforementioned target services and users. These policies are executed through the Session Management Function (SMF) configured with the User Plane Function (UPF), forming an automated closed loop from perception and decision-making to execution. This provides continuous, stable, and high-quality service assurance for the immersive services of core users in complex network environments, such as during periods of high load.

[0038] Thus, this solution provides a technology for dynamically adjusting user QoS policies when an AI module identifies high network traffic. By adding an AI module to the PCF network element of the mobile core network, this module can predict or instantly identify network congestion by analyzing network data in real time, and automatically and dynamically generate and distribute precise QoS policies for immersive communication services (such as XR, holographic communication, etc.), thereby prioritizing the experience of key users when network resources are scarce.

[0039] The embodiments provided in this application will now be described in detail with reference to the accompanying drawings.

[0040] The service quality strategy determination method provided in this application embodiment can be applied to, for example, Figure 1 The application environment shown. For example... Figure 1 As shown, the application environment includes a service quality policy determination device 101 and a front-end device 102. The service quality policy determination device 101 and the front-end device 102 are interconnected.

[0041] In some embodiments, the service quality policy determination device 101 may be a server cluster consisting of multiple servers, a single server, a computer, or a processor or processing chip in a server or computer, etc. This application embodiment does not limit the specific device form of the service quality policy determination device 101. Figure 1 The service quality policy determination device 101 is illustrated using a single server as an example.

[0042] In some embodiments, the front-end device 102 can be a device with wireless transceiver capabilities, such as a mobile phone, tablet computer, wearable device, in-vehicle device, augmented reality (AR) / virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc. This application embodiment does not limit the specific device form of the front-end device 102. Figure 1 The example shown is a mobile phone, with the front-end device 102 as the illustration.

[0043] In some embodiments, the quality of service policy determination device 101 may be a PCF.

[0044] In some embodiments, the Quality of Service (QoS) policy determination device 101 generates a dynamic QoS policy and distributes it to the Service Quality Filter (SMF) via a standard N7 interface. Thus, when the front-end device 102 initiates immersive services such as XR or holographic communication, and its user plane data stream is forwarded via the UPF, the UPF will prioritize and guarantee resources for the front-end device 102's service flow based on the dynamic QoS policy configured by the SMF and derived from the QoS policy determination device 101. This ensures that users of the front-end device 102 can continuously enjoy a low-latency, high-smoothness immersive communication experience even under high-load network conditions.

[0045] It should be noted that the system architecture described in the embodiments of this application is for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and does not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of system architecture, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0046] See Figure 2 This is a flowchart illustrating a service quality strategy determination method provided in an embodiment of this application. Figure 2 As shown, the service quality strategy determination method provided in this application can be implemented by the aforementioned service quality strategy determination device, specifically including the following steps 201 to 204.

[0047] Step 201: The service quality strategy determination device obtains key network parameters, application server information, and user subscription information through the AI ​​module in the PCF.

[0048] In some embodiments, the aforementioned quality of service policy determination device may be a PCF / PCF network element / PCF entity in the core network, or a specific hardware module or software function entity in the PCF network element.

[0049] In some embodiments, the "obtaining key network parameters, application server information and user subscription information" in step 201 above can be specifically implemented as steps 201a to 201c.

[0050] Step 201a: The Quality of Service (QoS) policy determination device obtains key network parameters from the NWDAF.

[0051] In some embodiments, the above-mentioned key network parameters may include at least one of the following: UPF load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency.

[0052] In some embodiments, the target interface described above may be an N3 interface or an N6 interface.

[0053] It is understandable that key network parameters are a set of quantitative indicators reflecting the real-time performance status of the network, and the quality of service policy determination device can subscribe to key parameters within the network from the NWDAF network element: Core network load: Subscribe to metrics for specific network slices, Data Network Access Information (DNAI), or Data Network Name (DNN) from NWDAF network elements through the Nnwdaf_EventsSubscription service.

[0054] Key metrics include: UPF load factor, N3 / N6 interface bandwidth utilization, packet drop rate, average / percentile latency, etc.

[0055] Threshold trigger: Set a threshold (such as bandwidth utilization) to 70%. When bandwidth utilization is greater than 70%, NWDAF can actively notify the PCF's AI module when the threshold is exceeded.

[0056] It should be noted that the AI ​​module in PCF can subscribe to events from NWDAF network elements through the Nnwdaf_EventsSubscription service. This subscription can be targeted at specific network slices, data network access locations, or data network names. As a centralized analysis node for network data, NWDAF continuously collects and processes raw data from network elements such as user plane functions and base stations.

[0057] In some embodiments, the above-mentioned UPF load rate refers to the percentage of processing resources occupied by user plane function network elements. If it is too high, it may become a bottleneck for data forwarding.

[0058] In some embodiments, the target interface bandwidth utilization rate refers to the ratio of the current traffic to the maximum available bandwidth of the N3 (between the radio access network and the core network) or N6 (between the core network and the data network) interface, which is a direct indicator for judging network congestion.

[0059] In some embodiments, the packet drop rate refers to the proportion of packets that cannot be forwarded due to congestion in the interface or UPF buffer and are ultimately dropped.

[0060] In some embodiments, the above-mentioned average latency or percentile latency refers to the average time required for a data packet to traverse the network or the high percentile (e.g., 95%) latency, the latter of which is crucial for ensuring an immersive service experience.

[0061] In this way, the service quality policy determination device adopts a threshold triggering mechanism, which presets a threshold (e.g., 70%) for key parameters (such as N3 interface bandwidth utilization). NWDAF continuously monitors these parameters, and once the monitored value exceeds the threshold, it will proactively and immediately send a notification to the AI ​​module of PCF, thereby triggering the subsequent intelligent decision-making process and avoiding the overhead caused by continuous polling of PCF.

[0062] Step 201b: The service quality policy determination device obtains application server information from NEF.

[0063] In some embodiments, the NEF network element acts as a gateway for exposing network capabilities to the outside world, storing and managing the registration information of third-party application servers. The AI ​​module in the PCF can query the NEF for information on registered application servers providing immersive communication services by calling the Nnef_ParameterProvision service.

[0064] In some embodiments, the QoS policy determination device can obtain application server information / application information from NEF network elements: through the Nnef_ParameterProvision service, it queries NEF for registered immersive application server information (such as IP 5-tuple, required basic QoS requirements) to accurately identify immersive communication service flows.

[0065] It is understandable that the core function of application server information is to achieve business flow identification, and its key fields include: The IP 5-tuple consists of the source IP address, destination IP address, source port number, destination port number, and transport layer protocol. This is the basis for uniquely identifying a service session in network traffic.

[0066] Required basic QoS requirements: The basic quality of service requirements of the network for its service flows declared by the application server during registration, such as the required 5G QoS Identifier, guaranteed bit rate, etc.

[0067] In this way, by acquiring this information, the AI ​​module can accurately match anonymous IP data streams on the network with known immersive applications (such as specific XR cloud rendering servers and holographic call servers).

[0068] Step 201c: The service quality strategy determination device obtains user subscription information from the UDR.

[0069] In some embodiments, the UDR is a central database that stores user subscription data. The AI ​​module in PCF queries the UDR through a standard interface to obtain the user's subscription information.

[0070] In some embodiments, the quality of service policy determination device can obtain user subscription information from the UDR network element: query user subscription information (such as ARP) and identify whitelisted users / high-value users / user groups that have subscribed to specific immersive communication services.

[0071] It is understandable that user contract information is used to identify user identity and value, and its key data includes: Allocation and Retention Priority: A numerical value that determines the priority of user or business sessions when resources are scarce.

[0072] Service subscription list: Indicates whether a user has subscribed to a specific immersive communication service package.

[0073] By analyzing this information, the AI ​​module can accurately identify high-value users (such as users with high ARP values), whitelisted users, or user groups that have subscribed to specific immersive services.

[0074] In this way, by obtaining key network parameters, application server information, and user subscription information from the three network elements NWDAF, NEF, and UDR respectively, multi-dimensional data is obtained, ensuring the validity, standardization, and systematic nature of the data source. This allows the AI ​​module's decisions to be based on comprehensive and accurate input data, enabling precise correlation and cross-analysis of network status, business attributes, and user identity. This facilitates the subsequent implementation of refined and personalized service quality strategies for specific businesses and users.

[0075] Step 202: The service quality strategy determination device uses the AI ​​module to determine whether network congestion is currently occurring or predicts whether network congestion will occur within a preset time period after the current moment, based on key network parameters.

[0076] In some embodiments, step 202 above can be specifically implemented as step 202a or step 202b.

[0077] Step 202a: The service quality strategy determination device determines network congestion if the key network parameters meet the congestion judgment conditions through the AI ​​module.

[0078] In some embodiments, the congestion judgment condition is that the network key parameters exceed their respective trigger thresholds. That is, the congestion judgment condition is a set of preset logical rules that compare the real-time acquired network key parameters with their respective trigger thresholds. For example, the UPF load rate is greater than / exceeds the threshold corresponding to the UPF load rate, or the average latency is greater than / exceeds the threshold corresponding to the average latency. This multi-condition judgment makes the congestion identification results more comprehensive and reliable.

[0079] For example, if the UPF load rate is >80%, or the N3 interface bandwidth utilization is >70%, or the average latency is >20ms, then network congestion is considered to be currently occurring.

[0080] Step 202b: The service quality strategy determination device uses the AI ​​module to predict network key parameters within a preset time period based on historical network key parameter data. If the predicted network key parameters meet the congestion judgment conditions, then network congestion is predicted to occur within the preset time period.

[0081] It should be noted that the key network parameters and congestion judgment conditions mentioned here can be found in the descriptions in the above embodiments, and will not be repeated here.

[0082] In some embodiments, the service quality strategy determination device employs a time series prediction model, i.e., the AI ​​module uses a time series prediction model to analyze historical network key parameter data to predict network key parameters within a preset time period. This time series prediction model can be a Long Short-Term Memory (LSTM) network model or a Transformer model. These models can capture the periodic and trend-based changes in indicators such as traffic and load.

[0083] In some embodiments, in the predictive identification of the congestion identification and prediction step, the quality of service policy determination device may use an LSTM time series prediction model, but is not limited to LSTM as a time series prediction model. Other time series prediction models (such as Transformer) can also be used to achieve predictive identification.

[0084] In some embodiments, the LSTM model, as a special type of recurrent neural network, has an internal gating mechanism that can effectively learn long-term dependencies, making it very suitable for processing time-dependent sequential data such as network traffic.

[0085] In some embodiments, the Transformer model is based on a self-attention mechanism, which can process sequential data in parallel and capture global dependencies, giving it advantages in prediction accuracy and efficiency.

[0086] In some embodiments, the quality of service policy determination device can perform congestion identification and prediction via an AI module: Real-time identification: The AI ​​module immediately determines whether network congestion has occurred based on the key parameters currently reported by NWDAF and in combination with predefined threshold rules.

[0087] Predictive identification: Historical traffic data is analyzed using an LSTM time series prediction model to predict potential congestion within a short period (e.g., the next 500ms). This step allows the AI ​​module to make preventative strategy adjustments.

[0088] It is understandable that the AI ​​module uses a time series prediction model to predict the values ​​of key network parameters within a preset time period (e.g., 500 milliseconds), and then substitutes the predicted values ​​into the congestion judgment conditions mentioned above for evaluation. If the prediction results meet the conditions, it is determined that the network will experience congestion in the near future, and preventive strategy adjustments can be made.

[0089] Thus, by setting congestion judgment conditions and combining them with real-time comparisons, the accuracy and timeliness of network congestion identification are improved. This not only enables rapid responses to existing congestion but also allows for the prediction of potential congestion risks in the near future based on historical data. By comprehensively considering multiple key parameters such as load rate, bandwidth utilization, and latency, the congestion identification results are more comprehensive and reliable.

[0090] Step 203: The service quality strategy determination device uses the AI ​​module to determine the target service flows and target users that need QoS protection based on application server information and user subscription information.

[0091] In some embodiments, combined with Figure 2 ,like Figure 3 As shown, step 203 above can be specifically implemented as steps 203a and 203b.

[0092] Step 203a: The service quality policy determination device uses the AI ​​module to match the service flows in the network with the application server information, and determines the service flows that belong to the target communication service type as the target service flows.

[0093] In some embodiments, the Quality of Service (QoS) policy determination device performs service flow identification to obtain target service flows. This involves matching service flows in the network with application server IPs obtained from the NEF (Network Application Function), accurately identifying affected immersive service flows. It can be understood that the AI ​​module performs service flow identification by matching the IP 5-tuple information of IP data flows currently being transmitted in the network (typically reported by SMF / UPF or obtained through deep packet inspection) with application server information (specifically, a list of IP addresses) obtained from the NEF. Successfully matched service flows are identified as target service flows belonging to the target communication service type (such as XR or holographic communication).

[0094] Step 203b: The service quality strategy determination device uses the AI ​​module to determine, from the user subscription information, at least one of the following: users who have signed up for the target communication service and users who belong to the high-priority subscription level, as target users.

[0095] In some embodiments, the AI ​​module identifies the user who initiates the target service flow by examining their user subscription information, and selects at least one of the following: users who have subscribed to the target communication service and users who belong to the high-priority subscription level (users with high ARP values) as the target user.

[0096] In this way, by matching network traffic with application server information to identify target business types and filtering high-value users from user subscription information, precise positioning of service quality assurance targets is achieved. This allows for accurate priority allocation of network resources to specific businesses (such as immersive communication) and specific user groups that truly require protection, thus avoiding the experience degradation caused by disordered resource allocation strategies during congestion. This dual identification mechanism based on business identity and user value ensures the precision and differentiation of service quality control, improving the efficiency of resource allocation.

[0097] Step 204: When network congestion occurs currently or is predicted, the service quality policy determination device generates target QoS policies for target service flows and target users through the AI ​​module, and then sends the target QoS policies to the SMF.

[0098] In some embodiments, the target QoS policy described above includes at least one of the following: Increase the ARP of the target user; Add a Session-AMBR or UE-AMBR related to the target service flow; Update the packet detection rules of UL or DL ​​to reserve bandwidth resources for the target service flow; New sessions of the target business flow are routed to UPFs with loads less than a preset load threshold.

[0099] It is understood that the Quality of Service (QoS) policy determination device can ensure the basic QoS corresponding to the 5QI (5G QoS Identifier), such as Guaranteed Bit Rate (GBR), for all identified immersive service users. When the judgment or prediction result of step 202 indicates network congestion, the AI ​​module can dynamically generate target QoS policies. These policies conform to communication standard policies and charging control rules, including one or more of the following policies: (1) Priority enhancement: Temporarily adjust the ARP priority of specific whitelist users / high-value users / specific immersive communication service users to the highest level to ensure that their resource requests are met first. That is, in the competition for network resources, their service requests are prioritized and processed by network elements (such as base stations and UPFs).

[0100] (2) Bandwidth guarantee: Dynamically increase the upper limit of Session-AMBR (Session AggregateMaximum Bit Rate) or UE-AMBR for immersive services, that is, dynamically increase the upper limit of the aggregated maximum bit rate for target users or target service sessions, and provide them with higher available bandwidth guarantee.

[0101] (3) Resource reservation: By interacting with SMF / UPF, the update of UL / DL Packet Detection rules is triggered to reserve some bandwidth resources for these service flows. That is, the SMF sends instructions to UPF to update the packet detection rules and reserve resources in the queue of UPF for data packets that match the characteristics of the target service flow, so as to ensure that their transmission is not affected by congestion.

[0102] (4) Traffic routing: By issuing an update command to the SMF, the newly initiated immersive session is routed to the UPF with a lighter load (based on DNAI level analysis provided by NWDAF). That is, based on the DNAI level UPF load information provided by NWDAF, the SMF is instructed to anchor the newly established service session to the UPF instance with a lighter load, thereby achieving load balancing and optimal path selection.

[0103] In some embodiments, the target QoS policy described above is used by the SMF to configure the UPF to execute the target QoS policy. That is, after receiving the target QoS policy, the SMF can send the target QoS policy to the UPF to instruct the UPF to execute the target QoS policy.

[0104] It is understandable that the PCF can send updated Policy and Charging Control (PCC) rules, i.e., the target QoS policy, generated by the AI ​​module to the SMF via the N7 interface. After receiving the new policy, the SMF configures the UPF through the N4 interface to execute the new traffic redirection, priority scheduling, and bandwidth limiting rules. The UPF performs the corresponding traffic redirection, priority scheduling, and bandwidth limiting operations, thereby completing end-to-end resource assurance for the target service flow.

[0105] Thus, by providing a variety of specific policy options, including prioritizing users, increasing bandwidth guarantees, reserving resources, and intelligent routing, the standard-defined QoS parameters can be directly applied to achieve end-to-end resource guarantees for target service flows. For example, increasing allocation and reservation priority ensures that service requests from high-value users are prioritized in resource contention; dynamically increasing the maximum aggregated bit rate for sessions or user devices provides higher bandwidth limits for critical services; and routing new sessions to lightly loaded user plane functions achieves load balancing, optimizing network resource utilization overall.

[0106] In this embodiment, the AI ​​module in the PCF predicts or instantly identifies network congestion and automatically and dynamically generates and distributes precise QoS control policies. This is not only applicable to immersive communication services (such as XR, holographic communication, etc.), but also to any communication service that requires real-time dynamic adjustment of QoS.

[0107] The service quality strategy determination method provided in this application deploys an AI module in the PCF (Process Control Function). By acquiring and comprehensively analyzing key network parameters from network data analysis functions, application server information from network open functions, and user subscription information from unified data storage, the decision-making basis is multidimensional and refined. Based on this real-time and comprehensive data, the AI ​​module can not only determine whether network congestion is currently occurring, but also predict whether network congestion will occur within a preset time period. When congestion is identified or predicted, the AI ​​module further accurately determines the target business flows (such as XR and holographic communication flows) and target users (such as high-value users) that need to be protected based on application server information and user subscription information, ensuring the precise targeting of policy control. Finally, the AI ​​module generates and distributes dynamic measurement policies to the session management function for the target business flows and target users, realizing dynamic and automated protection of critical business resources. In this way, service quality assurance is transformed into a new dynamic, accurate and predictive model, which solves the problems of delayed response and rigid strategies of traditional solutions in high dynamic load environments, improves the ability to stably guarantee the service quality of critical services such as immersive communication, and ensures the service quality of communication services.

[0108] In some embodiments of this application, the service quality strategy determination method further includes the following step 301.

[0109] Step 301: If the generation or issuance of the target QoS policy fails or times out, the QoS policy determination device executes the default QoS policy based on the user's subscription information.

[0110] In some embodiments, when the AI ​​module fails or the decision times out, PCF can fall back to the static default policy based on user subscription information to ensure basic network operation and always maintain the human operation and maintenance control authority of network operation and maintenance personnel above that of the AI ​​module, so as to ensure network operation and maintenance security.

[0111] Understandably, when the AI ​​module fails to generate or distribute dynamic policies due to reasons such as malfunction, insufficient computing resources, or timeout in communication with the SMF, the PCF will automatically fall back to the static default QoS policy determined based on the user subscription information obtained from the UDR. This default policy is usually pre-configured when the service is activated, ensuring basic network operation and effectively preventing the interruption of the entire service quality assurance function due to the failure of the intelligent decision-making unit, thus providing users with a stable and reliable service baseline guarantee.

[0112] Thus, the robustness and service continuity of the communication system are enhanced through a fallback mechanism in case of failure in intelligent policy generation or distribution. When the artificial intelligence module fails to function properly due to faults, timeouts, or any other reason, it can automatically switch to a static default quality of service policy based on user subscription information. This ensures that even under abnormal circumstances, the basic quality of service control functions of the network can still be maintained, preventing the entire quality of service assurance mechanism from failing due to the failure of the intelligent decision-making unit. This provides users with stable and reliable service guarantees, demonstrating the high reliability of the system design.

[0113] In some embodiments, the service quality policy determination device can train a reinforcement learning model based on historical network data; wherein, the state of the reinforcement learning model is a set of key network parameters obtained from NWDAF, the action of the reinforcement learning model is an instruction to adjust the control parameters, and the reward of the reinforcement learning model is a comprehensive evaluation value calculated based on the degree of improvement of the network key performance indicators (KPIs) after updating the policy control parameters according to the adjustment instructions and executing the subsequent congestion judgment and policy generation process.

[0114] In some embodiments, a reinforcement learning model is deployed in the AI ​​module of the service quality policy determination device; the service quality policy determination device optimizes the control parameters used in the policy generation process through the reinforcement learning model.

[0115] In some embodiments, the control parameters include at least one of the following: the time period for congestion prediction (such as the preset duration mentioned above), the trigger threshold for congestion judgment, the decision interval, the adjustment range of ARP priority, and the adjustment range of AMBR.

[0116] It is understandable that the AI ​​module will continuously obtain key network parameter data after policy changes from NWDAF. Specifically, a reinforcement learning model can be deployed in the AI ​​module of PCF, using "network state" as the state, "executable policy actions" as actions, and "KPI improvement degree" as a reward. This allows the AI ​​module to automatically learn the optimal policy parameter adjustment scheme under different congestion scenarios, achieving complete self-optimization. The adjusted parameters include the AI ​​module's prediction time window, decision interval, decision timeout threshold, bandwidth utilization trigger threshold, ARP priority, Session-AMBR, etc. Applicable parameter templates can be predefined, allowing the AI ​​module to quickly match parameters and call templates.

[0117] The following describes the service quality strategy determination method of this application according to a specific embodiment, such as... Figure 4The diagram shown illustrates the overall implementation flow of a service quality strategy determination method provided in this application embodiment. The specific implementation flow is as follows: S1 to S7: S1. Acquiring Multidimensional Data: The AI ​​module in PCF obtains multidimensional information required for decision-making from three key data sources in the core network concurrently or sequentially through standard interfaces, including key network parameters, application server information, and application server data.

[0118] Specifically, the AI ​​module subscribes to and obtains real-time key network parameters from the Network Data Analysis Function (NWDAF) via the Nnwdaf_EventsSubscription service. These parameters include, but are not limited to, User Plane Function (UPF) load rate, N3 / N6 interface bandwidth utilization, packet drop rate, average latency, and percentile latency. NWDAF can proactively push notifications based on preset thresholds (such as bandwidth utilization > 70%).

[0119] The AI ​​module uses the Nnef_ParameterProvision service to query information about registered immersive application servers from the Network Open Function (NEF), including the server's IP 5-tuple (source / destination IP, port, protocol) and its declared basic Quality of Service (QoS) requirements.

[0120] The AI ​​module queries user subscription information from the Unified Data Storage (UDR) through the Nudr_DataRepository service. The core information includes the user's Allocation and Retention Priority (ARP) and service subscription list (such as whether XR or holographic communication services are subscribed).

[0121] S2. Network congestion judgment and prediction: The AI ​​module performs congestion analysis on key network parameters.

[0122] Specifically, real-time congestion detection: The AI ​​module compares current key network parameters (such as UPF load rate and interface utilization) with their respective preset trigger thresholds in real time. If any key parameter exceeds its threshold, network congestion is immediately determined to have occurred.

[0123] Predictive congestion assessment: The AI ​​module calls the built-in time series prediction model (such as LSTM, Transformer) to analyze historical key network parameter data and predict parameter values ​​within a preset time period (such as 500ms). If the predicted value meets the congestion assessment criteria, it is determined that network congestion will occur in the future time period.

[0124] The S3 and AI modules identify the objects that need to be protected, namely the target business flow and the target users.

[0125] Specifically, target service flow identification: The AI ​​module matches the IP quintuple of active service flows in the network (whose characteristics can be provided by SMF / UPF) with the list of application server IP addresses obtained from NEF to accurately identify target service flows belonging to the immersive communication type.

[0126] Target user identification: The AI ​​module filters out users who have signed up for target immersive services and / or whose allocation and retention priority (ARP) is high from the user subscription information provided by UDR, and identifies them as target users.

[0127] S4. Dynamic policy generation and distribution: If the congestion judgment result indicates that network congestion has occurred or is predicted to occur, the AI ​​module will start policy generation.

[0128] Specifically, the AI ​​module dynamically generates a target QoS policy for the identified target service flows and target users. The PCF then sends this target QoS policy as an updated policy and charging control rule to the SMF through the standard N7 interface.

[0129] S5. Network policy execution: SMF and UPF act as execution units to execute the target QoS policy.

[0130] Specifically, after receiving the target QoS policy from the PCF, the SMF translates it into specific configuration instructions via the N4 interface and sends them to the User Plane Function (UPF). Based on the instructions from the SMF, the UPF performs corresponding QoS guarantee actions on data packets matching the characteristics of the target service flow, such as priority scheduling, bandwidth guarantee, and traffic routing.

[0131] S6. Closed-loop optimization: The reinforcement learning model deployed in the AI ​​module treats the network state as the state, the instruction to adjust the control parameters as the action, and the degree of improvement in the network KPI after the policy is executed as the reward.

[0132] It should be noted that the model automatically optimizes the control parameters on which the strategy depends through continuous interaction, such as the prediction time window, congestion judgment threshold, decision interval, and ARP / AMBR adjustment magnitude.

[0133] S7. Fault rollback: Continuously monitor the generation and distribution of dynamic policies. If the generation or distribution of the target QoS policy fails or times out, the PCF will immediately roll back to the execution of the static default QoS policy determined based on the user subscription information obtained from the UDR.

[0134] It is understandable that the dynamic policy generation and distribution of S4 are continuously monitored to determine whether the generated or distributed dynamic QoS policy has failed or timed out.

[0135] This ensures that the most basic service quality control functions are not interrupted.

[0136] It should be noted that the descriptions and beneficial effects of each step S1 to S7 in this embodiment can be found in the descriptions in the above embodiments, and will not be repeated here.

[0137] Figure 5 This is a schematic diagram of the architecture of a service quality policy determination system provided in an embodiment of this application. The service quality policy determination system 800 may include: PCF 801, NWDAF 802, NEF 803, UDR 804, SMF 805, and UPF 806.

[0138] PCF 801 connects to NWDAF 802 via a service-oriented interface. PCF 801 subscribes to network data analysis services from NWDAF through the Nnwdaf_EventsSubscription service, while NWDAF 802 provides PCF 801 with key network parameters and predictive analysis results.

[0139] PCF 801 connects to NEF 803 via a service-oriented interface. PCF 801 obtains application server information from NEF 803 through the Nnef_ParameterProvision service, while NEF 803 provides PCF 801 with the registration information and business characteristics of the third-party application server.

[0140] PCF 801 connects to UDR 804 via a service-oriented interface. PCF 801 queries user data from UDR 804 through the Nudr_DataRepository service, while UDR 804 provides PCF 801 with user subscription information and policy data.

[0141] The PCF 801 connects to the SMF 805 via an N7 interface. The PCF 801 sends PCC rules to the SMF 805 through the N7 interface, and sends target QoS policies and charging control rules to the SMF 805.

[0142] The SMF 805 connects to the UPF 806 via an N4 interface. The SMF 805 configures and manages the UPF 806 through the N4 interface, issuing packet processing rules, QoS enforcement policies, and traffic redirection commands to the UPF 806.

[0143] The NWDAF 802 connects to the SMF 805 / UPF 806 via a service-oriented interface. The NWDAF 802 collects network status data and performance metrics from the SMF 805 and UPF 806, while the SMF 805 and UPF 806 provide user plane status and session management information to the NWDAF 802.

[0144] The NEF 803 connects to the UDR 804 via a service-oriented interface. The NEF 803 may access certain policies and user data through the UDR 804.

[0145] The PCF 801 serves as the system's intelligent decision-making center, internally deploying an AI module. This AI module is the carrier for executing core algorithms and can be a software module, a hardware acceleration unit, or a combination of both. NWDAF 802, NEF 803, and UDR804 act as the system's data sources, providing multi-dimensional input data for the PCF 801's AI decisions. SMF 805 and UPF 806, as policy execution units, are responsible for translating the PCF 801's intelligent decisions into specific forwarding behaviors within the network.

[0146] The PCF 801 contains an AI module. This AI module is used to: determine whether network congestion is currently occurring or predict whether it will occur in the future based on key network parameters obtained from the NWDAF 802; identify target service flows and target users requiring QoS assurance based on application server information obtained from the NEF 803 and user subscription information obtained from the UDR 804; and generate target QoS policies for target service flows and target users in the event of current or predicted network congestion, and distribute these policies to the SMF 805. This applies to steps 201, 202, 203, and 204 and their related implementations, including all operations performed by the PCF's AI module in steps 201a to 201c, 202a to 202b, and 203a to 203b.

[0147] NWDAF 802 is used to provide key network parameters to the AI ​​module of PCF 801. It is applied to step 201a and its related implementation schemes, specifically including subscribing to and obtaining parameters such as UPF load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency through network data analysis services, and supporting threshold triggering mechanisms.

[0148] NEF 803 is used to provide application server information to the AI ​​module of PCF 801. Applied to step 201b above and its related implementations, this specifically includes querying registered immersive application server information, including the IP 5-tuple and required basic QoS requirements, through the Network Capability Open Service.

[0149] UDR 804 is used to provide user subscription information to the AI ​​module of PCF 801. Applied to step 201c and its related implementations, it specifically includes querying user allocation and retention priority (ARP), service subscription lists, and other information to identify high-value users and target user groups.

[0150] SMF 805 is used to receive the target QoS policy from PCF 801 and configure UPF 806 to execute the policy. Applied to step 204 and its related implementation schemes, this specifically includes receiving the policy and charging control rules through the N7 interface and sending the corresponding configuration instructions to the UPF through the N4 interface.

[0151] UPF 806 is used to perform corresponding QoS guarantee actions on the target service flow based on the SMF 805 configuration. Applied to step 204 above and its related implementation scheme, it specifically includes operations such as priority scheduling, bandwidth guarantee, resource reservation, and traffic routing to achieve end-to-end quality of service guarantee for the target service flow.

[0152] It should be noted that the specific steps performed for each module / network element and their beneficial effects can be found in the descriptions in the above embodiments, and will not be repeated here.

[0153] As can be seen, the above mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, the embodiments of this application provide corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the modules and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0154] This application embodiment can divide the service quality policy determination device into functional modules according to the above method example. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. Optionally, the module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0155] In some embodiments, this application also provides a service quality policy determination apparatus. This service quality policy determination apparatus may include one or more functional modules for implementing the service quality policy determination method of the above method embodiments.

[0156] For example, Figure 6 This is a schematic diagram illustrating the composition of a service quality strategy determination device provided in an embodiment of this application. Figure 6 As shown, the service quality policy determination device 900 includes: an acquisition module 901, a determination module 902, a processing module 903, and a sending module 904.

[0157] The acquisition module 901 is used to acquire key network parameters, application server information, and user subscription information through the AI ​​module in the PCF; the determination module 902 is used to determine, based on the key network parameters, whether network congestion is currently occurring or whether network congestion is predicted to occur within a preset time after the current moment, based on the application server information and user subscription information, based on the AI ​​module; the determination module 902 is also used to determine the target service flow and target user that need QoS protection based on the application server information and user subscription information, based on the AI ​​module; the processing module 903 is used to generate target QoS policies for the target service flow and target user in the case of current or predicted network congestion, based on the AI ​​module; and the sending module 904 is used to send the target QoS policies to the SMF.

[0158] The service quality policy determination device provided in this application deploys an AI module in the PCF (Process Control Function). By acquiring and comprehensively analyzing key network parameters from network data analysis functions, application server information from network open functions, and user subscription information from unified data storage, it achieves multidimensional and refined decision-making basis. Based on this real-time and comprehensive data, the AI ​​module can not only determine whether network congestion is currently occurring, but also predict whether network congestion will occur within a preset time period. When congestion is identified or predicted, the AI ​​module further accurately determines the target service flows (such as XR and holographic communication flows) and target users (such as high-value users) that need to be protected based on application server information and user subscription information, ensuring the precise targeting of policy control. Finally, the AI ​​module generates and distributes dynamic measurement policies to the session management function for the target service flows and target users, realizing dynamic and automated protection of critical service resources. In this way, service quality assurance is transformed into a new dynamic, accurate and predictive model, which solves the problems of delayed response and rigid strategies of traditional solutions in high dynamic load environments, improves the ability to stably guarantee the service quality of critical services such as immersive communication, and ensures the service quality of communication services.

[0159] In some embodiments, the acquisition module 901 is specifically used to: acquire network key parameters from NWDAF; acquire application server information from NEF; and acquire user subscription information from UDR.

[0160] In other embodiments, the determination module 902 is specifically used to: determine that network congestion has occurred if the network key parameters meet the congestion judgment conditions; or, predict the network key parameters within a preset time period based on historical network key parameter data, and predict that network congestion will occur within the preset time period if the predicted network key parameters meet the congestion judgment conditions; wherein the network key parameters include at least one of the following: UPF load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency.

[0161] In some other embodiments, the determination module 902 is specifically used to: match the service flow in the network with the application server information to determine the service flow belonging to the target communication service type as the target service flow; and determine, from the user subscription information, at least one of the users who have subscribed to the target communication service and the users who belong to the high priority subscription level as the target users.

[0162] In some other embodiments, the target QoS policy includes at least one of the following: increasing the ARP of the target user; adding a Session-AMBR or UE-AMBR related to the target service flow; updating the packet detection rules of the UL or DL ​​to reserve bandwidth resources for the target service flow; and routing the new session of the target service flow to a UPF with a load less than a preset load threshold.

[0163] In some other embodiments, the processing module 903 is further configured to execute a default QoS policy based on user subscription information in the event that the generation or issuance of the target QoS policy fails or times out.

[0164] It should be noted that the service quality strategy determination device can implement all the processes implemented in the above method embodiments and achieve the same beneficial effects. To avoid repetition, it will not be described again here.

[0165] In the case where the functions of the integrated modules described above are implemented in hardware, this application provides a possible structural schematic diagram of the electronic device involved in the above embodiments. For example... Figure 7 As shown, the electronic device 90 includes: a processor 92, a communication interface 93, and a bus 94. Optionally, the electronic device 90 may also include a memory 91.

[0166] Processor 92 may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 92 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 92 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0167] Communication interface 93 is used to connect with other devices via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0168] The memory 91 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0169] As one possible implementation, the memory 91 can exist independently of the processor 92. The memory 91 can be connected to the processor 92 via a bus 94 and is used to store instructions or program code. When the processor 92 calls and executes the instructions or program code stored in the memory 91, it can implement the quality of service strategy determination method provided in the embodiments of this application.

[0170] In another possible implementation, memory 91 can also be integrated with processor 92.

[0171] Bus 94 can be an Extended Industry Standard Architecture (EISA) bus, etc. Bus 94 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0172] Through the above description of the implementation methods, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the service calling device can be divided into different functional modules to complete all or part of the functions described above.

[0173] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by computer instructions instructing related hardware. The program can be stored in the aforementioned computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be any of the foregoing embodiments or memory. The aforementioned computer-readable storage medium can also be an external storage device of the aforementioned service invocation device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the aforementioned service invocation device. Further, the aforementioned computer-readable storage medium can include both internal storage units of the aforementioned service invocation device and external storage devices. The aforementioned computer-readable storage medium is used to store the aforementioned computer program and other programs and data required by the aforementioned service invocation device. The aforementioned computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0174] This application also provides a computer program product comprising a computer program that, when run on a computer, causes the computer to execute any of the service quality policy determination methods provided in the above embodiments.

[0175] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for determining service quality strategy, characterized in that, include: The AI ​​module in the Policy Control Function (PCF) is used to obtain key network parameters, application server information, and user subscription information. The AI ​​module, based on the key network parameters, determines whether network congestion is currently occurring or predicts whether network congestion will occur within a preset time period after the current moment. Based on the application server information and the user subscription information, the AI ​​module determines the target business flow and target user that require QoS assurance. In the event of current or predicted network congestion, the AI ​​module generates target QoS policies for the target service flow and the target user, and distributes the target QoS policies to the Session Management Function (SMF).

2. The service quality strategy determination method according to claim 1, characterized in that, The acquisition of key network parameters, application server information, and user subscription information includes: The key parameters of the network are obtained from the Network Data Analysis Function (NWDAF). Obtain the application server information from the Network Openness Function (NEF); The user's subscription information is obtained from the unified data storage unit (UDR).

3. The service quality strategy determination method according to claim 1, characterized in that, The step of determining whether network congestion has occurred or predicting whether network congestion will occur within a preset time period after the current moment based on the network key parameters includes: If the network key parameters meet the congestion judgment conditions, then it is determined that network congestion has occurred. or, Based on historical network key parameter data, predict the network key parameters within the preset time period. If the predicted network key parameters meet the congestion judgment conditions, then predict that network congestion will occur within the preset time period. The key network parameters include at least one of the following: User Plane Function (UPF) load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency.

4. The service quality strategy determination method according to claim 1, characterized in that, The step of determining the target service flows and target users requiring QoS protection based on the application server information and the user subscription information includes: The service flows in the network are matched with the application server information to determine the service flows that belong to the target communication service type, which are then identified as the target service flows. From the user subscription information, at least one of the following is identified: users who have subscribed to the target communication service and users who belong to the high-priority subscription level, as the target users.

5. The service quality strategy determination method according to any one of claims 1 to 4, characterized in that, The target QoS policy includes at least one of the following: Increase the allocation and retention priority of the target user in ARP; Increase the Session-AMBR or UE-AMBR associated with the target service flow; Update the packet detection rules for the uplink UL or downlink DL to reserve bandwidth resources for the target service flow; The newly created sessions of the target service flow are routed to UPFs with loads less than a preset load threshold.

6. The service quality strategy determination method according to claim 1, characterized in that, The method further includes: If the generation or issuance of the target QoS policy fails or times out, the default QoS policy based on the user's subscription information shall be executed.

7. A service quality strategy determination device, characterized in that, include: The module includes an acquisition module, a determination module, a processing module, and a sending module. The acquisition module is used to acquire key network parameters, application server information, and user subscription information through the artificial intelligence (AI) module in the policy control function (PCF). The determining module is used to determine, through the AI ​​module and based on the key network parameters, whether network congestion has occurred at present or to predict whether network congestion will occur within a preset time after the current moment. The determining module is further configured to, through the AI ​​module, determine the target business flow and target user that require QoS assurance based on the application server information and the user subscription information; The processing module is used to generate target QoS policies for the target service flow and the target user through the AI ​​module when network congestion occurs currently or is predicted to occur. The sending module is used to send the target QoS policy to the Session Management Function (SMF).

8. The service quality strategy determination device according to claim 7, characterized in that, The acquisition module is specifically used for: The key parameters of the network are obtained from the Network Data Analysis Function (NWDAF). Obtain the application server information from the Network Openness Function (NEF); The user's subscription information is obtained from the unified data storage unit (UDR).

9. The service quality strategy determination device according to claim 7, characterized in that, The determining module is specifically used for: If the network key parameters meet the congestion judgment conditions, then it is determined that network congestion has occurred. or, Based on historical network key parameter data, predict the network key parameters within the preset time period. If the predicted network key parameters meet the congestion judgment conditions, then predict that network congestion will occur within the preset time period. The key network parameters include at least one of the following: User Plane Function (UPF) load rate, target interface bandwidth utilization, packet drop rate, average latency, and percentile latency.

10. An electronic device, characterized in that, The device includes a processor and a memory, the processor being coupled to the memory; the memory is used to store computer instructions, which are loaded and executed by the processor to enable the computer device to implement the service quality policy determination method as described in any one of claims 1 to 6.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes computer-executable instructions that, when executed on a computer, cause the computer to perform the service quality policy determination method as described in any one of claims 1 to 6.

12. A computer program product, characterized in that, The computer program product includes a computer program that, when run on an electronic device, causes the electronic device to perform the service quality policy determination method as described in any one of claims 1 to 6.

Images