QoS control strategy generation method and device, communication device and storage medium

Abstract

The invention relates to a QoS control strategy generation method and device, a communication device and a storage medium. According to the method, the first network performance analysis result and the first user experience analysis result are received, and the QoS control strategy generation model is optimized by using the first network performance analysis result and the first user experience analysis result, so that the optimized QoS control strategy generation model considers the current actual network state and user experience; parameter values corresponding to the optimized QoS control strategy generation model are different according to different first network performance analysis results and different first user experience analysis results. The optimized QoS control strategy generation model can provide flexible and customized QoS control strategies according to network environments and user requirements, the QoS control strategy generation models optimized according to different network environments and user requirements are different, adaptive differentiated QoS control strategies can be provided for diversified service requirements, and differentiated QoS guarantee is achieved.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of communication technology, and in particular to a QoS control policy generation method, device, communication device, and storage medium. Background Art

[0002] With the development of communication technology, 6G networks and future network systems will integrate technologies such as artificial intelligence to provide intelligent services for diversified businesses. In response to the emergence of massive diversified business data flows, 6G networks need to provide differentiated Quality of Service (QoS) guarantees.

[0003] At present, the QoS control strategy of 5G networks is configured according to template parameters, and it is impossible to provide flexible and customized QoS control strategies based on the network environment and user needs. Therefore, it is impossible to provide adaptive and differentiated QoS guarantees for diversified business needs. Summary of the invention

[0004] At least one embodiment of the present disclosure provides a QoS control policy generation method, device, communication device, and storage medium.

[0005] In a first aspect, an embodiment of the present disclosure provides a QoS control policy generation method, which is applied to a QoS control policy generation function QPGF. The method includes:

[0006] The QPGF receives a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, where the first network performance analysis result is a result obtained by the NWDAF analyzing data generated by the network device executing the first QoS control policy, and the first user experience analysis result is a result obtained by the NWDAF analyzing data generated by the terminal executing the first QoS control policy;

[0007] QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result;

[0008] The QPGF generates a second QoS control policy based on the optimized QoS control policy generation model.

[0009] In some embodiments, the QoS control policy generation method further includes:

[0010] The QPGF sends the second QoS control policy to the policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

[0011] In some embodiments, the QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result, including:

[0012] The QPGF optimizes the parameters of the QoS control policy generation model corresponding to the first QoS control policy based on the first network performance analysis result and the first user experience analysis result;

[0013] The first QoS control policy is sent by a policy control function PCF to the network device and the terminal.

[0014] In some embodiments, the QoS control policy generation method further includes:

[0015] If the first QoS control policy is an initialized QoS control policy generated by the QPGF based on the QoS control policy generation model, the QPGF sends the first QoS control policy to the PCF.

[0016] In some embodiments, the first QoS control policy generated by the QPGF based on the QoS control policy generation model includes:

[0017] QPGF obtains the target network performance analysis results from NWDAF and the stored target QoS control policy from PCF;

[0018] The QPGF inputs the target network performance analysis result and the target QoS control policy into the QoS control policy generation model to obtain the first QoS control policy.

[0019] In some embodiments, the QPGF obtains the target network performance analysis result from the NWDAF including:

[0020] QPGF sends a network performance analysis request to NWDAF;

[0021] QPGF receives the network performance analysis response sent by NWDAF, which carries the target network performance analysis result; wherein the target network performance analysis result is obtained by NWDAF through training and reasoning of the network performance analysis model based on the network performance data.

[0022] In some embodiments, the QPGF sends the first QoS control policy to the PCF, including:

[0023] When the QPGF obtains the first QoS control policy, it sends a request for obtaining the terminal subscription data to the unified data management (UDM) network element.

[0024] QPGF receives the terminal subscription data response sent by UDM, which carries the terminal capability information;

[0025] When determining that the terminal capability information satisfies the first QoS control policy, the QPGF sends the first QoS control policy to the PCF.

[0026] In some embodiments, the QoS control policy generation model is a generative adversarial network model, and the generative adversarial network model includes a generator and a discriminator;

[0027] QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result, including:

[0028] Based on the first network performance analysis result and the first user experience analysis result, QPGF trains the model parameters corresponding to the generator and outputs the training results;

[0029] QPGF inputs the training result into the discriminator, and stops training the generator when it is determined that the discrimination result meets the preset training goal, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number;

[0030] The QPGF generates a second QoS control policy based on the optimized QoS control policy generation model, including:

[0031] The QPGF generates a second QoS control policy based on the trained generator.

[0032] In a second aspect, the embodiment of the present disclosure further proposes a QoS control policy generation method, which is applied to a policy control function PCF. The QoS control policy generation method includes:

[0033] The PCF receives a second QoS control policy sent by the QoS control policy generation function QPGF, where the second QoS control policy is a policy generated by the QoS control policy generation method according to any embodiment of the first aspect;

[0034] The PCF sends the second QoS control policy to the network device and the terminal respectively.

[0035] In some embodiments, the PCF sends a second QoS control policy to the network device and the terminal respectively, including:

[0036] The PCF sends a protocol data unit PDU session modification message to the session management function SMF, and the PDU session modification message carries the second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

[0037] In a third aspect, the embodiment of the present disclosure further proposes a QoS control policy generation device, which is applied to a QoS control policy generation function QPGF, and the device includes:

[0038] The first unit is used to receive a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, where the first network performance analysis result is a result obtained by the NWDAF analyzing data generated by the network device executing the first QoS control policy, and the first user experience analysis result is a result obtained by the NWDAF analyzing data generated by the terminal executing the first QoS control policy;

[0039] The second unit is used to optimize the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result;

[0040] The third unit is used to generate a second QoS control policy based on the optimized QoS control policy generation model.

[0041] In a fourth aspect, the embodiment of the present disclosure further proposes a QoS control policy generation device, which is applied to a policy control function PCF, and the device includes:

[0042] The first unit is used to receive a second QoS control policy sent by a QoS control policy generation function QPGF, where the second QoS control policy is a policy generated by the QoS control policy generation method according to any embodiment of the first aspect;

[0043] The second unit is used to send a second QoS control policy to the network device and the terminal respectively.

[0044] In a fifth aspect, an embodiment of the present disclosure further provides a communication device, including a memory, a transceiver, and a processor:

[0045] A memory for storing computer programs; a transceiver for transmitting and receiving data under the control of a processor; and a processor for reading the computer programs in the memory and executing:

[0046] receiving a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, wherein the first network performance analysis result is a result obtained by NWDAF analyzing data generated by a network device executing a first QoS control policy, and the first user experience analysis result is a result obtained by NWDAF analyzing data generated by a terminal executing the first QoS control policy;

[0047] Optimizing the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result;

[0048] Based on the optimized QoS control policy generation model, a second QoS control policy is generated.

[0049] In some embodiments, the processor is further configured to read a computer program in the memory and execute:

[0050] The second QoS control policy is sent to the policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

[0051] In some embodiments, based on the first network performance analysis result and the first user experience analysis result, optimizing the QoS control policy generation model includes:

[0052] Based on the first network performance analysis result and the first user experience analysis result, optimizing parameters of the QoS control policy generation model corresponding to the first QoS control policy;

[0053] The first QoS control policy is sent by a policy control function PCF to the network device and the terminal.

[0054] In some embodiments, the processor is further configured to read a computer program in the memory and execute:

[0055] If the first QoS control policy is an initialized QoS control policy generated by the processor based on the QoS control policy generation model, the first QoS control policy is sent to the PCF.

[0056] In some embodiments, the first QoS control policy generated based on the QoS control policy generation model includes:

[0057] Obtain target network performance analysis results from NWDAF and stored target QoS control policies from PCF;

[0058] The target network performance analysis result and the target QoS control policy are input into the QoS control policy generation model to obtain the first QoS control policy.

[0059] In some embodiments, obtaining the target network performance analysis result from the NWDAF includes:

[0060] Send a network performance analysis request to NWDAF;

[0061] A network performance analysis response sent by NWDAF is received, where the network performance analysis response carries a target network performance analysis result; wherein the target network performance analysis result is obtained by NWDAF through training and reasoning of a network performance analysis model based on network performance data.

[0062] In some embodiments, sending the first QoS control policy to the PCF includes:

[0063] When the first QoS control policy is obtained, a terminal subscription data acquisition request is sent to a unified data management (UDM) network element;

[0064] Receive the terminal subscription data response sent by UDM, which carries the terminal capability information;

[0065] When it is determined that the terminal capability information satisfies the first QoS control policy, the first QoS control policy is sent to the PCF.

[0066] In some embodiments, the QoS control policy generation model is a generative adversarial network model, and the generative adversarial network model includes a generator and a discriminator;

[0067] Based on the first network performance analysis result and the first user experience analysis result, the QoS control policy generation model is optimized, including:

[0068] Based on the first network performance analysis result and the first user experience analysis result, train the model parameters corresponding to the generator and output the training results;

[0069] Input the training result into the discriminator, and stop training the generator when it is determined that the discrimination result meets the preset training goal, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number;

[0070] Based on the optimized QoS control policy generation model, a second QoS control policy is generated, including:

[0071] Based on the trained generator, a second QoS control policy is generated.

[0072] In a sixth aspect, an embodiment of the present disclosure further provides a communication device, including a memory, a transceiver, and a processor:

[0073] A memory for storing computer programs; a transceiver for transmitting and receiving data under the control of a processor; and a processor for reading the computer programs in the memory and executing:

[0074] Receive a second QoS control policy sent by a QoS control policy generation function QPGF, where the second QoS control policy is a policy generated by the QoS control policy generation method according to any embodiment of the first aspect;

[0075] The second QoS control policy is sent to the network device and the terminal respectively.

[0076] In some embodiments, sending the second QoS control policy to the network device and the terminal respectively includes:

[0077] A protocol data unit PDU session modification message is sent to the session management function SMF, where the PDU session modification message carries the second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

[0078] In the seventh aspect, the embodiments of the present disclosure also propose a processor-readable storage medium, which stores a program, and the program is used to enable the processor to execute the QoS control policy generation method of any embodiment of the first aspect or the QoS control policy generation method of any embodiment of the second aspect.

[0079] It can be seen that in at least one embodiment of the present disclosure, by receiving the first network performance analysis result and the first user experience analysis result, the first network performance analysis result and the first user experience analysis result can be used to optimize the QoS control policy generation model, so that the optimized QoS control policy generation model takes into account the current actual network status and user experience. Different first network performance analysis results and different first user experience analysis results have different parameter values ​​corresponding to the optimized QoS control policy generation model. Therefore, the new QoS control policy (referred to as the second QoS control policy) generated by the optimized QoS control policy generation model can provide flexible and customized QoS control policies according to the network environment and user needs. Different network environments and user needs have different optimized QoS control policy generation models, which can provide adaptive differentiated QoS control policies for diversified business needs and achieve differentiated QoS guarantees. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure, and a person skilled in the art can also obtain other drawings based on these drawings.

[0081] Figure 1 A flowchart of a QoS control policy generation method provided by an embodiment of the present disclosure;

[0082] Figure 2 A flowchart of another QoS control policy generation method provided by an embodiment of the present disclosure;

[0083] Figure 3 A schematic diagram of a QPGF for implementing self-generation of QoS control policies provided in an embodiment of the present disclosure;

[0084] Figure 4 A schematic diagram of a QoS control policy generation process provided by an embodiment of the present disclosure;

[0085] Figure 5 An interactive schematic diagram of a QPGF generating an initialization QoS control policy provided by an embodiment of the present disclosure;

[0086] Figure 6An interactive schematic diagram of a QoS control policy generation process provided by an embodiment of the present disclosure;

[0087] Figure 7 A schematic diagram of a QoS control strategy generating device provided by an embodiment of the present disclosure;

[0088] Figure 8 A schematic diagram of another QoS control strategy generating device provided by an embodiment of the present disclosure;

[0089] Fig. 9 A schematic diagram of a communication device provided in an embodiment of the present disclosure;

[0090] Fig.10 A schematic diagram of another communication device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0091] In order to more clearly understand the above-mentioned purposes, features and advantages of the present disclosure, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the described embodiments are part of the embodiments of the present disclosure, rather than all of the embodiments. The specific embodiments described herein are only used to explain the present disclosure, rather than to limit the present disclosure. Based on the described embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in the field belong to the scope of protection of the present disclosure.

[0092] It should be noted that, in this document, relational terms such as “first” and “second” are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0093] QoS is a description of the quality of service transmission. Its purpose is to provide end-to-end service quality assurance for various services according to their different needs. 5G network introduces the concept of QoS flow, which is the smallest granularity of QoS processing. The identifier of QoS flow is QoS Flow ID (QFI), which is unique in a protocol data unit (PDU) session. QFI can be dynamically allocated by the network, or the same as 5QI (5G QoS identifier). The network side performs data packet forwarding based on QFI, and user plane data with the same QFI in the PDU session will receive the same forwarding processing.

[0094] Whether a QoS flow is guaranteed bit rate (GBR) or non-GBR depends on the QoS configuration; the QoS configuration of a QoS flow is sent to the Radio Access Network (RAN). The QoS configuration contains the following QoS parameters:

[0095] 1. The QoS configuration of each QoS flow will include the following QoS parameters: 5QI, ARP (Allocation and Retention Priority);

[0096] 2. The QoS configuration of each Non-GBR QoS flow may also include parameters: ReflectiveQoS Attribute (RQA);

[0097] 3. The QoS configuration of each GBR QoS flow will also include parameters: uplink and downlink guaranteed flow bit rate (GuaranteedFlow Bit Rate, GFBR), uplink and downlink maximum flow bit rate (MaximumFlowBitRate, MFBR);

[0098] 4. The QoS configuration of each GBR QoS flow may also include: indication control, maximum packet loss rate;

[0099] 5. The QoS configuration of each QoS flow has a corresponding QFI, which is included in the QoS configuration.

[0100] Each 5QI represents a set of 5G QoS characteristics, which describe how QoS flows should be treated in packet forwarding between the terminal (User Equipment, UE) and the user plane function (User Plane Function, UPF). 5G QoS characteristics include: resource type, priority, expected delay, bit error rate and other information.

[0101] Standardized 5QI values ​​are defined for frequently used services, so using standardized QoS features can optimize signaling. Services that do not define standardized 5QIs can use dynamically assigned 5QI values ​​(requires QoS features to be included in the QoS profile). There is a fixed mapping relationship between standardized 5QI and 5G QoS features.

[0102] The QoS control strategy mainly includes QoS parameter sets, QoS flow assurance measures, etc. Among them, the QoS parameter set includes: uplink and downlink guaranteed flow bit rate (GFBR), uplink and downlink maximum flow bit rate (MFBR), jitter range, maximum bit error rate, priority, packet delay budget, maximum data burst, semantic similarity and other parameter labels and corresponding data.

[0103] The existing QoS control policy adjustment configures QoS parameters based on a fixed mapping relationship from standardized 5QI to 5G QoS characteristics. It cannot provide flexible and customized QoS control policies based on network environment and user needs, and therefore cannot provide adaptive and differentiated QoS guarantees for diversified business needs.

[0104] At least one embodiment of the present disclosure discloses a QoS control policy generation method, device, communication device or storage medium. By receiving a first network performance analysis result and a first user experience analysis result, the first network performance analysis result and the first user experience analysis result can be used to optimize the QoS control policy generation model, so that the optimized QoS control policy generation model takes into account the current actual network status and user experience. Different first network performance analysis results and different first user experience analysis results have different parameter values ​​corresponding to the optimized QoS control policy generation model. Therefore, the new QoS control policy (referred to as the second QoS control policy) generated by the optimized QoS control policy generation model can provide flexible and customized QoS control policies according to network environments and user needs. Different QoS control policy generation models are optimized for different network environments and user needs, which can provide adaptive differentiated QoS control policies for diversified business needs and achieve differentiated QoS guarantees.

[0105] Figure 1 The present invention provides a flow chart of a QoS control policy generation method provided in an embodiment of the present invention, wherein the QoS control policy generation method is applied to a QoS control policy generation function (QoS Policy Generation Function, QPGF). Figure 1 As shown, the QoS control policy generation method may include but is not limited to the following steps 101 to 103:

[0106] In step 101, QPGF receives a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, wherein the first network performance analysis result is a result obtained by NWDAF analyzing data generated by a network device executing a first QoS control policy, and the first user experience analysis result is a result obtained by NWDAF analyzing data generated by a terminal executing a first QoS control policy.

[0107] In this embodiment, the Policy Control Function (PCF) sends the first QoS control policy to the network device and the terminal to implement the issuance of the first QoS control policy. The issuance method follows the conventional issuance method in this field and will not be repeated here. The first QoS control policy can be understood as the initialization QoS control policy. At present, all QoS control policies, including the first QoS control policy, are generated by the PCF. The generation method is based on a fixed mapping relationship from standardized 5QI to 5G QoS characteristics or based on manual configuration. It is not combined with the current actual network status and user experience. The flexibility of QoS parameter configuration is poor, and it is impossible to provide flexible and customized QoS control strategies according to the network environment and user needs.

[0108] Therefore, in order to realize the intelligent generation of QoS control policies, a new network function is added in this embodiment: QoS Policy Generation Function (QPGF). QPGF can obtain the current actual network status and user experience, and intelligently generate QoS control policies. For example, in this embodiment, QPGF receives a first network performance analysis result and a first user experience analysis result, wherein the first network performance analysis result reflects the current actual network status, and the first user experience analysis result reflects the current actual user experience.

[0109] In this embodiment, the Network Data Analytics Function (NWDAF) can collect network performance data and terminal data, wherein the network performance data is the data generated by the network device executing the first QoS control strategy, and the terminal data is the data generated by the terminal executing the first QoS control strategy. Furthermore, NWDAF analyzes the network performance data to obtain the first network performance analysis result, and NWDAF analyzes the terminal data to obtain the first user experience analysis result. Thus, NWDAF can send the first network performance analysis result and the first user experience analysis result to QPGF. In some embodiments, NWDAF can actively send the first network performance analysis result and the first user experience analysis result. In other embodiments, after receiving the analysis result request sent by QPGF, NWDAF sends the first network performance analysis result and the first user experience analysis result to QPGF as a response.

[0110] In step 102, the QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result.

[0111] In this embodiment, QPGF includes a QoS control policy generation model, the QoS control policy generation model is a neural network model, the output of the QoS control policy generation model is the QoS control policy, and the input of the QoS control policy generation model is the first network performance analysis result and the first user experience analysis result.

[0112] In step 103, the QPGF generates a second QoS control policy based on the optimized QoS control policy generation model.

[0113] It can be seen that in this embodiment, the QoS control policy generation model is optimized using the first network performance analysis result and the first user experience analysis result, so that the optimized QoS control policy generation model takes into account the current actual network status and user experience. Different first network performance analysis results and different first user experience analysis results have different parameter values ​​corresponding to the optimized QoS control policy generation model. Therefore, the new QoS control policy (referred to as the second QoS control policy) generated by the optimized QoS control policy generation model can provide flexible and customized QoS control policies according to network environment and user needs. Different QoS control policy generation models are optimized for different network environments and user needs, which can provide adaptive differentiated QoS control policies for diversified business needs and achieve differentiated QoS guarantees.

[0114] On the basis of the above embodiment, after the second QoS control policy is generated in step 103, the QoS control policy generation method further includes the following steps:

[0115] The QPGF sends the second QoS control policy to the policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

[0116] It can be seen that, unlike the prior art, after the PCF provides the initial QoS control policy, the subsequent newly generated QoS control policy is no longer generated by the PCF, but by the QPGF, and the QPGF provides the new QoS control policy to the PCF. The same as the prior art is that the PCF still sends the new QoS control policy to the network equipment and terminals.

[0117] In some embodiments, after QPGF sends the second QoS control policy to PCF, QPGF receives the second network performance analysis result and the second user experience analysis result generated by NWDAF, the second network performance analysis result is the result of NWDAF analyzing the data generated by the network device executing the second QoS control policy, and the second user experience analysis result is the result of NWDAF analyzing the data generated by the terminal executing the second QoS control policy; further, QPGF optimizes the optimized QoS control policy generation model again based on the second network performance analysis result and the second user experience analysis result; thus, QPGF generates a third QoS control policy based on the re-optimized QoS control policy generation model. QPGF sends the third QoS control policy to PCF, so that PCF sends the third QoS control policy to the network device and the terminal. By analogy, continuous optimization of the QoS control policy is achieved until the feedback user experience reaches the best and the best QoS control policy is output.

[0118] On the basis of the above embodiment, in step 102, the QPGF optimizes the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result, specifically:

[0119] The QPGF optimizes parameters of a QoS control policy generation model corresponding to the first QoS control policy based on the first network performance analysis result and the first user experience analysis result.

[0120] In this embodiment, the QoS control policy generation model corresponding to the first QoS control policy can be understood as the QoS control policy generation model before optimization, and the first network performance analysis result and the first user experience analysis result used in this optimization are related to the first QoS control policy, that is, the first network performance analysis result is the result of analyzing the data generated by the network device executing the first QoS control policy, and the first user experience analysis result is the result of analyzing the data generated by the terminal executing the first QoS control policy.

[0121] In some embodiments, if the first QoS control policy is an initialized QoS control policy generated by the QPGF based on the QoS control policy generation model, the QPGF sends the first QoS control policy to the PCF, and then the PCF sends the first QoS control policy to the network device and the terminal.

[0122] In some embodiments, the QPGF sends the first QoS control policy to the PCF, specifically:

[0123] When the QPGF obtains the first QoS control policy, it sends a terminal subscription data acquisition request to the Unified Data Management (UDM) network element; then, the QPGF receives the terminal subscription data response sent by the UDM, and the terminal subscription data response carries the terminal capability information; thus, when the QPGF determines that the terminal capability information meets the first QoS control policy, the QPGF sends the first QoS control policy to the PCF.

[0124] On the basis of the above embodiment, the first QoS control policy generated by the QPGF based on the QoS control policy generation model is specifically:

[0125] QPGF obtains the target network performance analysis result from NWDAF and the stored target QoS control policy from PCF; then, QPGF inputs the target network performance analysis result and the target QoS control policy into the QoS control policy generation model to obtain the first QoS control policy.

[0126] In this embodiment, the target network performance analysis result is a result obtained by NWDAF analyzing data generated by the network device executing the target QoS control policy.

[0127] In this embodiment, QPGF obtains the target network performance analysis result from NWDAF, specifically: QPGF sends a network performance analysis request to NWDAF; then, QPGF receives a network performance analysis response sent by NWDAF, and the network performance analysis response carries the target network performance analysis result; wherein, the target network performance analysis result is obtained by NWDAF through training and reasoning of a network performance analysis model based on network performance data.

[0128] In some embodiments, NWDAF includes an analytics logical function (AnLF) and a model training logical function (MTLF). QPGF obtains the target network performance analysis result from NWDAF, specifically: QPGF sends a network performance analysis request to AnLF; then, QPGF receives a network performance analysis response sent by AnLF, and the network performance analysis response carries the target network performance analysis result; wherein, the target network performance analysis result is the target network performance analysis result obtained by MTLF training and reasoning the network performance analysis model based on the target network performance data carried in the model training requirement message after MTLF receives the model training requirement message sent by AnLF. wherein, the target network performance data is the data generated by the network device executing the target QoS control policy.

[0129] Based on the above embodiment, the QoS control policy generation model is a Generative Adversarial Network (GAN) model, which includes a generator and a discriminator. Accordingly, in step 102, "QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result", specifically:

[0130] Based on the first network performance analysis result and the first user experience analysis result, QPGF trains the model parameters corresponding to the generator and outputs the training results. Then, the training results are input into the discriminator, and when it is determined that the discrimination result meets the preset training goal, the training of the generator is stopped, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number.

[0131] For example, based on the first network performance analysis results and the first user experience analysis results, QPGF jointly trains the generator and the discriminator (adjusts the model parameters of the generator and the model parameters of the discriminator) until the preset training goals of the generative adversarial network model are reached and the joint training is stopped to obtain an optimized generator and an optimized discriminator; wherein the training goals include: the discrimination result of the discriminator converges to 0.5 or the number of iterations reaches a preset number of times.

[0132] Accordingly, in step 103, "QPGF generates a second QoS control policy based on the optimized QoS control policy generation model", specifically, QPGF generates a second QoS control policy based on the optimized generator. That is, when the aforementioned joint training is completed and the optimized generator is obtained, the second QoS control policy generated by the optimized generator can be obtained.

[0133] Figure 2 Another QoS control policy generation method provided in the embodiment of the present disclosure is applied to the policy control function PCF. The QoS control policy generation method may include but is not limited to the following steps 201 and 202:

[0134] In step 201, the PCF receives a second QoS control policy sent by a QoS control policy generation function QPGF.

[0135] In this embodiment, the second QoS control strategy is based on Figure 1 The policies generated by the QoS control policy generation method provided in the relevant embodiments.

[0136] In step 202, the PCF sends the second QoS control policy to the network device and the terminal respectively.

[0137] In this embodiment, the difference from the prior art is that the second QoS control policy is generated by the QoS control policy generation function QPGF instead of PCF. Since QPGF can provide flexible and customized QoS control policies based on the network environment and user needs, different QoS control policy generation models are optimized for different network environments and user needs, which can provide adaptive differentiated QoS control strategies for diversified business needs and achieve differentiated QoS guarantees.

[0138] In some embodiments, in step 202, “the PCF sends the second QoS control policy to the network device and the terminal respectively” is specifically:

[0139] The PCF sends a protocol data unit (PDU) session modification message to the session management function (SMF), and the PDU session modification message carries a second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

[0140] Based on the above embodiments, Figure 3 A schematic diagram of a QPGF for implementing self-generation of QoS control policies provided in an embodiment of the present disclosure. The QPGF has the function of adaptively generating new QoS control policies according to the current network status. The QPGF stores a generative adversarial network model as a QoS control policy generation model. The QPGF maintains and trains the generative adversarial network model to implement self-generation of QoS control policies.

[0141] The generative adversarial network model includes a generator and a discriminator. The generator is used to generate QoS control policies (QoS policies for short), and the discriminator is used to determine whether the input QoS policy is a real QoS policy adopted in history or a QoS policy generated by the generator. The output of the discriminator is the discrimination probability, which is the ratio of the number of real and generated QoS policies.

[0142] The generator and the discriminator are in a mutually adversarial game relationship. The goal of the discriminator is to do its best to correctly distinguish the generated strategy from the real strategy. The goal of the generator is to generate the most accurate QoS strategy as much as possible, minimize the gap between the generated strategy and the real strategy, and avoid being distinguished by the discriminator. When the discriminator cannot determine whether the input QoS policy is real or generated, for example, the discriminator outputs a judgment probability of 0.5, then it can be considered that the QoS policy generated by the generator is the real QoS policy.

[0143] The input of the generator is divided into two parts. The first part is random information that satisfies the Gaussian distribution and the discrimination probability output by the discriminator as feedback. The second part is feedback from the external network, including the classification label of the current network transmission data flow, network status and rewards. Among them, the classification labels include images, videos, texts, etc., the network status includes the expected resource occupancy rate, throughput, etc., and the reward is the user experience (Quality of Experience, QoE), etc. QoE represents the UE's experience and feelings of application services and network quality. The internal generator and discriminator of QPGF play against each other. After training both for a certain number of cycles (that is, joint training), a smaller learning rate is used to update the parameters of the generator once. When the discrimination probability of the discriminator is equal to 0.5, the training can be terminated.

[0144] Figure 4 A schematic diagram of a QoS control strategy generation process provided by an embodiment of the present disclosure, in Figure 4 In the process, QPGF realizes self-generation of QoS policy, and QPGF sends the generated QoS policy to PCF. PCF sends the QoS policy to the session management function (SMF), and sends the QoS policy to the terminal (UE) and network equipment (such as RAN equipment) through SMF. The network data analysis function (NWDAF) collects analysis data (including network performance data and terminal data), and then NWDAF performs data analysis based on the collected analysis data. Specifically, NWDAF performs network performance analysis on network performance data and user experience analysis on terminal data. Among them, network performance analysis includes network resource occupancy rate calculation and future resource occupancy rate prediction, which are used to characterize the impact of the current QoS policy on network performance, so as to judge the impact of the customized QoS policy adopted by all users in the current network on the overall performance of the network.

[0145] exist Figure 4 In the process, NWDAF provides QPGF with network performance analysis results and user experience analysis results, wherein the network performance analysis results provide the network status that needs to be input for the training of the GAN model inside the QPGF, and the user experience analysis results provide the rewards that need to be input for the training of the GAN model inside the QPGF. Combining the advantages of interactive learning of reinforcement learning, it is considered to issue the generated QoS policy through the PDU session modification process, and truly act on the network. The corresponding network element function executes the QoS policy, thereby obtaining feedback from the terminal as a reward and feeding it back to QPGF. The current network status and the generated QoS policy will affect the status and reward received from NWDAF in the next step. QPGF continuously trains the GAN model by receiving the status and reward feedback from NWDAF, as well as the classification label of the current network transmission data flow, and optimizes the QoS policy generated by the generator of the GAN model each time.

[0146] visible, Figure 4 The QoS control policy generation process shown is summarized in steps 1 to 5 as follows:

[0147] 1. QPGF initialization, training of QPGF internal generator and discriminator, and finally generating a preliminary QoS control strategy (that is, initializing the QoS control strategy);

[0148] 2. PCF receives the QoS control policy sent by QPGF, stores the QoS control policy, and then executes the issuance of the QoS control policy;

[0149] 3. The QoS control policies are exchanged between SMF, RAN and UE, and RAN and UE execute the corresponding QoS control policies;

[0150] 4. NWDAF collects data for network performance analysis and user experience analysis from RAN, UE, etc. and performs data analysis. It then feeds back the network performance analysis results and user experience analysis results obtained from the analysis to QPGF as status and reward respectively;

[0151] 5. QPGF combines the received status and rewards with a reinforcement learning algorithm to continuously train and optimize the GAN model, thereby making the generated QoS policy closer to the optimal one.

[0152] Figure 5 A schematic diagram of an interaction of generating an initialization QoS control policy for a QPGF according to an embodiment of the present disclosure is provided. Figure 5 In the process, QPGF generates an initial QoS control policy including the following steps 1 to 8:

[0153] 1. QPGF sends a network performance analysis request to AnLF, which includes QPGF Profile (configuration information), Analytics ID, etc. Analytics ID is the network performance analysis identifier.

[0154] In this embodiment, AnLF determines the NFs (Network Function) of the current network service and triggers the data collection process.

[0155] 2. AnLF initiates a network performance data collection request to NFs etc. through the data coordination function, including the event group ID. The event group ID is a group of network functions corresponding to the data that AnLF needs to collect for the network performance analysis event.

[0156] 3. NFs sends network performance data messages to AnLF through the data coordination function.

[0157] 4. AnLF initiates a model training request message to MTLF, which includes Analytics ID, AI Model ID, network performance data, etc., where AI Model ID is the identifier of the network performance analysis model.

[0158] In this embodiment, MTLF performs network performance analysis model training and reasoning, and outputs network performance analysis results.

[0159] 5. MTLF returns the network performance analysis results to AnLF.

[0160] 6. AnLF sends a network performance analysis request response to QPGF, which includes the network performance analysis results, etc.

[0161] 7. QPGF sends a QoS flow policy query request to PCF, which includes information such as QFI and timestamp.

[0162] 8. PCF sends QoS flow policy to QPGF.

[0163] In this embodiment, the QPGF trains the internal generator and discriminator of the QPGF according to the fed-back network performance analysis results and the QoS flow policy, and finally generates a preliminary QoS control policy (ie, the initialization QoS control policy).

[0164] Figure 6 An interactive schematic diagram of a QoS control strategy generation process provided by an embodiment of the present disclosure, in Figure 6 In the QoS control policy generation process, the QoS control policy generation process includes the following steps 1 to 14:

[0165] First, QPGF follows Figure 5 The process of generating the initialization QoS control policy shown initializes the generator and the discriminator, and generates a preliminary QoS control policy (ie, the initialization QoS control policy).

[0166] 1. QPGF requests UE subscription data from UDM to determine whether UE capabilities can meet the currently generated QoS policy.

[0167] In this embodiment, if the UE capability cannot meet the proposed QoS policy, the previous QoS configuration is maintained.

[0168] 2. UDM sends UE subscription data to QPGF, including UE capabilities and other information.

[0169] 3. QPGF sends a QoS policy modification message to PCF, where the QoS policy modification message includes information such as QFI and QoS parameter set.

[0170] 4. PCF records the received QoS policy and triggers the PDU (Protocol Data Units) session modification process, sends a session modification message to the SMF (Session Management function), and issues the newly generated QoS control policy, including QFI, QoS parameter set and other information.

[0171] In this embodiment, the SMF derives the PDR (Packet Detection Rule) used by the UPF (User Plane Function) according to the QoS control policy, derives the QoS Profile used by the RAN, and derives the QoS Rule (QoS Rule) used by the UE.

[0172] 5.SMF sends a session modification message to UPF, including PDR, etc.

[0173] 6. UPF sends a session modification response to SMF.

[0174] In this embodiment, the UPF performs corresponding operations according to the PDR generated by the SMF;

[0175] 7.SMF sends an N1 N2 transparent transmission request message to AMF (Authentication Management Function), which includes PDU session ID, QFI, QoS Profile. The QoS Profile includes information such as QoS parameter set, which is used by the base station to perform QoS control, and also includes related PDU session modification commands.

[0176] 8.AMF sends N1 N2 Transfer Request message response.

[0177] 9.AMF sends an N2 message to RAN, which includes the message content sent by SMF to AMF in step 7, namely, PDU session ID, QFI, QoS Profile, related PDU session modification commands, etc.

[0178] In this embodiment, the RAN searches for matching QoS parameters in the received QoS Profile according to the QFI, and performs QoS control.

[0179] 10. Signaling exchange between RAN and UE: RAN sends QFI and QoS Rule to UE, where QoS Rule includes the unique QoS Rule identifier in the PDU session, QFI of the corresponding QoS Flow and other information.

[0180] In this embodiment, the UE performs QoS control according to the QoS Rule, and associates the uplink data with the QoS Flow according to the QoS Rule.

[0181] 11.RAN sends an N2 message response, i.e. PDU session confirmation, to AMF.

[0182] 12. The UE sends a NAS (Non-Access Stratum) message to confirm the execution of the QoS policy, and the UE feeds back user experience data to the network.

[0183] 13. NWDAF initiates the data collection process to collect data related to network performance analysis and user experience from UE, RAN, AMF, etc. and perform data analysis. Figure 5 Same as steps 1 to 5.

[0184] 14. NWDAF feeds back the network performance analysis results and user experience analysis results to QPGF as status and rewards respectively.

[0185] In this embodiment, the QPGF retrains the generator and the discriminator based on the received network performance analysis results as the state and the user experience analysis results as the reward, and outputs a new QoS control policy.

[0186] Then, steps 1 to 14 are executed again until the feedback user experience reaches the best and the best QoS control policy is output.

[0187] Figure 6 Relative to Figure 5 The difference is that the generated QoS policy is issued through the PDU session modification process, and actually acts on the network. The corresponding network element function performs QoS control, so the feedback from the user is fed back to the QPGF as a reward.

[0188] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all described as a series of action combinations, but those skilled in the art can understand that the embodiments of the present disclosure are not limited by the described action sequence, because according to the embodiments of the present disclosure, some steps can be performed in other sequences or simultaneously. In addition, those skilled in the art can understand that the embodiments described in the specification are all optional embodiments.

[0189] Figure 7 This is a schematic diagram of a QoS control policy generation device provided by an embodiment of the present disclosure, which is applied to a QoS control policy generation function QPGF. Figure 7 As shown, the QoS control strategy generating device includes but is not limited to: a first unit 71, a second unit 72 and a third unit 73. The functions of each unit are described as follows:

[0190] The first unit 71 is used to receive a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, where the first network performance analysis result is a result obtained by the NWDAF analyzing data generated by the network device executing the first QoS control policy, and the first user experience analysis result is a result obtained by the NWDAF analyzing data generated by the terminal executing the first QoS control policy;

[0191] The second unit 72 is used to optimize the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result;

[0192] The third unit 73 is configured to generate a second QoS control policy based on the optimized QoS control policy generation model.

[0193] In some embodiments, the QoS control policy generating device further includes:

[0194] The fourth unit is used to send the second QoS control policy to the policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

[0195] In some embodiments, the second unit 72 is configured to:

[0196] Based on the first network performance analysis result and the first user experience analysis result, optimizing parameters of the QoS control policy generation model corresponding to the first QoS control policy;

[0197] The first QoS control policy is sent by the policy control function PCF to the network device and the terminal.

[0198] In some embodiments, the QoS control policy generating device further includes:

[0199] A generating unit, configured to generate an initialization QoS control policy as a first QoS control policy based on a QoS control policy generating model;

[0200] The sending unit is used to send the first QoS control policy to the PCF.

[0201] In some embodiments, the generating unit is configured to:

[0202] Obtain target network performance analysis results from NWDAF and stored target QoS control policies from PCF;

[0203] The target network performance analysis result and the target QoS control policy are input into the QoS control policy generation model to obtain the first QoS control policy.

[0204] In some embodiments, the generating unit obtains the target network performance analysis result from the NWDAF, including:

[0205] Send a network performance analysis request to NWDAF;

[0206] A network performance analysis response sent by NWDAF is received, where the network performance analysis response carries a target network performance analysis result; wherein the target network performance analysis result is obtained by NWDAF through training and reasoning of a network performance analysis model based on network performance data.

[0207] In some embodiments, the sending unit is configured to:

[0208] When the first QoS control policy is obtained, a terminal subscription data acquisition request is sent to a unified data management (UDM) network element;

[0209] Receive the terminal subscription data response sent by UDM, which carries the terminal capability information;

[0210] When it is determined that the terminal capability information satisfies the first QoS control policy, the first QoS control policy is sent to the PCF.

[0211] In some embodiments, the QoS control policy generation model is a generative adversarial network model, and the generative adversarial network model includes a generator and a discriminator;

[0212] The second unit 72 is used for:

[0213] Based on the first network performance analysis result and the first user experience analysis result, train the model parameters corresponding to the generator and output the training results;

[0214] Input the training result into the discriminator, and stop training the generator when it is determined that the discrimination result meets the preset training goal, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number;

[0215] The third unit 73 is used to generate a second QoS control policy based on the trained generator.

[0216] Figure 7 For details of each embodiment of the QoS control policy generation device shown in FIG. Figure 1 The details of the various embodiments of the QoS control policy generation method shown are not repeated here.

[0217] Figure 8 This is a schematic diagram of another QoS control policy generation device provided by an embodiment of the present disclosure, which is applied to a policy control function PCF. Figure 8As shown, the QoS control strategy generating device includes but is not limited to: a first unit 81 and a second unit 82. The functions of each unit are described as follows:

[0218] The first unit 81 is used to receive a second QoS control policy sent by a QoS control policy generation function QPGF, where the second QoS control policy is based on Figure 1 The policy generated by the QoS control policy generation method provided in the relevant embodiment;

[0219] The second unit 82 is used to send the second QoS control policy to the network device and the terminal respectively.

[0220] In some embodiments, the second unit 82 is configured to:

[0221] A protocol data unit PDU session modification message is sent to the session management function SMF, where the PDU session modification message carries the second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

[0222] Figure 8 For details of each embodiment of the QoS control policy generation device shown in FIG. Figure 2 The details of the various embodiments of the QoS control policy generation method shown are not repeated here.

[0223] The present disclosure also provides a processor-readable storage medium, which stores a program for causing the processor to execute Figure 1 The steps of each embodiment of the QoS control policy generation method shown in the figure or Figure 2 The steps of each embodiment of the QoS control policy generation method are shown. The processor-readable storage medium may be a non-transitory computer-readable storage medium.

[0224] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD)), etc.

[0225] Fig. 9 A schematic diagram of a communication device provided in an embodiment of the present disclosure, such as Fig. 9 As shown, the communication device provided in the embodiment of the present disclosure includes a memory 91, a transceiver 92, and a processor 93:

[0226] The memory 91 is used to store computer programs; the transceiver 92 is used to send and receive data under the control of the processor 93; the processor 93 is used to read the computer program in the memory 91 and execute:

[0227] Receiving a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, where the first network performance analysis result is a result obtained by NWDAF analyzing data generated by a network device executing a first QoS control policy, and the first user experience analysis result is a result obtained by NWDAF analyzing data generated by a terminal executing the first QoS control policy;

[0228] Optimizing the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result;

[0229] Based on the optimized QoS control policy generation model, a second QoS control policy is generated.

[0230] In some embodiments, the processor is further configured to read a computer program in the memory and execute:

[0231] The second QoS control policy is sent to the policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

[0232] In some embodiments, based on the first network performance analysis result and the first user experience analysis result, optimizing the QoS control policy generation model includes:

[0233] Based on the first network performance analysis result and the first user experience analysis result, optimizing parameters of the QoS control policy generation model corresponding to the first QoS control policy;

[0234] The first QoS control policy is sent by a policy control function PCF to the network device and the terminal.

[0235] In some embodiments, the processor is further configured to read a computer program in the memory and execute:

[0236] If the first QoS control policy is an initialized QoS control policy generated by the QPGF based on the QoS control policy generation model, the first QoS control policy is sent to the PCF.

[0237] In some embodiments, the first QoS control policy generated based on the QoS control policy generation model includes:

[0238] Obtain target network performance analysis results from NWDAF and stored target QoS control policies from PCF;

[0239] The target network performance analysis result and the target QoS control policy are input into the QoS control policy generation model to obtain the first QoS control policy.

[0240] In some embodiments, obtaining the target network performance analysis result from the NWDAF includes:

[0241] Send a network performance analysis request to NWDAF;

[0242] A network performance analysis response sent by NWDAF is received, where the network performance analysis response carries a target network performance analysis result; wherein the target network performance analysis result is obtained by NWDAF through training and reasoning of a network performance analysis model based on network performance data.

[0243] In some embodiments, sending the first QoS control policy to the PCF includes:

[0244] When the first QoS control policy is obtained, a terminal subscription data acquisition request is sent to a unified data management (UDM) network element;

[0245] Receive the terminal subscription data response sent by UDM, which carries the terminal capability information;

[0246] When it is determined that the terminal capability information satisfies the first QoS control policy, the first QoS control policy is sent to the PCF.

[0247] In some embodiments, the QoS control policy generation model is a generative adversarial network model, and the generative adversarial network model includes a generator and a discriminator;

[0248] Based on the first network performance analysis result and the first user experience analysis result, the QoS control policy generation model is optimized, including:

[0249] Based on the first network performance analysis result and the first user experience analysis result, train the model parameters corresponding to the generator and output the training results;

[0250] Input the training result into the discriminator, and stop training the generator when it is determined that the discrimination result meets the preset training goal, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number;

[0251] Based on the optimized QoS control policy generation model, a second QoS control policy is generated, including:

[0252] Based on the trained generator, a second QoS control policy is generated.

[0253] Fig. 9In the embodiment, the transceiver 92 is used to receive and send data under the control of the processor 93. The bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by the processor 93 and various circuits of the memory represented by the memory 91 are linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 92 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, which transmission medium includes a wireless channel, a wired channel, an optical cable, and other transmission media. The processor 93 is responsible for managing the bus architecture and general processing, and the memory 91 may store data used by the processor 93 when performing operations.

[0254] Fig. 9 In the embodiment, the processor 93 can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 93 or the instructions in the form of software. The processor 93 can be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0255] Fig.10 A schematic diagram of a communication device provided in an embodiment of the present disclosure, such as Fig.10 As shown, the communication device provided in the embodiment of the present disclosure includes a memory 1001, a transceiver 1002, and a processor 1003:

[0256] The memory 1001 is used to store computer programs; the transceiver 1002 is used to send and receive data under the control of the processor 1003; the processor 1003 is used to read the computer program in the memory 1001 and execute:

[0257] Receive a second QoS control policy sent by a QoS control policy generation function QPGF, where the second QoS control policy is a policy generated by the QoS control policy generation method according to any embodiment of the first aspect;

[0258] The second QoS control policy is sent to the network device and the terminal respectively.

[0259] In some embodiments, sending the second QoS control policy to the network device and the terminal respectively includes:

[0260] A protocol data unit PDU session modification message is sent to the session management function SMF, where the PDU session modification message carries the second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

[0261] Fig.10 In the embodiment, the transceiver 1002 is used to receive and send data under the control of the processor 1003. The bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by the processor 1003 and various circuits of the memory represented by the memory 1001 are linked together. The bus architecture may also link together various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 1002 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, which transmission medium includes a wireless channel, a wired channel, an optical cable, and other transmission media. The processor 93 is responsible for managing the bus architecture and general processing, and the memory 91 may store data used by the processor 1003 when performing operations.

[0262] Fig.10 In the embodiment, the processor 1003 can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 1003 or the instruction in the form of software. The processor 1003 can be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0263] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "includes..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0264] Those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present disclosure and to form different embodiments.

[0265] Those skilled in the art will appreciate that the description of each embodiment has its own emphasis, and for parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0266] Although the embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A QoS control policy generation method, applied to a QoS control policy generation function QPGF, the method include: The QPGF receives a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, where the first network performance analysis result is a result obtained by the NWDAF analyzing data generated by the network device executing the first QoS control policy, and the first user experience analysis result is a result obtained by the NWDAF analyzing data generated by the terminal executing the first QoS control policy; The QPGF optimizes a QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result; The QPGF generates a second QoS control policy based on the optimized QoS control policy generation model.

2. The method according to claim 1, in, The method further comprises: The QPGF sends the second QoS control policy to a policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

3. The method according to claim 2, in, The QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result, including: The QPGF optimizes parameters of a QoS control policy generation model corresponding to the first QoS control policy based on the first network performance analysis result and the first user experience analysis result; The first QoS control policy is sent to the network device and the terminal by a policy control function PCF.

4. The method according to claim 2, in, The method further comprises: If the first QoS control policy is an initialized QoS control policy generated by the QPGF based on the QoS control policy generation model, the QPGF sends the first QoS control policy to the PCF.

5. The method according to claim 4, in, The first QoS control policy generated by the QPGF based on the QoS control policy generation model includes: The QPGF obtains the target network performance analysis result from the NWDAF, and obtains the stored target QoS control policy from the PCF; The QPGF inputs the target network performance analysis result and the target QoS control policy into the QoS control policy generation model to obtain a first QoS control policy.

6. The method according to claim 5, in, The QPGF obtains the target network performance analysis result from the NWDAF including: The QPGF sends a network performance analysis request to the NWDAF; The QPGF receives a network performance analysis response sent by the NWDAF, wherein the network performance analysis response carries a target network performance analysis result; wherein the target network performance analysis result is obtained by the NWDAF through training and reasoning of a network performance analysis model based on network performance data.

7. The method according to claim 4, in, The QPGF sends the first QoS control policy to the PCF, including: The QPGF sends a terminal subscription data acquisition request to the unified data management UDM network element when obtaining the first QoS control policy; The QPGF receives a terminal subscription data response sent by the UDM, where the terminal subscription data response carries terminal capability information; When determining that the terminal capability information satisfies the first QoS control policy, the QPGF sends the first QoS control policy to the PCF.

8. The method according to claim 1, in, The QoS control strategy generation model is a generative adversarial network model, and the generative adversarial network model includes a generator and a discriminator; The QPGF optimizes the QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result, including: The QPGF trains the model parameters corresponding to the generator based on the first network performance analysis result and the first user experience analysis result, and outputs the training result; The QPGF inputs the training result into the discriminator, and stops training the generator when it is determined that the discrimination result meets a preset training goal, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number; The QPGF generates a second QoS control policy based on the optimized QoS control policy generation model, including: The QPGF generates a second QoS control policy based on the trained generator.

9. A QoS control policy generation method, applied to a policy control function PCF, the method include: The PCF receives a second QoS control policy sent by a QoS control policy generation function QPGF, where the second QoS control policy is a policy generated based on the QoS control policy generation method according to any one of claims 1 to 8; The PCF sends the second QoS control policy to the network device and the terminal respectively.

10. The method according to claim 9, in, The PCF sends the second QoS control policy to the network device and the terminal respectively, including: The PCF sends a protocol data unit PDU session modification message to the session management function SMF, and the PDU session modification message carries the second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

11. A QoS control policy generation device, applied to a QoS control policy generation function QPGF, the device include: A first unit is used to receive a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, wherein the first network performance analysis result is a result obtained by the NWDAF analyzing data generated by the network device executing the first QoS control policy, and the first user experience analysis result is a result obtained by the NWDAF analyzing data generated by the terminal executing the first QoS control policy; A second unit is used to optimize a QoS control policy generation model based on the first network performance analysis result and the first user experience analysis result; The third unit is used to generate a second QoS control policy based on the optimized QoS control policy generation model.

12. A QoS control policy generating device, applied to a policy control function PCF, the device include: A first unit is used to receive a second QoS control policy sent by a QoS control policy generation function QPGF, wherein the second QoS control policy is a policy generated based on the QoS control policy generation method according to any one of claims 1 to 8; The second unit is used to send the second QoS control policy to the network device and the terminal respectively.

13. A communication device, It is characterized in that Including memory, transceiver, processor: A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and executing: Receive a first network performance analysis result and a first user experience analysis result generated by a network data analysis function NWDAF, where the first network performance analysis result is a result obtained by the NWDAF analyzing data generated by the network device executing the first QoS control policy, and the first user experience analysis result is a result obtained by the NWDAF analyzing data generated by the terminal executing the first QoS control policy; Optimizing a QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result; Based on the optimized QoS control policy generation model, a second QoS control policy is generated.

14. The communication device according to claim 13, in, The processor is further configured to read the computer program in the memory and execute: The second QoS control policy is sent to a policy control function PCF, and the second QoS control policy is sent by the PCF to the network device and the terminal.

15. The communication device according to claim 14, in, The optimizing the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result includes: Based on the first network performance analysis result and the first user experience analysis result, optimizing parameters of a QoS control policy generation model corresponding to the first QoS control policy; The first QoS control policy is sent to the network device and the terminal by a policy control function PCF.

16. The communication device according to claim 14, in, The processor is further configured to read the computer program in the memory and execute: If the first QoS control policy is an initialized QoS control policy generated by the processor based on the QoS control policy generation model, the first QoS control policy is sent to the PCF.

17. The communication device according to claim 16, in, The first QoS control policy generated based on the QoS control policy generation model includes: Obtaining target network performance analysis results from the NWDAF, and obtaining stored target QoS control policies from the PCF; The target network performance analysis result and the target QoS control policy are input into the QoS control policy generation model to obtain a first QoS control policy.

18. The communication device according to claim 17, in, The obtaining of the target network performance analysis result from the NWDAF includes: Sending a network performance analysis request to the NWDAF; A network performance analysis response sent by the NWDAF is received, wherein the network performance analysis response carries a target network performance analysis result; wherein the target network performance analysis result is obtained by the NWDAF through training and reasoning of a network performance analysis model based on network performance data.

19. The communication device according to claim 16, in, The sending the first QoS control policy to the PCF includes: When the first QoS control policy is obtained, a terminal subscription data acquisition request is sent to a unified data management (UDM) network element; Receive a terminal subscription data response sent by the UDM, where the terminal subscription data response carries terminal capability information; In a case where it is determined that the terminal capability information satisfies the first QoS control policy, the first QoS control policy is sent to the PCF.

20. The communication device according to claim 13, in, The QoS control strategy generation model is a generative adversarial network model, and the generative adversarial network model includes a generator and a discriminator; The optimizing the QoS control strategy generation model based on the first network performance analysis result and the first user experience analysis result includes: Based on the first network performance analysis result and the first user experience analysis result, training the model parameters corresponding to the generator and outputting the training results; Input the training result into the discriminator, and stop training the generator when it is determined that the discrimination result meets a preset training goal, wherein the training goal includes: the discrimination result of the discriminator converges to a preset value or the number of iterations reaches a preset number; The generating a second QoS control policy based on the optimized QoS control policy generating model comprises: Based on the trained generator, a second QoS control policy is generated.

21. A communication device, It is characterized in that Including memory, transceiver, processor: A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and executing: Receive a second QoS control policy sent by a QoS control policy generation function QPGF, where the second QoS control policy is a policy generated based on the QoS control policy generation method according to any one of claims 1 to 8; The second QoS control policy is sent to the network device and the terminal respectively.

22. The communication device according to claim 21, It is characterized in that The sending the second QoS control policy to the network device and the terminal respectively includes: A protocol data unit PDU session modification message is sent to a session management function SMF, wherein the PDU session modification message carries the second QoS control policy, so that the SMF sends the second QoS control policy to the network device and the terminal through the PDU session modification process.

23. A processor-readable storage medium, It is characterized in that The processor-readable storage medium stores a program, and the program is used to enable the processor to execute the QoS control policy generation method according to any one of claims 1 to 8 or the QoS control policy generation method according to claim 9 or 10.

Images