User plane traffic flow analysis method, network function, and storage medium
By identifying the characteristics of user plane service flows and adjusting QoS policies, the problem of not being able to identify service flow characteristics in existing technologies is solved, enabling dynamic optimization of service flows and improvement of user experience.
Patent Information
- Application Number
- PCT/CN2026/092545
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-09
- Filing Date
- 2026-04-23
- Publication Date
- 2026-11-12
AI Technical Summary
Existing user plane functions cannot effectively identify the characteristics of service flows, such as audio or video streams, preventing operators from dynamically adjusting optimization strategies to improve QoS.
By receiving and analyzing indication information sent by various network functions, the characteristics of service flows are identified using models, and QoS policies are adjusted based on the identification results, including obtaining and using mapping tables, model training and inference to identify and match service types.
It enables the identification and performance enhancement of specific service types, and can adjust QoS policies according to service characteristics to improve user experience.
Smart Images

Figure CN2026092545_12112026_PF_FP_ABST
Abstract
Description
User plane service flow analysis methods, network functions, and storage media
[0001] Cross-reference to related applications
[0002] This disclosure claims priority to Chinese Patent Application No. 202510598877.3, filed in China on May 9, 2025, the entire contents of which are incorporated herein by reference. Technical Field
[0003] This disclosure relates to the field of mobile communication technology, specifically to a user plane service flow analysis method, network function, and storage medium. Background Technology
[0004] The transmission performance of service flows in the network is guaranteed by the Policy Control Function (PCF) based on the application function (AF) request to pre-configure policies. Operators typically do not perform dynamic optimization based on the identified performance and quality of service (QoS).
[0005] Currently, User Plane Function (UPF) only supports identifying a flow through the IP 5-tuple and cannot fuzzily identify certain service characteristics, such as audio streams and video streams. Related technologies' UPF can only identify the QoS of a flow through the IP 5-tuple and enhance the QoS guarantee of that flow by modifying the Policy and Charging Control (PCC) policy through the PCF. Summary of the Invention
[0006] At least one embodiment of this disclosure provides a user plane service flow analysis method, network function, and storage medium to address the problem of how to identify service characteristics using user plane functions.
[0007] To solve the above-mentioned technical problems, this disclosure is implemented as follows:
[0008] In a first aspect, embodiments of this disclosure provide a user plane service flow analysis method, including:
[0009] The first network function receives first information sent by the second network function, the first information including at least one of the following:
[0010] First indication information used to indicate the first business flow;
[0011] Secondary indication information used to indicate the business characteristics that identify the first business flow.
[0012] Optionally, it may also include one of the following:
[0013] The first network function sends second information to the third network function, wherein the second information includes at least one of the following: the first indication information, the second indication information, and model-related information;
[0014] The first network function sends third information to the fourth network function, wherein the third information includes the first indication information and the second indication information.
[0015] Optionally, the model-related information is a first model or a third indication information, wherein the third indication information is used to indicate the provision of a fifth network function for the first model;
[0016] The first model is used to identify the business characteristics of the business flow of the first business type, where the first business type is the business type of the first business flow.
[0017] Optionally, the method further includes:
[0018] Send fourth information to the fifth network function, the fourth information including the first service type.
[0019] Optionally, the fifth information sent by the fifth network function is received, the fifth information including the first model.
[0020] Optionally, if the model-related information is the third indication information, the method further includes:
[0021] Send sixth information to the sixth network function for locating the fifth network function. The sixth information includes at least one of the following: slice information, service area information, network function capability information, and AoI information.
[0022] Optionally, it also includes:
[0023] Receive the seventh information sent by the sixth network function, the seventh information including the third indication information.
[0024] Optionally, it also includes:
[0025] The system receives an eighth message sent by a third network function, the eighth message including at least one of the following: a fourth indication message for indicating a service flow to be adjusted, the service type of the service flow to be adjusted, and a first QoS policy determined by the third network function; wherein the service flow to be adjusted is a service flow for which the QoS policy needs to be adjusted.
[0026] Optionally, it also includes:
[0027] Send a ninth message to the fourth network function and / or the third network function, the ninth message including at least one of the following: a second QoS policy determined by the first network function; the second QoS policy being applied to the QoS policy of switching the service flow to be adjusted.
[0028] Optionally, it also includes:
[0029] The system receives a tenth message sent by a third network function, the tenth message including a fifth indication that the third network function has switched the QoS policy of the first service flow.
[0030] Secondly, embodiments of this disclosure provide a user plane service flow analysis method, including:
[0031] The third network function receives second information sent by the first network function or the fourth network function. The second information includes at least one of the following: first indication information for indicating a first service flow, second indication information for indicating the service characteristics of the first service flow, and model-related information; the model-related information is a first model or a third indication information, and the third indication information is used to indicate a fifth network function that provides the first model; the first model is used to identify the service characteristics of a service flow of a first service type, and the first service type is the service type of the first service flow.
[0032] Optionally, it also includes:
[0033] The business characteristics of the first business flow are identified using the first model.
[0034] Optionally, it may also include at least one of the following:
[0035] If the model-related information is the first model, the first model can be obtained directly from the model-related information;
[0036] If the model-related information is the third indication information, send eleventh information to the fifth network function, the eleventh information including the first service type; receive twelfth information sent by the fifth network function, the twelfth information including the first model; obtain the first model from the twelfth information;
[0037] If the second information does not include the model-related information, a thirteenth piece of information for locating the fifth network function is sent to the sixth network function. The thirteenth piece of information includes at least one of the following: slice information, service area information, network function capability information, and AoI information. The fourteenth piece of information sent by the sixth network function is received, and the fourteenth piece of information includes the third indication information. A fifteenth piece of information is sent to the fifth network function, and the fifteenth piece of information includes the first service type. The sixteenth piece of information sent by the fifth network function is received, and the sixteenth piece of information includes the first model. The first model is obtained from the sixteenth piece of information.
[0038] Optionally, the service characteristics of the first service flow include at least one of the following: audio service flow, video service flow, local service flow, remote service flow, importance of the service flow, packet burst size, packet burst interval, frame interval, and frame rate.
[0039] Optionally, the business characteristics of the first business flow are identified using the first model, including:
[0040] The first business flow is input into the first model to obtain the business characteristics of the first business flow output by the first model.
[0041] Optionally, it also includes:
[0042] Based on the service characteristics of the first service flow, determine whether the QoS policy of the first service flow needs to be adjusted.
[0043] Optionally, it also includes:
[0044] If it is determined that the QoS policy of the first service flow needs to be adjusted, an eighth message is sent to the first network function. The eighth message includes at least one of the following: a fourth indication message for indicating the service flow to be adjusted, the service type of the service flow to be adjusted, and the first QoS policy determined by the third network function; wherein the service flow to be adjusted is the first service flow.
[0045] Optionally, it also includes:
[0046] If it is determined that the QoS policy of the first service flow needs to be adjusted, a first QoS policy is determined.
[0047] Optionally, it also includes:
[0048] The system receives a ninth message sent by the first network function, the ninth message including at least one of the following: a second QoS policy determined by the first network function; the second QoS policy being applied to the QoS policy of switching the service flow to be adjusted.
[0049] Optionally, it also includes:
[0050] Switch the QoS policy of the first service flow to the second QoS policy.
[0051] Optionally, it also includes:
[0052] If it is determined that the QoS policy of the first service flow needs to be adjusted, a first QoS policy is determined, and the QoS policy of the first service flow is switched to the first QoS policy.
[0053] Optionally, it also includes:
[0054] The tenth message sent to the first network function includes a fifth indication that the third network function has switched the QoS policy of the first service flow.
[0055] Optionally, it also includes:
[0056] If the first model fails to identify the service characteristics of the first service flow, a seventeenth message is sent to the fifth network function. The seventeenth message includes at least one of the following: the first service type, the first model, and a sixth indication message. The sixth indication message indicates that the first model failed to identify the service characteristics of the service flow of the first service type.
[0057] Optionally, it also includes:
[0058] Receive the eighteenth message sent by the fifth network function, the eighteenth message including the updated first model.
[0059] Optionally, it also includes:
[0060] The updated first model is used to identify the business characteristics of the first business flow.
[0061] Thirdly, embodiments of this disclosure provide a user plane service flow analysis method, including:
[0062] The fifth network function receives the twenty-fifth information sent by the first network function, the third network function, or the fourth network function, and the twenty-fifth information includes the first service type.
[0063] Optionally, the method further includes:
[0064] The fifth network function sends a twenty-sixth message to the first network function, the third network function, or the fourth network function. The twenty-sixth message includes a first model, which is a model corresponding to the first service type.
[0065] Optionally, it also includes:
[0066] The fifth network function maintains a mapping table between service types and models, and the mapping table includes mapping probabilities between various service types and various models.
[0067] Optionally, it also includes:
[0068] The fifth network function searches the mapping table to obtain the first model corresponding to the first service type, wherein the first model is the model with the highest mapping probability among all the models mapped to the first service type.
[0069] Optionally, it also includes:
[0070] If the first business type does not exist in the mapping table, a first model for identifying the business characteristics of the business flow of the first business type is generated through learning and reasoning, and the mapping table is updated according to the first model and the first business type.
[0071] Optionally, the multiple models in the mapping table are multiple small models locally configured by the fifth network function. The small models are obtained by distillation training based on the large model in an external database, and the parameter specifications of the small models are smaller than those of the large models.
[0072] Optionally, the distillation training includes:
[0073] For the first service type, obtain the first probability distribution between the multiple small models transmitted by the large model and the first service type, and obtain the first model corresponding to the first service type;
[0074] Based on the weight parameters, the distillation loss of the first model and the supervision loss of the first model in local supervised learning are weighted and summed to obtain the comprehensive loss;
[0075] The first business type training samples are input into the large model and the first model. By comparing the recognition results of the large model and the multiple small models and adjusting the weight parameters, the overall loss is minimized, and the trained first model is obtained.
[0076] Fourthly, embodiments of this disclosure provide a first network function, including:
[0077] The receiving module is configured to receive first information sent by the second network function, the first information including at least one of the following:
[0078] First indication information used to indicate the first business flow;
[0079] Secondary indication information used to indicate the business characteristics that identify the first business flow.
[0080] Fifthly, embodiments of this disclosure provide a third network function, including:
[0081] The receiving module is configured to receive second information sent by a first network function or a fourth network function, the second information including at least one of the following: first indication information for indicating a first service flow, second indication information for indicating service characteristics of identifying the first service flow, and model-related information; the model-related information is a first model or a third indication information, the third indication information being used to indicate a fifth network function providing the first model; the first model is used to identify service characteristics of a service flow of a first service type, the first service type being the service type of the first service flow.
[0082] Sixthly, embodiments of this disclosure provide a fifth network function, including:
[0083] The receiving module is used to receive the twenty-fifth information sent by the first network function, the third network function or the fourth network function, wherein the twenty-fifth information includes the first service type;
[0084] The sending module is used to send the 26th information to the first network function, the third network function, or the fourth network function. The 26th information includes a first model, which is a model corresponding to the first service type.
[0085] Seventhly, embodiments of this disclosure provide a user plane service analysis method, including:
[0086] The third network function determines the third QoS policy for the first service flow;
[0087] The third network function configures the third QoS policy for the first service flow.
[0088] Optionally, the third network function determines a third QoS policy for the first service flow, including at least one of the following:
[0089] The third network function determines the first service type of the first service flow;
[0090] The third network function obtains the second model from the fifth network function, and uses the second model to match and obtain the third QoS policy for the first service type.
[0091] Optionally, the third network function determines the first service type of the first service flow, including:
[0092] The third network function determines the first service type of the first service flow based on at least one of the following information and / or models: a pre-configured strategy, a locally trained model, a model obtained from the fifth network function, the flow description information of the service flow, the application detection information, and the application ID information.
[0093] Optionally, the third network function determines a third QoS policy for the first service flow, including:
[0094] The third network function sends the nineteenth information to the fifth network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate.
[0095] The third network function receives the twentieth information sent by the fifth network function, the twentieth information including the third QoS policy of the first service flow.
[0096] Optionally, it also includes:
[0097] The third network function sends a twenty-first message to the fourth network function, the twenty-first message being used to notify the fourth network function to configure the third QoS policy for the first service flow in the radio access network.
[0098] Optionally, it may also include at least one of the following:
[0099] When the third network function detects a change in the service characteristics of the first service flow, it determines a fourth QoS policy based on the new service characteristics of the first service flow.
[0100] The third network function executes the fourth QoS policy for the first service flow and sends twenty-second information to the first network function. The twenty-second information includes at least one of the following: new service characteristics of the first service flow and the fourth QoS policy.
[0101] The third network function receives the twenty-third or twenty-fourth information sent by the first network function, wherein the twenty-third information is used to instruct the third network function to execute the fourth QoS policy, and the twenty-fourth information is used to instruct the third network function to execute the third QoS policy;
[0102] The third network function may, based on the second-third information, continue to execute the fourth QoS policy for the first service flow, or, based on the second-fourth information, execute the third QoS policy for the first service flow.
[0103] Eighthly, embodiments of this disclosure provide a user plane service analysis method, including:
[0104] The fifth network function sends a second model to the third network function, which is used to match the QoS policy corresponding to the service type.
[0105] Ninthly, embodiments of this disclosure provide a user plane service analysis method, including:
[0106] The fifth network function receives the nineteenth information sent by the third network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate.
[0107] The fifth network function analyzes the information included in the nineteenth information to obtain the analysis result; and determines the third QoS strategy for the first service flow based on the analysis result.
[0108] The fifth network function sends twentieth information to the third network function, the twentieth information including the third QoS policy of the first service flow.
[0109] Optionally, the first network function includes PCF, the second network function includes AF, the third network function includes UPF, the fourth network function includes SMF, the fifth network function includes NDWAF, and the sixth network function includes NRF.
[0110] In a tenth aspect, embodiments of this disclosure provide a third network function, including:
[0111] The determination module is used to determine the third QoS policy for the first service flow;
[0112] The configuration module is used to configure the third QoS policy for the first service flow by the third network function.
[0113] In one aspect, embodiments of this disclosure provide a fifth network function, including:
[0114] The second sending module is used to send the second model to the third network function, and the second model is used to match the QoS policy corresponding to the service type.
[0115] In a twelfth aspect, embodiments of this disclosure provide a fifth network function, including:
[0116] The second receiving module is used to receive the nineteenth information sent by the third network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate.
[0117] The analysis module is used to analyze the information included in the nineteenth piece of information to obtain analysis results; and to determine the third QoS strategy for the first service flow based on the analysis results.
[0118] The third sending module is used to send twentieth information to the third network function, the twentieth information including the third QoS policy of the first service flow.
[0119] In a thirteenth aspect, embodiments of this disclosure provide a network function, including: a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the method as described in any of the preceding embodiments.
[0120] In a fourteenth aspect, embodiments of this disclosure provide a computer-readable storage medium storing a program that, when executed by a processor, implements the steps of the method described in any of the preceding embodiments.
[0121] In a fifteenth aspect, embodiments of this disclosure provide a computer program product including computer instructions that, when executed by a processor, implement the steps of the method described in any of the preceding embodiments.
[0122] Compared with related technologies, the user plane service flow analysis method, network function and storage medium provided in this disclosure can enable the network to identify and enhance the performance of specific service types / feature service flows, help the network identify a certain type of service, configure policies to improve service experience, such as providing application B's QoS policy for video service of application A to improve service experience. Attached Figure Description
[0123] Figure 1 is a schematic diagram of an application scenario according to an embodiment of this disclosure;
[0124] Figure 2 is a flowchart of a user plane service flow analysis method according to an embodiment of the present disclosure;
[0125] Figure 3 is another flowchart of the user plane service flow analysis method according to an embodiment of this disclosure;
[0126] Figure 4 is another flowchart of the user plane service flow analysis method according to an embodiment of the present disclosure;
[0127] Figure 5 is an interactive flowchart of the user plane service flow analysis method of Example 1 of this disclosure;
[0128] Figure 6 is an interactive flowchart of the user plane service flow analysis method of Example 2 of this disclosure;
[0129] Figure 7 is an interactive flowchart of the user plane service flow analysis method of Example 3 of this disclosure;
[0130] Figure 8 is an interactive flowchart of the user plane service flow analysis method of Example 4 of this disclosure;
[0131] Figure 9 is an interactive flowchart of the user plane service flow analysis method of Example 5 of this disclosure;
[0132] Figure 10 is a schematic diagram of the structure of a first network function according to an embodiment of the present disclosure;
[0133] Figure 11 is a schematic diagram of the structure of the third network function according to an embodiment of the present disclosure;
[0134] Figure 12 is a schematic diagram of the structure of the fifth network function according to an embodiment of the present disclosure;
[0135] Figure 13 is a flowchart of a user plane service flow analysis method according to another embodiment of the present disclosure;
[0136] Figure 14 is a flowchart of a user plane service flow analysis method according to another embodiment of the present disclosure;
[0137] Figure 15 is another flowchart of a user plane service flow analysis method according to yet another embodiment of this disclosure;
[0138] Figure 16 is an interactive flowchart of the user plane service flow analysis method of Example 6 of this disclosure;
[0139] Figure 17 is an interactive flowchart of the user plane service flow analysis method of Example 7 of this disclosure;
[0140] Figure 18 is a schematic diagram of the structure of the third network function according to another embodiment of the present disclosure;
[0141] Figure 19 is a schematic diagram of the structure of the fifth network function according to another embodiment of the present disclosure;
[0142] Figure 20 is a schematic diagram of the structure of the fifth network function according to yet another embodiment of the present disclosure;
[0143] Figure 21 is a schematic diagram of a network function according to an embodiment of the present disclosure. Detailed Implementation
[0144] The terms "first," "second," etc., used in this disclosure are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this disclosure can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, not limited in number; for example, the first object can be one or more. Furthermore, "or" in this disclosure indicates at least one of the connected objects. For example, "A or B" covers three scenarios: Scenario 1: including A but not B; Scenario 2: including B but not A; Scenario 3: including both A and B. The character " / " generally indicates that the preceding and following objects are in an "or" relationship.
[0145] The term "instruction" in this disclosure can be either a direct instruction (or explicit instruction) or an indirect instruction (or implicit instruction). A direct instruction can be understood as one in which the sender explicitly informs the receiver of specific information, the operation to be performed, or the requested result, etc.; an indirect instruction can be understood as one in which the receiver determines the corresponding information based on the instruction sent by the sender, or makes a judgment and determines the operation to be performed or the requested result, etc., based on the judgment result.
[0146] It is worth noting that the technologies described in this disclosure are not limited to Long Term Evolution (LTE) / LTE-Advanced (LTE-A) systems, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA), or other systems. The terms "system" and "network" in this disclosure are often used interchangeably, and the described technologies can be used with the systems and radio technologies mentioned above, as well as with other systems and radio technologies. The following description describes New Radio (NR) systems for illustrative purposes, and the term NR is used in most of the following description; however, these technologies can also be applied to systems other than NR systems, such as 6th Generation (6G) communication systems.
[0147] Figure 1 shows a block diagram of a wireless communication system applicable to embodiments of the present disclosure. The wireless communication system includes a terminal 11 and a network device 12. The terminal 11 can be a mobile phone, tablet computer, laptop computer, notebook computer, personal digital assistant (PDA), handheld computer, netbook, ultra-mobile personal computer (UMPC), mobile internet device (MID), augmented reality (AR), virtual reality (VR) device, robot, wearable device, flight vehicle, vehicle user equipment (VUE), shipboard equipment, pedestrian user equipment (PUE), smart home (home devices with wireless communication capabilities, such as refrigerators, televisions, washing machines, or furniture), game console, personal computer (PC), ATM, or self-service machine, etc. Wearable devices include: smartwatches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart chains, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. Among these, in-vehicle devices can also be referred to as in-vehicle terminals, in-vehicle controllers, in-vehicle modules, in-vehicle components, in-vehicle chips, or in-vehicle units, etc. It should be noted that the specific type of terminal 11 is not limited in this disclosure embodiment. Network device 12 may include access network devices or core network devices, wherein access network devices may also be referred to as Radio Access Network (RAN) devices, radio access network functions, or radio access network units. Access network devices may include base stations, Wireless Local Area Network (WLAN) access points (APs), or Wireless Fidelity (WiFi) nodes, etc.In this context, a base station may be referred to as a Node B (NB), Evolved Node B (eNB), Next Generation Node B (gNB), New Radio Node B (NR Node B), Access Point, Relay Base Station (RBS), Serving Base Station (SBS), Base Transceiver Station (BTS), Radio Base Station, Radio Transceiver, Basic Service Set (BSS), Extended Service Set (ESS), Home Node B (HNB), Home Evolved Node B, Transmission Reception Point (TRP), or any other suitable term in the relevant field, as long as the same technical effect is achieved. The base station is not limited to any specific technical terminology. It should be noted that in this disclosure, only a base station in an NR system is used as an example for description, and the specific type of base station is not limited.
[0148] Core network equipment may include, but is not limited to, at least one of the following: core network node, core network function, Mobility Management Entity (MME), Access and Mobility Management Function (AMF), Session Management Function (SMF), User Plane Function (UPF), Policy Control Function (PCF), Policy and Charging Rules Function (PCRF), Edge Application Server Discovery Function (EASDF), Unified Data Management (UDM), Unified Data Repository (UDR), Home Subscriber Server (HSS), Centralized network configuration (CNC), Network Repository Function (NRF), Network Exposure Function (NEF), Local NEF (or L-NEF), Binding Support Function (BSF), and Application Function. Functions, AFs, etc. It should be noted that this disclosure only uses the core network equipment in the NR system as an example for introduction, and does not limit the specific type of core network equipment.
[0149] As described in the background section, if user plane functionality can identify specific business characteristics, it can intelligently analyze whether the business has better performance guarantees under other strategies based on these characteristics, thereby enhancing the user experience of a certain type of business. This disclosure provides a user plane business flow analysis method that can identify business characteristics, thereby supporting the enhancement of the user experience of the business.
[0150] To facilitate understanding, the network functions involved in the embodiments of this disclosure will be described first.
[0151] First network function: A network function used to perform policy control. In NR systems, this first network function may specifically be PCF, and in 6G systems it may also be called PCF or other names.
[0152] Second network function: Application function. In NR systems, this second network function may specifically be Application Function (AF), and in 6G systems it may also be called AF or other names.
[0153] Third network function: A network function used to perform user plane functions. Specifically, the third network function in an NR system may be a UPF, and in a 6G system it may also be called UPF or other names.
[0154] Fourth network function: A network function used to perform session management. In NR systems, this fourth network function may specifically be SMF, and in 6G systems it may also be called SMF or other names.
[0155] Fifth Network Function: A network function used to perform network data analysis and processing. Specifically, in NR systems, this fifth network function may be the Network Data Analytics Function (NDWAF), and in 6G systems, it may also be called NDWAF or other names.
[0156] Sixth Network Function: A network function used to store network data. In NR systems, this sixth network function may specifically be NRF, and in 6G systems it may also be referred to as NRF or other names.
[0157] Figure 2 is a flowchart illustrating the application of a user plane service flow analysis method provided in this embodiment of the present disclosure to a first network function. As shown in Figure 2, the method includes:
[0158] Step 21: The first network function receives first information sent by the second network function, the first information including at least one of the following:
[0159] First indication information used to indicate the first business flow;
[0160] Secondary indication information used to indicate the business characteristics that identify the first business flow.
[0161] Here, the first indication information may specifically be the IP 5-tuple and / or flow identifier of the first service flow, etc. The second indication information is used to indicate the service characteristics that identify the first service flow. The service characteristics of a certain service flow may specifically include at least one of the following: whether it is an audio service flow, whether it is a video service flow, whether it is a local service flow, whether it is a remote service flow, the importance of the service flow, etc.
[0162] Through the above steps, the first network function receives the first information sent by the second network function, and thereby instructs the relevant network function to identify the service characteristics of the first service flow through the second indication information in the first information.
[0163] For example, after step 21 above, the method further includes:
[0164] Step 22: The first network function sends the second information to the third network function, or sends the third information to the fourth network function;
[0165] In step 22, the first network function can forward the second information to the third network function through the fourth network function. Here, the fourth network function can specifically be a function such as SMF. The second information includes at least one of the following: the first indication information, the second indication information, and model-related information; the model-related information is either the first model or the third indication information, and the third indication information is used to indicate the fifth network function providing the first model; the first model is used to identify the service characteristics of the service flow of the first service type, and the first service type is the service type of the first service flow. In this embodiment, the service characteristics can be a service type, a service pattern, a traffic pattern, a service behavior, or traffic behavior, etc. The service type can be a traffic pattern, a service pattern, a service behavior, or traffic behavior.
[0166] The third information includes the first indication information and the second indication information. Through the third information, the fourth network function can acquire the first model and then send at least one of the first indication information, the second indication information, and the first model to the third network function.
[0167] Thus, through the above steps, the third network function can acquire the first indication information and / or the second indication information, then determine the first service flow based on the first indication information, determine the need to identify service characteristics of the first service flow based on the second indication information, obtain the first model based on the model-related information, and finally identify the service characteristics of the first service flow based on the first model. Since the third network function is a user plane network function (such as UPF), this embodiment of the disclosure can identify the service characteristics of the service flow at the user plane, providing support for improving the user experience of the service flow based on service characteristics.
[0168] In this embodiment of the disclosure, a mapping table between service types and models is pre-maintained at the fifth network function. The mapping table includes mapping probabilities between various service types and various models, such as the probability of service type 1 mapping to model 1, the probability of service type 1 mapping to model 2, and so on. Through the mapping table, the model corresponding to each service type can be determined, wherein the model corresponding to a certain service type is the model with the highest mapping probability for that service type.
[0169] Before sending the second information, the first network function may send a fourth information to the fifth network function, the fourth information including the first service type. The fourth information is used to request the first model corresponding to the first service type. Furthermore, the first network function may also receive a fifth information sent by the fifth network function, the fifth information including the first model. Specifically, the fifth information may include model parameters of the first model. Thus, when sending the second information, the first model can be carried in the first information and sent to the third network function.
[0170] In some embodiments of this disclosure, the model-related information is the third indication information, which can indicate the fifth network function. Specifically, it can be information such as the address or identifier of the fifth network function. Thus, after obtaining the third indication information, the third network function can request the first model from the fifth network function.
[0171] Since multiple fifth network functions may exist in the network, and different fifth network functions may have different service areas, the first network function may send sixth information to the sixth network function to locate the fifth network function before sending the second information carrying the third indication information. The sixth information includes at least one of the following: slice information, service area information, network function capability information, and Area of Interest (AoI) information. Further, the first network function receives seventh information sent by the sixth network function, which includes the third indication information and is used to indicate a specific fifth network function. Thus, by sending the third indication information to the third network function, the first network function enables the third network function to request the first model from the corresponding fifth network function based on the third indication information.
[0172] In this embodiment of the disclosure, the third network function can use the first model to identify the service characteristics of the first service flow, and then determine whether the QoS policy of the first service flow needs to be adjusted based on the service characteristics of the first service flow.
[0173] For example, if the third network function determines that the QoS policy of the first service flow needs to be adjusted, the third network function can request the first network function to adjust the QoS policy of the first service flow. Specifically, the third network function can send an eighth message to the first network function, the eighth message including at least one of the following: a fourth indication message for indicating the service flow to be adjusted, the service type of the service flow to be adjusted, and a first QoS policy determined by the third network function; wherein the service flow to be adjusted is the first service flow. The first network function then receives the eighth message sent by the third network function. Furthermore, the first network function can determine a second QoS policy applied to the first service flow based on the eighth message. The first QoS policy is a QoS policy suggested by the third network function, for the first network function to refer to when determining the QoS policy applied to the first service flow. The second QoS policy ultimately determined by the first network function for the first service flow may be the same as or different from the first QoS policy. Then, the first network function sends ninth information to the fourth network function and / or the third network function, the ninth information including at least one of the following: a second QoS policy determined by the first network function; the second QoS policy being applied to switch the QoS policy of the service flow to be adjusted. In this way, the QoS policy of the first service flow can be switched to the second QoS policy, realizing the adjustment of the QoS policy of the service flow based on its service characteristics.
[0174] For example, if the third network function determines that the QoS policy of the first service flow needs to be adjusted, the third network function can adjust the QoS policy of the first service flow itself. In this case, the third network function automatically switches the QoS policy of the first service flow and sends a tenth message to the first network function to notify that the QoS policy of the first service flow has been switched. The first network function then receives the tenth message sent by the third network function, which includes a fifth indication that the third network function has switched the QoS policy of the first service flow. Thus, the first network function can know that the QoS policy of the first service flow has been switched.
[0175] Figure 3 is a flowchart illustrating the application of a user plane service flow analysis method provided in this embodiment of the present disclosure to a third network function. As shown in Figure 3, the method includes:
[0176] Step 31: The third network function receives second information sent by the first network function or the fourth network function. The second information includes at least one of the following: first indication information for indicating the first service flow, second indication information for indicating the service characteristics of the first service flow, and model-related information; the model-related information is a first model or a third indication information, and the third indication information is used to indicate a fifth network function that provides the first model; the first model is used to identify the service characteristics of the service flow of the first service type, and the first service type is the service type of the first service flow.
[0177] Through the above steps, the third network function can obtain one or more of the first indication information, the second indication information, and model-related information from the second information, thereby providing support for the third network function to identify the business characteristics of the first business flow.
[0178] The third network function can obtain the first model based on the model-related information, or it can request the first model from the fifth network function.
[0179] For example, if the model-related information is the first model, the third network function directly obtains the first model from the model-related information;
[0180] For example, when the model-related information is the third indication information, the third network function sends an eleventh message to the fifth network function, the eleventh message including the first service type; receives a twelfth message sent by the fifth network function, the twelfth message including the first model; and obtains the first model from the twelfth message.
[0181] For example, if the second information does not include the model-related information, the third network function sends a thirteenth piece of information to the sixth network function for locating the fifth network function. The thirteenth piece of information includes at least one of the following: slice information, service area information, network function capability information, and AoI information; receives a fourteenth piece of information sent by the sixth network function, the fourteenth piece of information including the third indication information; sends a fifteenth piece of information to the fifth network function, the fifteenth piece of information including the first service type; receives a sixteenth piece of information sent by the fifth network function, the sixteenth piece of information including the first model; and obtains the first model from the sixteenth piece of information.
[0182] Thus, the third network function can utilize the first model to identify the service characteristics of the first service flow. In this embodiment, the service characteristics can be service type, service pattern, traffic pattern, service behavior, or traffic behavior, etc. Specifically, the third network function can input the first service flow into the first model to obtain the service characteristics of the first service flow output by the first model. The service characteristics of the first service flow include at least one of the following: audio service flow, video service flow, local service flow, remote service flow, service flow importance, packet burst size, packet burst interval, frame interval, and frame rate.
[0183] Furthermore, the third network function can determine whether the QoS policy of the first service flow needs to be adjusted based on the service characteristics of the first service flow.
[0184] For example, if the third network function determines that the QoS policy of the first service flow needs to be adjusted, it sends an eighth message to the first network function. This eighth message includes at least one of the following: a fourth indication message for indicating the service flow to be adjusted, and the service type of the service flow to be adjusted; wherein the service flow to be adjusted is the first service flow. Subsequently, the third network function can receive a ninth message sent by the first network function. This ninth message includes at least one of the following: a second QoS policy determined by the first network function; the second QoS policy is applied to switch the QoS policy of the service flow to be adjusted. Then, the third network function switches the QoS policy of the first service flow to the first QoS policy.
[0185] For example, if the third network function determines that the QoS policy of the first service flow needs to be adjusted, it can determine a first QoS policy and send eighth information to the first network function. The eighth information includes at least one of the following: fourth indication information for indicating the service flow to be adjusted, the service type of the service flow to be adjusted, and the first QoS policy determined by the third network function; wherein the service flow to be adjusted is the first service flow. The first QoS policy is a QoS policy suggested by the third network function and is provided to the first network function as reference information for QoS policy decision-making. Subsequently, the third network function can receive ninth information sent by the first network function, which includes at least one of the following: a second QoS policy determined by the first network function; the second QoS policy is applied to switch the QoS policy of the service flow to be adjusted. Then, the third network function switches the QoS policy of the first service flow to the first QoS policy.
[0186] For example, if the third network function determines that the QoS policy of the first service flow needs to be adjusted, it can also determine a first QoS policy and switch the QoS policy of the first service flow to the first QoS policy. Furthermore, the tenth information sent by the third network function to the first network function includes a fifth indication that the third network function has switched the QoS policy of the first service flow. Thus, the tenth information notifies the first network function that the QoS policy of the first service flow has been switched.
[0187] When the third network function uses the first model to identify the service characteristics of the first service flow, if it fails to identify the service characteristics of the first service flow, the third network function sends a seventeenth piece of information to the fifth network function. The seventeenth piece of information includes at least one of the following: the first service type, the first model, and a sixth indication. The sixth indication indicates that the first model failed to identify the service characteristics of the service flow of the first service type. The fifth network function updates the first model based on the information in the seventeenth piece of information. Subsequently, the third network function can also receive an eighteenth piece of information sent by the fifth network function, which includes the updated first model. Then, the third network function can also use the updated first model to identify the service characteristics of the first service flow.
[0188] Figure 4 is a flowchart illustrating the application of a user plane service flow analysis method provided in this embodiment of the present disclosure to a fifth network function. As shown in Figure 4, the method includes:
[0189] Step 41: The fifth network function receives the twenty-fifth information sent by the first network function, the third network function, or the fourth network function, wherein the twenty-fifth information includes the first service type;
[0190] Step 42: The fifth network function sends the twenty-sixth information to the first network function, the third network function, or the fourth network function. The twenty-sixth information includes a first model, which is a model corresponding to the first service type.
[0191] Through the above steps, the fifth network function returns the first model to the first network function, the third network function, or the fourth network function, enabling the relevant network function to obtain the first model and then identify the service characteristics of the first service flow, thereby providing support for identifying the service characteristics of the first service flow.
[0192] Optionally, the fifth network function maintains a mapping table between service types and models, the mapping table including mapping probabilities between multiple service types and multiple models. The fifth network function provides a method to look up the mapping table to obtain a first model corresponding to the first service type, wherein the first model is the model with the highest mapping probability among all models mapped to the first service type. Thus, in step 42, the first model can be sent to the relevant network function.
[0193] In this embodiment of the disclosure, if the mapping table does not contain the first service type, the fifth network function can generate a first model for identifying the service characteristics of the service flow of the first service type through learning and reasoning, and update the mapping table according to the first model and the first service type.
[0194] Specifically, the multiple models in the mapping table are multiple small models locally configured by the fifth network function. The small models are obtained by distillation training based on the large model in the external database, and the parameter specifications of the small models are smaller than those of the large models.
[0195] Taking the first model as an example, the distillation training includes: for the first service type, obtaining the first probability distribution between the multiple small models transmitted by the large model and the first service type, and obtaining the first model corresponding to the first service type; according to the weight parameters, weighting and summing the distillation loss of the first model and the supervision loss of the first model in local supervised learning to obtain a comprehensive loss; inputting the training samples of the first service type into the large model and the first model, and by comparing the recognition results of the large model and the multiple small models and adjusting the weight parameters, minimizing the comprehensive loss to obtain the trained first model.
[0196] The methods of the embodiments of this disclosure have been described above from the perspectives of the first network function, the third network function, and the fifth network function.
[0197] The above method will be further explained below through more specific interaction examples between network functions. It should be noted that in the following examples, the first network function is PCF, the second network function is AF, the third network function is UPF, the fourth network function is SMF, the fifth network function is NDWAF, the sixth network function is NRF, and the seventh network function is NEF.
[0198] Example 1:
[0199] In this example, PCF acquires the model, and UPF uses the model provided by PCF to identify business features. As shown in Figure 5, this example specifically includes:
[0200] Step 51: The AF sends an AF request to the PCF via the NEF, which carries indication information (i.e., the second indication information mentioned above) to indicate the service characteristics of the service flow.
[0201] Step 52: The PCF sends a request to the AI analysis network element (e.g., NWDAF) to obtain a model for acquiring and identifying the service characteristics of the service flow. The request carries indication information of the service type of the service flow (i.e., the first service type mentioned above), such as IP packets, service flows of a certain application, etc. The NWDAF finds the model corresponding to the service type and sends the model to the PCF.
[0202] Additionally, in step 52, the PCF can determine the NWDAF in the following way: the PCF sends a request to the NRF to find the NWDAF serving the PCF. This request carries information such as slices (e.g., Single Network Slice Selection Assistance Information (S-NSSAI)), service area information (TAIs list), NWDAF capabilities, and AoI. The NRF then finds and returns the corresponding NWDAF information, such as the NWDAF's address. Thus, in step 52, the PCF can send a request to the NWDAF based on its address.
[0203] Step 53: The PCF sends a policy configuration to the UPF through the SMF, which carries the IP 5-tuple of the service flow, indication information indicating the service characteristics of the service flow, and the model.
[0204] Step 54: After receiving the policy configuration, the UPF obtains the model and then uses the model to infer and identify the service characteristics of the service flow. Service characteristics may include, but are not limited to, audio, video, local services, remote services, and the importance of the flow. Specific judgment criteria include, but are not limited to, frame interval, periodicity, size, I / P frames, and traffic characteristics. For example, frames with a certain size range are considered video streams; frame interval and periodicity determine frame importance; and traffic within a certain range indicates local services.
[0205] Different models correspond to different business types and can be stored locally on the NWDAF. Furthermore, the models can be dynamically updated based on the emergence of new business types. For example, when the PCF receives an AF request indicating that the business flow to be identified is a new business type, the PCF configures this new business type in the NWDAF, and the NWDAF learns and infers the model corresponding to this new business type.
[0206] NWDAF Model Training (Distillation Training): The model configured locally in NWDAF is a small model with limited information, resulting in low deployment costs. An external database contains a large model with full information, which transmits the probability distribution (soft labels) of a model matching a specific business type to the small model. A comprehensive loss function is established for the small model, such as a * distillation loss + (1-a) * true supervision loss, where parameter a is an adjustable weight. By combining the hard labels from the original local supervised learning and the soft labels from the large model, the optimal solution for the comprehensive loss is obtained. Training samples of the same business type are input into both the large and small models. By comparing the recognition results of the large and small models, parameter a is continuously adjusted to minimize the comprehensive loss of the small model, thus completing the local model training for NWDAF.
[0207] Example 2:
[0208] In this example, SMF acquires the model, and UPF uses the model provided by SMF to identify business features. As shown in Figure 6, this example specifically includes:
[0209] Step 61: The AF sends an AF request to the PCF via the NEF, which carries indication information (i.e., the second indication information mentioned above) to indicate the service characteristics of the service flow.
[0210] Step 62: The PCF sends a policy configuration to the UPF through the SMF, which carries the IP 5-tuple of the service flow and indication information for identifying the service characteristics of the service flow (i.e., the second indication information mentioned above).
[0211] Step 63: After receiving the policy configuration from the PCF, the SMF sends a request to the AI analysis network element (e.g., NWDAF) to obtain a model for acquiring and identifying the service characteristics of the service flow. The request carries service type indication information of the service flow (i.e., the first service type mentioned above), such as IP packets, service flows of a certain application, etc. The NWDAF looks up the model corresponding to the service type and sends the model to the PCF.
[0212] Additionally, in step 63, the SMF can determine the NWDAF in the following way: the SMF sends a request to the NRF to find the NWDAF serving the PCF, the request carrying information such as slice (S-NSSAI), service area information (TAIs list), NWDAF capabilities, and AoI. The NRF finds and returns the corresponding NWDAF information, such as the NWDAF address. Thus, in step 63, the SMF can send a request to the NWDAF based on the NWDAF address.
[0213] Step 64: SMF sends policy configuration to UPF, wherein the policy configuration carries the IP 5-tuple of the service flow, indication information indicating the service characteristics of the service flow, and the model.
[0214] Step 65: After receiving the policy configuration, the UPF obtains the model and then uses the model to infer and identify the service characteristics of the service flow. Service characteristics may include, but are not limited to, audio, video, local services, remote services, and the importance of the flow. Specific judgment criteria include, but are not limited to, frame interval, periodicity, size, I / P frames, and traffic characteristics. For example, frames with a certain size range are considered video streams; frame interval and periodicity determine frame importance; and traffic within a certain range indicates local services.
[0215] Example 3:
[0216] In this example, the UPF requests the model from the NWDAF using the address provided by the PCF; the UPF then uses the model to identify business characteristics. As shown in Figure 7, this example specifically includes:
[0217] Step 71: The AF sends an AF request to the PCF via the NEF, which carries indication information (i.e., the second indication information mentioned above) to indicate the service characteristics of the service flow.
[0218] Step 72: The PCF sends a policy configuration to the UPF through the SMF, which carries the IP 5-tuple of the service flow, indication information for identifying the service characteristics of the service flow (i.e., the second indication information mentioned above), and the address of the AI analysis network element (such as NWDAF) (i.e., the third indication information mentioned above).
[0219] Additionally, in step 72, the PCF can determine the NWDAF in the following way: the PCF sends a request to the NRF to find the NWDAF serving the PCF, the request carrying information such as slice (S-NSSAI), service area information (TAIs list), NWDAF capabilities, and AoI. The NRF finds and returns the corresponding NWDAF information, such as the NWDAF address.
[0220] Step 73: After receiving the policy configuration, the UPF sends a request to the AI analysis network element (e.g., NWDAF) to obtain a model for acquiring and identifying the service characteristics of the service flow. The request carries service type indication information of the service flow (i.e., the first service type mentioned above), such as IP packets, service flows of a certain application, etc. The NWDAF finds the model corresponding to the service type and sends the model to the UPF.
[0221] Step 74: UPF identifies service characteristics through model inference. Service characteristics may include audio, video, local services, remote services, and the importance of the stream. Judgment criteria include frame interval, periodicity, size, I / P frames, and traffic characteristics. For example, frames with a certain size range are considered video streams.
[0222] Step 75: If the UPF model fails to identify business features, it can be fed back to the NWDAF to dynamically update the model.
[0223] Example 4:
[0224] In this example, the UPF determines the address of the NWDAF and requests the model from the NWDAF; the UPF uses the model to identify business characteristics. As shown in Figure 8, this example specifically includes:
[0225] Step 81: The AF sends an AF request to the PCF via the NEF, which carries indication information (i.e., the second indication information mentioned above) to indicate the service characteristics of the service flow.
[0226] Step 82: The PCF sends a policy configuration to the UPF through the SMF, which carries the IP 5-tuple of the service flow and indication information for identifying the service characteristics of the service flow (i.e., the second indication information mentioned above).
[0227] Step 83: After receiving the policy configuration, the UPF sends a request to the sixth network function (such as the NRF) to find the NWDAF serving the PCF. This request carries information such as slice (S-NSSAI), service area information (TAIs list), NWDAF capabilities, and AoI. The NRF then finds and returns the corresponding NWDAF information, such as the NWDAF's address.
[0228] Step 84: The UPF sends a request to the NWDAF based on the NWDAF's address, requesting to obtain a model for identifying the service characteristics of the service flow. The request carries service type indication information of the service flow (i.e., the first service type mentioned above), such as IP packets, service flows of a certain application, etc. The NWDAF finds the model corresponding to the service type and sends the model to the UPF.
[0229] Step 85: UPF identifies service characteristics through model inference. Service characteristics may include audio, video, local services, remote services, and the importance of the stream. Judgment criteria include frame interval, periodicity, size, I / P frames, and traffic characteristics. For example, frames with a certain size range are considered video streams.
[0230] Example 5:
[0231] In this example, the UPF reports the modified QoS policy for the service flow to the PCF. For instance, after intelligently identifying a specific service flow and service type, the UPF reports the service flow parameters (IP 5-tuple, service type) to the PCF, and the PCF then issues the policy to the SMF-UPF to enhance the QoS guarantee for that flow. As shown in Figure 9, this example specifically includes:
[0232] Step 90: The UPF identifies and detects the service flow to be adjusted. Specifically, the service flow to be adjusted may be a pre-determined service flow requiring enhanced QoS guarantees, or a service flow whose QoS does not meet preset requirements.
[0233] Step 91: The UPF reports an event to the PCF. The event may include the following information: the identification information of the service flow to be adjusted, such as the IP 5-tuple, flow description, and the service characteristics of the service flow to be adjusted. Alternatively, the event may include the following information: the identification information of the service flow to be adjusted and the QoS determined by the UPF through self-learning (for reference by the PCF when making decisions).
[0234] In this example, the UPF can also learn to directly switch the QoS of the service flow to be adjusted and report to the PCF that the QoS policy of the flow has been updated. In this case, it is no longer necessary to execute subsequent steps 92-93.
[0235] Step 92: PCF changes the QoS policy of the service flow to be adjusted based on the service characteristics of the service flow to be adjusted fed back by UPF, and sends the new QoS policy to SMF-UPF; or, PCF updates the QoS policy based on the QoS policy fed back by UPF and the service flow characteristics of the service flow to be adjusted.
[0236] Step 93: SMF updates the N4 rule to improve the QoS guarantee of the QoS flow. Various methods of embodiments of this disclosure have been described above. Apparatus for implementing the above methods will now be provided.
[0237] Please refer to Figure 10. This embodiment of the disclosure also provides a first network function, including:
[0238] The first receiving module 1001 is configured to receive first information sent by the second network function, the first information including at least one of the following:
[0239] First indication information used to indicate the first business flow;
[0240] Secondary indication information used to indicate the business characteristics that identify the first business flow.
[0241] Optionally, the first network function further includes:
[0242] The first sending module is used to send the second information to the third network function, or to send the third information to the fourth network function;
[0243] The second information includes at least one of the following: the first indication information, the second indication information, and model-related information; the third information includes the first indication information and the second indication information; the model-related information is either the first model or the third indication information, and the third indication information is used to indicate the provision of the fifth network function of the first model; the first model is used to identify the service characteristics of the service flow of the first service type, and the first service type is the service type of the first service flow.
[0244] Optionally, the first network function further includes:
[0245] The second sending module is used to send fourth information to the fifth network function before the first sending module sends the second information, the fourth information including the first service type.
[0246] Optionally, the first network function further includes:
[0247] The second receiving module is used to receive the fifth information sent by the fifth network function, the fifth information including the first model.
[0248] Optionally, the first network function further includes:
[0249] The third sending module is used to send sixth information for finding the fifth network function to the sixth network function when the model-related information is the third indication information. The sixth information includes at least one of the following: slice information, service area information, network function capability information, and AoI information.
[0250] Optionally, the first network function further includes:
[0251] The third receiving module is used to receive the seventh information sent by the sixth network function, the seventh information including the third indication information.
[0252] Optionally, the first network function further includes:
[0253] The fourth receiving module is used to receive the eighth information sent by the third network function. The eighth information includes at least one of the following: fourth indication information for indicating the service flow to be adjusted, the service type of the service flow to be adjusted, and the first QoS policy determined by the third network function; wherein the service flow to be adjusted is a service flow that requires adjustment of the QoS policy.
[0254] Optionally, the first network function further includes:
[0255] The fourth sending module is configured to send ninth information to the fourth network function and / or the third network function, the ninth information including at least one of the following: a second QoS policy determined by the first network function; the second QoS policy being applied to the QoS policy of switching the service flow to be adjusted.
[0256] Optionally, the first network function further includes:
[0257] The fifth receiving module is used to receive the tenth information sent by the third network function, the tenth information including: the fifth indication information that the third network function has switched the QoS policy of the first service flow.
[0258] It should be noted that the device in this embodiment corresponds to the method applied to the first network function side described above. The implementation methods in each of the above embodiments are applicable to the embodiments of this device and can achieve the same technical effect. The device provided in this disclosure can implement all the method steps implemented in the above method embodiments and can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail here.
[0259] Please refer to Figure 11. This embodiment of the disclosure also provides a third network function, including:
[0260] The first receiving module 1101 is configured to receive second information sent by a first network function or a fourth network function. The second information includes at least one of the following: first indication information for indicating a first service flow, second indication information for indicating service characteristics of the first service flow, and model-related information. The model-related information is either a first model or a third indication information. The third indication information is used to indicate a fifth network function that provides the first model. The first model is used to identify the service characteristics of a service flow of a first service type. The first service type is the service type of the first service flow.
[0261] Optionally, the third network function further includes:
[0262] The first identification module is used to identify the business characteristics of the first business flow using the first model.
[0263] Optionally, the third network function further includes:
[0264] The model acquisition module is used to perform at least one of the following:
[0265] If the model-related information is the first model, the first model can be obtained directly from the model-related information;
[0266] If the model-related information is the third indication information, send eleventh information to the fifth network function, the eleventh information including the first service type; receive twelfth information sent by the fifth network function, the twelfth information including the first model; obtain the first model from the twelfth information;
[0267] If the second information does not include the model-related information, a thirteenth piece of information for locating the fifth network function is sent to the sixth network function. The thirteenth piece of information includes at least one of the following: slice information, service area information, network function capability information, and AoI information. The fourteenth piece of information sent by the sixth network function is received, and the fourteenth piece of information includes the third indication information. A fifteenth piece of information is sent to the fifth network function, and the fifteenth piece of information includes the first service type. The sixteenth piece of information sent by the fifth network function is received, and the sixteenth piece of information includes the first model. The first model is obtained from the sixteenth piece of information.
[0268] Optionally, the service characteristics of the first service flow include at least one of the following: audio service flow, video service flow, local service flow, remote service flow, importance of the service flow, packet burst size, packet burst interval, frame interval, and frame rate.
[0269] Optionally, the first identification module is further configured to input the first service flow into the first model to obtain the service characteristics of the first service flow output by the first model.
[0270] Optionally, the third network function further includes:
[0271] The first determining module is used to determine whether the QoS policy of the first service flow needs to be adjusted based on the service characteristics of the first service flow.
[0272] Optionally, the third network function further includes:
[0273] A first sending module is configured to send eighth information to the first network function when it is determined that the QoS policy of the first service flow needs to be adjusted. The eighth information includes at least one of the following: fourth indication information for indicating the service flow to be adjusted, the service type of the service flow to be adjusted, and the first QoS policy determined by the third network function; wherein the service flow to be adjusted is the first service flow.
[0274] Optionally, the third network function further includes:
[0275] The second determining module is used to determine the first QoS policy when it is determined that the QoS policy of the first service flow needs to be adjusted.
[0276] Optionally, the third network function further includes:
[0277] The second receiving module is configured to receive the ninth information sent by the first network function, the ninth information including at least one of the following: a second QoS policy determined by the first network function; the second QoS policy being applied to the QoS policy for switching the service flow to be adjusted.
[0278] Optionally, the third network function further includes:
[0279] The first switching module is used to switch the QoS policy of the first service flow to the second QoS policy.
[0280] Optionally, the third network function further includes:
[0281] The second switching module is used to determine a first QoS policy and switch the QoS policy of the first service flow to the first QoS policy when it is determined that the QoS policy of the first service flow needs to be adjusted.
[0282] Optionally, the third network function further includes:
[0283] The second sending module is used to send tenth information to the first network function, the tenth information including: fifth indication information that the third network function has switched the QoS policy of the first service flow.
[0284] Optionally, the third network function further includes:
[0285] The third sending module is configured to send a seventeenth message to the fifth network function when the first model fails to identify the service characteristics of the first service flow. The seventeenth message includes at least one of the following: the first service type, the first model, and a sixth indication message. The sixth indication message indicates that the first model failed to identify the service characteristics of the service flow of the first service type.
[0286] Optionally, the third network function further includes:
[0287] The third receiving module is used to receive the eighteenth message sent by the fifth network function, the eighteenth message including the updated first model.
[0288] Optionally, the third network function further includes:
[0289] The second identification module is used to identify the business characteristics of the first business flow using the updated first model.
[0290] It should be noted that the device in this embodiment corresponds to the method applied to the third network function side described above. The implementation methods in each of the above embodiments are applicable to the embodiments of this device and can achieve the same technical effect. The device provided in this disclosure can implement all the method steps implemented in the above method embodiments and can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail here.
[0291] Please refer to Figure 12. This embodiment of the disclosure also provides a fifth network function, including:
[0292] The first receiving module 1201 is used to receive the twenty-fifth information sent by the first network function, the third network function or the fourth network function, wherein the twenty-fifth information includes the first service type.
[0293] The first sending module 1202 is used to send the 26th information to the first network function, the third network function, or the fourth network function, wherein the 26th information includes a first model, and the first model is a model corresponding to the first service type.
[0294] Optionally, the fifth network function further includes:
[0295] The maintenance module is used to maintain a mapping table between business types and models. The mapping table includes the mapping probabilities between various business types and various models.
[0296] Optionally, the fifth network function further includes:
[0297] The lookup module is used by the fifth network function to look up the mapping table and obtain the first model corresponding to the first service type, wherein the first model is the model with the highest mapping probability among all the models mapped to the first service type.
[0298] Optionally, the fifth network function further includes:
[0299] The update module is configured to generate a first model of business characteristics for identifying the business flow of the first business type through learning and reasoning when the first business type does not exist in the mapping table, and update the mapping table according to the first model and the first business type.
[0300] Optionally, the multiple models in the mapping table are multiple small models locally configured by the fifth network function. The small models are obtained by distillation training based on the large model in an external database, and the parameter specifications of the small models are smaller than those of the large models.
[0301] Optionally, the fifth network function further includes:
[0302] A training module is used to obtain the small model through distillation training based on a large model in an external database, wherein the distillation training includes:
[0303] For the first service type, obtain the first probability distribution between the multiple small models transmitted by the large model and the first service type, and obtain the first model corresponding to the first service type;
[0304] Based on the weight parameters, the distillation loss of the first model and the supervision loss of the first model in local supervised learning are weighted and summed to obtain the comprehensive loss;
[0305] The first business type training samples are input into the large model and the first model. By comparing the recognition results of the large model and the multiple small models and adjusting the weight parameters, the overall loss is minimized, and the trained first model is obtained.
[0306] It should be noted that the device in this embodiment corresponds to the method applied to the fifth network function side described above. The implementation methods in each of the above embodiments are applicable to the embodiments of this device and can achieve the same technical effect. The device provided in this disclosure can implement all the method steps implemented in the above method embodiments and can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail here.
[0307] Other embodiments of this disclosure also provide a user plane service analysis method, performed by a third network function. As shown in FIG13, the method includes:
[0308] Step 131: The third network function determines the third QoS policy for the first service flow.
[0309] Step 132: The third network function configures the third QoS policy for the first service flow.
[0310] As one implementation, in step 131 above, the third network function determines the third QoS policy for the first service flow, specifically including:
[0311] Step a, the third network function determines the first service type of the first service flow.
[0312] Here, the third network function determines the first service type of the first service flow based on at least one of the following information and / or models: a pre-configured strategy, a locally trained model, a model obtained from the fifth network function, the flow description information of the service flow, application detection information, application ID information, etc.
[0313] Step b: The third network function obtains the second model from the fifth network function, and uses the second model to match and obtain the third QoS policy for the first service type.
[0314] As another implementation, in step 131 above, the third network function determines the third QoS policy of the first service flow, specifically including:
[0315] Step A: The third network function sends the nineteenth information to the fifth network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate.
[0316] Step B: The third network function receives the twentieth information sent by the fifth network function, the twentieth information including the third QoS policy of the first service flow.
[0317] In this embodiment of the disclosure, the third network function may further send a twenty-first message to the fourth network function. The twenty-first message is used to notify the fourth network function to configure the third QoS policy for the first service flow in the radio access network. The fourth network function configures the third QoS policy for the radio access network (RAN) according to the twenty-first message.
[0318] Furthermore, in this embodiment of the disclosure, when the third network function detects a change in the service characteristics of the first service flow, it determines a fourth QoS policy based on the new service characteristics of the first service flow after the change. Then, the third network function executes the fourth QoS policy for the first service flow and sends twenty-second information to the first network function. The twenty-second information includes at least one of the following: the new service characteristics of the first service flow and the fourth QoS policy.
[0319] The first network function sends a twenty-sixth message to the third network function, instructing the third network function to first execute the fourth QoS policy. The first network function may send a request to the fifth network function, requesting the fifth network function to judge or verify the rationality of the fourth QoS policy, and, based on the judgment / verification result returned by the fifth network function, send a twenty-third or twenty-fourth message to the first network function. For example, if the judgment / verification result indicates that the fourth QoS policy is reasonable, the twenty-third message is sent to the first network function; if the judgment / verification result indicates that the fourth QoS policy is unreasonable, the twenty-fourth message is sent to the first network function.
[0320] The third network function receives either the twenty-third or twenty-fourth message sent by the first network function, wherein the twenty-third message instructs the third network function to execute the fourth QoS policy, and the twenty-fourth message instructs the third network function to execute the third QoS policy. Then, the third network function, based on the twenty-third message, continues to execute the fourth QoS policy for the first service flow, or, based on the twenty-fourth message, executes the third QoS policy for the first service flow.
[0321] Other embodiments of this disclosure also provide a user plane service analysis method, performed by a fifth network function. As shown in FIG14, the method includes:
[0322] Step 141: The fifth network function sends the second model to the third network function. The second model is used to match the QoS policy corresponding to the service type.
[0323] Other embodiments of this disclosure also provide a user plane service analysis method, performed by a fifth network function. As shown in Figure 15, the method includes:
[0324] Step 151: The fifth network function receives the nineteenth information sent by the third network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate.
[0325] Step 152: The fifth network function analyzes the information included in the nineteenth information to obtain the analysis result; and determines the third QoS strategy for the first service flow based on the analysis result.
[0326] Step 153: The fifth network function sends twentieth information to the third network function, the twentieth information including the third QoS policy of the first service flow.
[0327] The methods of the embodiments of this disclosure have been described above from the perspectives of the third network function and the fifth network function.
[0328] The above method will be further explained below through more specific interaction examples between network functions. It should be noted that in the following examples, the first network function is PCF, the second network function is AF, the third network function is UPF, the fourth network function is SMF, the fifth network function is NDWAF, the sixth network function is NRF, and the seventh network function is NEF.
[0329] Example 6:
[0330] As shown in Figure 16, this example specifically includes:
[0331] Step 161: UPF determines the type of service flow based on the following information: pre-configured strategy (e.g., which type of IP 5-tuple corresponds to which service type, and which 5QI is used), locally trained model (e.g., an internal model is trained based on the previous correspondence between service types and IP 5-tuples / service features, which can determine the service type corresponding to a new IP 5-tuple / service feature), model obtained from NWDAF, service flow description, application detection, application ID, etc.
[0332] Step 162: The UPF obtains the model from the NWDAF, matches the QoS guarantee to the service flow type, and sets the QFI or 5QI value for the service flow. In this article, QoS guarantee and QoS policy have the same meaning.
[0333] Step 163: The UPF sends the QoS guarantee (the value of QFI or 5QI) to the SMF, notifying the SMF to configure a new QoS for the service flow.
[0334] Steps 164a-164b: The SMF verifies with the NWDAF whether the QoS guarantees recommended by the UPF are reasonable. If reasonable, the SMF configures new QoS enforcement for the RAN.
[0335] Step 165: The UPF notifies the PCF that a new flow has been discovered and its locally determined QoS guarantee. The new flow refers to a flow whose service characteristics have changed. In this case, the UPF re-determines a new QoS guarantee for the new flow.
[0336] Step 166: PCF determines whether the QoS guarantee sent by UPF is reasonable, or verifies it with NWDAF.
[0337] In step 167, the PCF replies to the UPF that it needs to verify the service type. The UPF then performs QoS guarantees according to the service type it determines.
[0338] Step 168: If the verification is correct, the PCF replies to the UPF to execute the new QoS guarantee and issues the new QoS policy to the RAN.
[0339] Step 169: If the verification is incorrect, PCF replies to UPF to perform the original QoS guarantee.
[0340] Example 7:
[0341] As shown in Figure 17, this example specifically includes:
[0342] Step 170: The data provider (such as UPF, RAN) provides data to the NWDAF, including: UPF open latency, packet loss rate, QoS, burst data streams, malicious packets, duplicate packets, bandwidth utilization, jitter, packet arrival interval, importance of service flow, packet burst size, packet burst interval, frame interval, frame rate, and other information. And / or, the RAN sends QoS monitoring information to the NWDAF. (In some cases, the UPF and NWDAF may be co-located, with the UPF internally calling the NWDAF's services and providing it with the aforementioned data.)
[0343] Step 171: UPF requests / subscribes to data analysis from NWDAF, which may include the following information: latency, packet loss rate, QoS, burst data stream, malicious packets, duplicate packets, bandwidth utilization, jitter, packet arrival interval, importance of service flow, packet burst size, packet burst interval, frame interval, frame rate, etc.
[0344] Step 172: NWDAF analyzes this data and provides analysis results, such as: service characteristics of the service flows received by the UPF in the next ten minutes (e.g., changes in bandwidth and traffic); and selects the optimal QoS for the UPF from the alternative QoS options.
[0345] Step 173, NWDAF responds to UPF analysis results.
[0346] Step 174: UPF dynamically adjusts its QoS or QoS type (e.g., the type of PDU set) based on the NWDAF analysis results.
[0347] Step 175, optionally, the UPF reports the adjusted QoS to the SMF, and the SMF issues a new QoS enforcement to the RAN.
[0348] Step 176, optionally, the UPF notifies the PCF of the discovery of a new flow and its locally determined QoS guarantee. The new flow refers to a flow whose service characteristics have changed. In this case, the UPF re-determines a new QoS guarantee for the service flow.
[0349] Step 177: PCF determines whether the QoS guarantee sent by UPF is reasonable, or verifies it with NWDAF.
[0350] Step 178: If the verification is correct, the PCF replies to the UPF to execute the new QoS guarantee and issue the new QoS policy.
[0351] Step 179: If the verification is incorrect, PCF replies to UPF to perform the original QoS guarantee.
[0352] Please refer to Figure 18. One embodiment of this disclosure also provides a third network function, including:
[0353] The determination module is used to determine the third QoS policy for the first service flow;
[0354] The configuration module is used to configure the third QoS policy for the first service flow by the third network function.
[0355] Optionally, the determining module is further configured to determine a first service type of the first service flow; obtain a second model from the fifth network function; and use the second model to match and obtain a third QoS policy for the first service type.
[0356] Optionally, the determining module is further configured to determine the first service type of the first service flow based on at least one of the following information and / or models: a pre-configured strategy, a locally trained model, a model obtained from the fifth network function, the flow description information of the service flow, the application detection information, and the application ID information.
[0357] Optionally, the determining module is further configured to send a nineteenth piece of information to the fifth network function, the nineteenth piece of information including at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate; and receive a twentieth piece of information sent by the fifth network function, the twentieth piece of information including a third QoS policy for the first service flow.
[0358] Optionally, the third network function further includes:
[0359] The sending module is used to send twenty-first information to the fourth network function, the twenty-first information being used to notify the fourth network function to configure the third QoS policy for the first service flow in the radio access network.
[0360] Optionally, the third network function further includes:
[0361] The update module is configured to: upon detecting a change in the service characteristics of the first service flow, determine a fourth QoS policy based on the new service characteristics of the first service flow; execute the fourth QoS policy for the first service flow and send a twenty-second message to the first network function, the twenty-second message including at least one of the following: the new service characteristics of the first service flow and the fourth QoS policy; receive a twenty-third message or a twenty-fourth message sent by the first network function, wherein the twenty-third message is used to instruct the third network function to execute the fourth QoS policy, and the twenty-fourth message is used to instruct the third network function to execute the third QoS policy; continue to execute the fourth QoS policy for the first service flow according to the twenty-third message, or execute the third QoS policy for the first service flow according to the twenty-fourth message.
[0362] It should be noted that the device in this embodiment corresponds to the method applied to the third network function side described above. The implementation methods in each of the above embodiments are applicable to the embodiments of this device and can achieve the same technical effect. The device provided in this disclosure can implement all the method steps implemented in the above method embodiments and can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail here.
[0363] Please refer to Figure 19. One embodiment of this disclosure also provides a fifth network function, including:
[0364] The second sending module is used to send the second model to the third network function, and the second model is used to match the QoS policy corresponding to the service type.
[0365] Please refer to Figure 20. Another embodiment of this disclosure also provides a fifth network function, including:
[0366] The second receiving module is used to receive the nineteenth information sent by the third network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate.
[0367] The analysis module is used to analyze the information included in the nineteenth piece of information to obtain analysis results; and to determine the third QoS strategy for the first service flow based on the analysis results.
[0368] The third sending module is used to send twentieth information to the third network function, the twentieth information including the third QoS policy of the first service flow.
[0369] It should be noted that the device in this embodiment corresponds to the method applied to the fifth network function side described above. The implementation methods in each of the above embodiments are applicable to the embodiments of this device and can achieve the same technical effect. The device provided in this disclosure can implement all the method steps implemented in the above method embodiments and can achieve the same technical effect. Therefore, the parts that are the same as those in the method embodiments and the beneficial effects will not be described in detail here.
[0370] Another embodiment of the network function disclosed herein may specifically be the first network function, the third network function, or the fifth network function described above, as shown in FIG21. It includes a transceiver 1310, a processor 1300, a memory 1320, and a program or instructions stored in the memory 1320 and executable on the processor 1300. When the processor 1300 executes the program or instructions, it implements the various processes of the above-described method embodiment on the network device side and can achieve the same technical effect. To avoid repetition, it will not be described again here.
[0371] The transceiver 1310 is used to receive and send data under the control of the processor 1300.
[0372] In Figure 21, the bus architecture may include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 1300 and memory represented by memory 1320. The bus architecture may 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. Transceiver 1310 may be multiple elements, including transmitters and receivers, providing units for communicating with various other devices over a transmission medium. Processor 1300 is responsible for managing the bus architecture and general processing, and memory 1320 may store data used by processor 1300 during operation.
[0373] This disclosure also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the above-described user plane service flow analysis method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.
[0374] This disclosure also provides a computer program product, including computer instructions. When the computer instructions are executed by a processor, they implement the various processes of the above-described user plane service flow analysis method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.
[0375] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0376] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this disclosure.
[0377] The embodiments of this disclosure have been described above with reference to the accompanying drawings. However, this disclosure is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this disclosure without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this disclosure.
Claims
1. A user plane business flow analysis method, comprising: The first network function receives first information sent by the second network function, the first information including at least one of the following: First indication information used to indicate the first business flow; Secondary indication information used to indicate the business characteristics that identify the first business flow.
2. The method of claim 1, further comprising at least one of the following: The first network function sends second information to the third network function, wherein, The second information includes at least one of the following: the first indication information, the second indication information, and model-related information; The first network function sends third information to the fourth network function, wherein the third information includes the first indication information and the second indication information.
3. The method according to claim 2, wherein, The model-related information is either a first model or a third indication information, wherein the third indication information is used to indicate the provision of a fifth network function for the first model; The first model is used to identify the business characteristics of the business flow of the first business type, where the first business type is the business type of the first business flow.
4. The method according to claim 2, further comprising: Send fourth information to the fifth network function, the fourth information including the first service type.
5. The method according to claim 4, further comprising: The system receives fifth information sent by the fifth network function, the fifth information including the first model.
6. The method according to claim 3, wherein, When the model-related information is the third indication information, the method further includes: Send sixth information to the sixth network function for locating the fifth network function. The sixth information includes at least one of the following: slice information, service area information, network function capability information, and area of interest (AoI) information.
7. The method according to claim 6, further comprising: Receive the seventh information sent by the sixth network function, the seventh information including the third indication information.
8. The method according to claim 1, further comprising: The system receives an eighth message sent by a third network function, the eighth message including at least one of the following: a fourth indication message for indicating a service flow to be adjusted, the service type of the service flow to be adjusted, and a first quality of service (QoS) policy determined by the third network function; wherein the service flow to be adjusted is a service flow for which the QoS policy needs to be adjusted.
9. The method according to claim 8, further comprising: Send a ninth message to the fourth network function and / or the third network function, the ninth message including at least one of the following: a second QoS policy determined by the first network function; the second QoS policy being applied to the QoS policy of switching the service flow to be adjusted.
10. The method according to claim 1, further comprising: The system receives a tenth message sent by a third network function, the tenth message including a fifth indication that the third network function has switched the QoS policy of the first service flow.
11. A user plane business flow analysis method, comprising: The third network function receives second information sent by the first network function or the fourth network function. The second information includes at least one of the following: first indication information for indicating a first service flow, second indication information for indicating the service characteristics of the first service flow, and model-related information; the model-related information is a first model or a third indication information, and the third indication information is used to indicate a fifth network function that provides the first model; the first model is used to identify the service characteristics of a service flow of a first service type, and the first service type is the service type of the first service flow.
12. The method of claim 11, further comprising: The business characteristics of the first business flow are identified using the first model.
13. The method of claim 11, further comprising at least one of the following: If the model-related information is the first model, the first model can be obtained directly from the model-related information; If the model-related information is the third indication information, send eleventh information to the fifth network function, the eleventh information including the first service type; receive twelfth information sent by the fifth network function, the twelfth information including the first model; obtain the first model from the twelfth information; If the second information does not include the model-related information, a thirteenth piece of information for locating the fifth network function is sent to the sixth network function. The thirteenth piece of information includes at least one of the following: slice information, service area information, network function capability information, and AoI information. The fourteenth piece of information sent by the sixth network function is received, and the fourteenth piece of information includes the third indication information. A fifteenth piece of information is sent to the fifth network function, and the fifteenth piece of information includes the first service type. The sixteenth piece of information sent by the fifth network function is received, and the sixteenth piece of information includes the first model. The first model is obtained from the sixteenth piece of information.
14. The method according to claim 11, wherein, The service characteristics of the first service flow include at least one of the following: audio service flow, video service flow, local service flow, remote service flow, importance of the service flow, packet burst size, packet burst interval, frame interval, and frame rate.
15. The method according to claim 11, wherein, The business characteristics of the first business flow are identified using the first model, including: The first business flow is input into the first model to obtain the business characteristics of the first business flow output by the first model.
16. The method of claim 11, further comprising: Based on the service characteristics of the first service flow, determine whether the QoS policy of the first service flow needs to be adjusted.
17. The method of claim 16, further comprising: If it is determined that the QoS policy of the first service flow needs to be adjusted, an eighth message is sent to the first network function. The eighth message includes at least one of the following: a fourth indication message for indicating the service flow to be adjusted, the service type of the service flow to be adjusted, and the first QoS policy determined by the third network function; wherein the service flow to be adjusted is the first service flow.
18. The method of claim 17, further comprising: If it is determined that the QoS policy of the first service flow needs to be adjusted, a first QoS policy is determined.
19. The method of claim 17, further comprising: Receive the ninth information sent by the first network function, the ninth information including at least one of the following: the second QoS policy determined by the first network function; The second QoS policy is applied to switch the QoS policy of the service flow to be adjusted.
20. The method of claim 19, further comprising: Switch the QoS policy of the first service flow to the second QoS policy.
21. The method of claim 16, further comprising: If it is determined that the QoS policy of the first service flow needs to be adjusted, a first QoS policy is determined, and the QoS policy of the first service flow is switched to the first QoS policy.
22. The method of claim 21, further comprising: The tenth message sent to the first network function includes a fifth indication that the third network function has switched the QoS policy of the first service flow.
23. The method of claim 11, further comprising: If the service characteristics of the first service flow cannot be identified using the first model, the seventeenth information is sent to the fifth network function. The seventeenth information includes at least one of the following: the first service type, the first model, and the sixth indication information. The sixth indication information indicates that the first model failed to identify the business characteristics of the business flow of the first business type.
24. The method of claim 23, further comprising: Receive the eighteenth message sent by the fifth network function, the eighteenth message including the updated first model.
25. The method of claim 24, further comprising: The updated first model is used to identify the business characteristics of the first business flow.
26. A user plane business analysis method, comprising: The fifth network function receives the twenty-fifth information sent by the first network function, the third network function, or the fourth network function, and the twenty-fifth information includes the first service type.
27. The method of claim 26, further comprising: The fifth network function sends a twenty-sixth message to the first network function, the third network function, or the fourth network function. The twenty-sixth message includes a first model, which is a model corresponding to the first service type.
28. The method of claim 26, further comprising: The fifth network function maintains a mapping table between service types and models, and the mapping table includes mapping probabilities between various service types and various models.
29. The method of claim 28, further comprising: The fifth network function searches the mapping table to obtain the first model corresponding to the first service type, wherein the first model is the model with the highest mapping probability among all the models mapped to the first service type.
30. The method of claim 28, further comprising: If the first business type does not exist in the mapping table, a first model for identifying the business characteristics of the business flow of the first business type is generated through learning and reasoning, and the mapping table is updated according to the first model and the first business type.
31. The method according to claim 28, wherein, The multiple models in the mapping table are multiple small models locally configured by the fifth network function. The small models are obtained by distillation training based on the large model in the external database, and the parameter specifications of the small models are smaller than those of the large models.
32. The method according to claim 29, wherein, The distillation training includes: For the first service type, obtain the first probability distribution between the multiple small models transmitted by the large model and the first service type, and obtain the first model corresponding to the first service type; Based on the weight parameters, the distillation loss of the first model and the supervision loss of the first model in local supervised learning are weighted and summed to obtain the comprehensive loss; The first business type training samples are input into the large model and the first model. By comparing the recognition results of the large model and the multiple small models and adjusting the weight parameters, the overall loss is minimized, and the trained first model is obtained.
33. A first network function, comprising: A first receiving module is configured to receive first information sent by a second network function, the first information including at least one of the following: First indication information used to indicate the first business flow; Secondary indication information used to indicate the business characteristics that identify the first business flow.
34. A third network function, comprising: The first receiving module is configured to receive second information sent by a first network function or a fourth network function. The second information includes at least one of the following: first indication information for indicating a first service flow, second indication information for indicating service characteristics of the first service flow, and model-related information; the model-related information is a first model or a third indication information, the third indication information being used to indicate a fifth network function that provides the first model; the first model is used to identify service characteristics of a service flow of a first service type, the first service type being the service type of the first service flow.
35. A fifth network function, comprising: The first receiving module is used to receive the twenty-fifth information sent by the first network function, the third network function or the fourth network function, wherein the twenty-fifth information includes the first service type. The first sending module is used to send the 26th information to the first network function, the third network function, or the fourth network function. The 26th information includes a first model, which is a model corresponding to the first service type.
36. A user-plane business analysis method, comprising: The third network function determines the third QoS policy for the first service flow; The third network function configures the third QoS policy for the first service flow.
37. The method of claim 36, wherein, The third network function determines a third QoS policy for the first service flow, including at least one of the following: The third network function determines the first service type of the first service flow; The third network function obtains the second model from the fifth network function, and uses the second model to match and obtain the third QoS policy for the first service type.
38. The method according to claim 37, wherein, The third network function determines the first service type of the first service flow, including: The third network function determines the first service type of the first service flow based on at least one of the following information and / or models: a pre-configured strategy, a locally trained model, a model obtained from the fifth network function, the flow description information of the service flow, the application detection information, and the application ID information.
39. The method according to claim 36, wherein, The third network function determines the third QoS policy for the first service flow, including: The third network function sends the nineteenth information to the fifth network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate. The third network function receives the twentieth information sent by the fifth network function, the twentieth information including the third QoS policy of the first service flow.
40. The method of claim 36, further comprising: The third network function sends a twenty-first message to the fourth network function, the twenty-first message being used to notify the fourth network function to configure the third QoS policy for the first service flow in the radio access network.
41. The method of claim 36, further comprising at least one of the following: When the third network function detects a change in the service characteristics of the first service flow, it determines a fourth QoS policy based on the new service characteristics of the first service flow. The third network function executes the fourth QoS policy for the first service flow and sends twenty-second information to the first network function. The twenty-second information includes at least one of the following: new service characteristics of the first service flow and the fourth QoS policy. The third network function receives the twenty-third or twenty-fourth message sent by the first network function, wherein... The 23rd information is used to instruct the third network function to execute the fourth QoS policy, and the 24th information is used to instruct the third network function to execute the third QoS policy; The third network function may, based on the second-third information, continue to execute the fourth QoS policy for the first service flow, or, based on the second-fourth information, execute the third QoS policy for the first service flow.
42. A user plane business analysis method, comprising: The fifth network function sends a second model to the third network function, which is used to match the QoS policy corresponding to the service type.
43. The method according to any one of claims 1 to 32 or claims 36 to 42, wherein, The first network function includes Policy Control Function (PCF), the second network function includes Application Function (AF), the third network function includes User Plane Function (UPF), the fourth network function includes Session Management Function (SMF), the fifth network function includes Network Data Analysis Function (NDWAF), and the sixth network function includes Network Storage Function (NRF).
44. A user-plane business analysis method, comprising: The fifth network function receives the nineteenth information sent by the third network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate. The fifth network function analyzes the information included in the nineteenth information to obtain the analysis result; and determines the third QoS strategy for the first service flow based on the analysis result. The fifth network function sends twentieth information to the third network function, the twentieth information including the third QoS policy of the first service flow.
45. A third network function, comprising: The determination module is used to determine the third QoS policy for the first service flow; The configuration module is used to configure the third QoS policy for the first service flow by the third network function.
46. A fifth network function, comprising: The second sending module is used to send the second model to the third network function, and the second model is used to match the QoS policy corresponding to the service type.
47. A fifth network function, comprising: The second receiving module is used to receive the nineteenth information sent by the third network function. The nineteenth information includes at least one of the following: latency, packet loss rate, fourth QoS policy, burst data stream, malicious data packet, duplicate data packet, bandwidth utilization, latency jitter, data packet arrival interval, importance of service flow, data packet burst size, data packet burst interval, frame interval, and frame rate. The analysis module is used to analyze the information included in the nineteenth piece of information to obtain analysis results; and to determine the third QoS strategy for the first service flow based on the analysis results. The third sending module is used to send twentieth information to the third network function, the twentieth information including the third QoS policy of the first service flow.
48. A network function, comprising: Transceiver, processor, memory, and programs or instructions stored in the memory and executable on the processor; When the processor executes the program or instructions, it implements the steps of the method as claimed in any one of claims 1 to 32 or 36 to 43.
49. A computer-readable storage medium having a computer program stored thereon, the computer program, when executed by a processor, implementing the steps of the method as claimed in any one of claims 1 to 32 or 36 to 43.
50. A computer program product comprising computer instructions that, when executed by a processor, implement the steps of the method as claimed in any one of claims 1 to 32 or 36 to 43.
