Network slice subnet operation and maintenance management method, device, system, equipment and medium

Abstract

The embodiment of the application relates to the technical field of communication, and discloses a network slice subnet operation and maintenance management method, comprising the following steps: acquiring performance analysis results of each NF in a network slice subnet, wherein the performance analysis results are obtained through an analysis model on the NF, and the analysis model is a model obtained based on federated learning by a plurality of NFs and a common node; and performing operation and maintenance management on the network slice subnet according to the performance analysis results. The embodiment of the application also discloses a network slice subnet operation and maintenance management device, system, equipment and medium. The network slice subnet operation and maintenance management method, device, system, equipment and medium provided by the embodiment of the application can reduce the operation and maintenance management cost of the network slice subnet.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method, apparatus, system, device, and medium for network slice subnet operation and maintenance management. Background Technology

[0002] Network slicing operation and maintenance management is a significant challenge currently faced by 5G networks. In performing network slicing operation and maintenance management, the slice management system needs to translate tenant SLA (Service Level Agreement) requirements into instantiation parameters and service parameters for slice subnets in the radio, transport, and core network domains, and then complete the operation and maintenance management of network slices and their subnets within the slice management system.

[0003] Currently, when performing operation and maintenance management on a slice (network slice subnet), the slice management system needs to obtain management data and service data from the physical site or VNF (Virtual Network Function) corresponding to the NF in the network slice subnet for analysis, and then perform operation and maintenance management of the network slice subnet based on the analysis results.

[0004] However, the data obtained from the physical site or VNF corresponding to the NF is a large amount of periodic data. The acquisition and processing of this data will greatly increase the operation and maintenance management cost of the network slice subnet (such as network bandwidth and data storage cost). At the same time, the current optimization methods for network slice subnets are still based on expert experience, and there is no effective means to achieve intelligent optimization and improvement of the SLA of the network slice subnet. Summary of the Invention

[0005] The main objective of this application is to provide a method, apparatus, system, device, and medium for the operation and maintenance management of network slice subnets, which can reduce the operation and maintenance management cost of network slice subnets.

[0006] To achieve the above objectives, this application provides a method for network slice subnet operation and maintenance management, including: obtaining the performance analysis results of each NF in the network slice subnet, wherein the performance analysis results are obtained through an analysis model on the NF, and the analysis model is a model obtained by federated learning from several NFs and a common node; and performing operation and maintenance management on the network slice subnet based on the performance analysis results.

[0007] To achieve the above objectives, this application also provides a network slice subnet operation and maintenance management device, including: an acquisition module, used to acquire the performance analysis results of each NF in the network slice subnet, wherein the performance analysis results are obtained through an analysis model on the NF, and the analysis model is a model obtained by federated learning from several NFs and a common node; and a management module, used to perform operation and maintenance management on the network slice subnet according to the performance analysis results.

[0008] To achieve the above objectives, this application also provides a network slice subnet operation and maintenance management system, including a public node and several NFs belonging to a network slice subnet. The public node is used to issue model training tasks to each NF, perform model aggregation based on the model parameters reported by each NF, issue the aggregated model parameters to each NF for iterative learning, obtain the performance analysis results of the analysis model obtained by the NF based on the iterative learning, and perform operation and maintenance management of the network slice subnet based on the performance analysis results. The model training task includes federated learning strategy information, machine learning algorithm information, and subscription data information. The NF is used to receive the model training tasks issued by the public node, perform model training based on the model training tasks, report the trained model parameters to the public node, receive the model parameters issued by the public node, perform iterative learning based on the issued model parameters, and obtain the analysis model after reaching a preset number of iterations or a preset model accuracy.

[0009] To achieve the above objectives, embodiments of this application also provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the above-described network slice subnet operation and maintenance management method.

[0010] To achieve the above objectives, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described network slice subnet operation and maintenance management method.

[0011] The network slice subnet operation and maintenance management method proposed in this application obtains the performance analysis results of each NF in the network slice subnet and performs operation and maintenance management based on the performance analysis results. The performance analysis results are obtained through the analysis model of the NF, which is a model derived from several NFs and a common node based on federated learning. Because of the federated learning approach, each NF can analyze its own data using its own model. Therefore, during the operation and maintenance management of the network slice subnet, the NSSMF of the network slice subnet can directly obtain the analysis results from the corresponding physical sites or VNFs of the NFs, eliminating the need to obtain and process large amounts of periodic data from physical sites or VNFs, thus reducing the demand for network bandwidth and data storage costs. Simultaneously, by learning and managing the network slice subnet through the federated learning mechanism, the reliance on experienced professionals can be reduced, achieving intelligent and automated optimization and improvement of the network slice subnet SLA, thereby further reducing the operation and maintenance management cost of the network slice subnet. Attached Figure Description

[0012] One or more embodiments are illustrated by way of example with reference to the accompanying drawings, and these illustrative descriptions do not constitute a limitation on the embodiments.

[0013] Figure 1 This is a flowchart illustrating the network slice subnet operation and maintenance management method provided in the first embodiment of the present invention;

[0014] Figure 2 This is a flowchart illustrating the federated learning process of the network slice subnet operation and maintenance management method provided in the first embodiment of the present invention.

[0015] Figure 3 This is a flowchart illustrating the network slice subnet operation and maintenance management method provided in the second embodiment of the present invention;

[0016] Figure 4 This is a flowchart illustrating the implementation of the network slice subnet operation and maintenance management method provided in the second embodiment of the present invention on the RAN side.

[0017] Figure 5 This is a flowchart illustrating the implementation of the network slice subnet operation and maintenance management method provided in the second embodiment of the present invention on the CN side.

[0018] Figure 6 This is a flowchart illustrating the network slice subnet operation and maintenance management method provided in the third embodiment of the present invention;

[0019] Figure 7 This is a flowchart illustrating the detailed steps following S303 of the network slice subnet operation and maintenance management method provided in the third embodiment of the present invention.

[0020] Figure 8 This is a flowchart illustrating the network slice subnet operation and maintenance management method provided in the third embodiment of the present invention;

[0021] Figure 9 This is a schematic diagram of the module structure of the network slicing subnet operation and maintenance management device provided in the fourth embodiment of the present invention;

[0022] Figure 10 This is a schematic diagram of the network slice subnet operation and maintenance management system provided in the fifth embodiment of the present invention;

[0023] Figure 11 This is a structural example diagram of the network slice subnet operation and maintenance management system provided in the fifth embodiment of the present invention;

[0024] Figure 12 This is a schematic diagram of the structure of the electronic device provided in the sixth embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this application to help readers better understand this application. However, the technical solutions claimed in this application can be implemented even without these technical details and various changes and modifications based on the following embodiments. The division of the various embodiments below is for the convenience of description and should not constitute any limitation on the specific implementation of this application. The various embodiments can be combined with and referenced by each other without contradiction.

[0026] The first embodiment of this invention relates to a network slice subnet operation and maintenance management method. This method obtains the performance analysis results of each NF (Network Function) in the network slice subnet. The performance analysis results are obtained through an analysis model on each NF, which is a model derived from several NFs and a common node based on federated learning. The network slice subnet is then managed and maintained according to the performance analysis results. Because of the federated learning approach, each NF can analyze its own data using its own model. Therefore, during the operation and maintenance management of the network slice subnet, the NSSMF (Network Slice Formatting Formatting Function) of the network slice subnet can directly obtain the analysis results from the corresponding physical sites or VNFs of the NFs, eliminating the need to obtain and process large amounts of periodic data from physical sites or VNFs, thus reducing the demand for network bandwidth and data storage costs. Simultaneously, by learning and managing the network slice subnet through the federated learning mechanism, the reliance on experienced professionals can be reduced, enabling intelligent and automated optimization and improvement of the network slice subnet's SLA (Service Level Agreement), thereby further reducing the operation and maintenance management costs of the network slice subnet.

[0027] It should be noted that the execution entity of the network slice subnet operation and maintenance management method provided in this embodiment of the invention can be a slice management system, specifically an NSSMF (Network Slice Subnet Management Function) entity on the Radio Access Network (RAN) side, or an NSSMF entity on the Core Network (CN) side. It is understood that implementing the network slice subnet operation and maintenance management method through NSSMF instead of NSMF (Network Slice Management Function) can reduce the complexity of the operation and maintenance management process on the NSMF side and improve the timeliness of slice operation and maintenance management; at the same time, implementing the network slice subnet operation and maintenance management method through NSSMF can enable the network slice subnet to have automated management functions, improving the level of intelligence in operation and maintenance management.

[0028] The specific process of the network slice subnet operation and maintenance management method provided in the embodiments of the present invention is as follows: Figure 1 As shown, it includes the following steps:

[0029] S101: Obtain the performance analysis results of each NF in the network slice subnet. The performance analysis results are obtained through the analysis model on the above NF. The analysis model is a model obtained by federated learning based on several of the above NFs and a common node.

[0030] The performance analysis results of NFs can include the analysis results of the NF's own operating performance or the performance impact of NFs on network slice subnets, which can serve as the basis for the operation and maintenance management of network slice subnets.

[0031] After obtaining the analysis model through federated learning, each NF can use the obtained analysis model to analyze its own data, thereby obtaining the performance analysis results of its NF. It should be understood that when using federated learning to train the analysis model, it can be a subset of NFs in the network slice subnet and a common node for federated learning. Other NFs not participating in federated learning can download the analysis model from the common node after training. Preferably, to make the sample data more comprehensive and the prediction results of the analysis model more accurate, federated learning can be performed on all NFs in the network slice subnet and the common node. The common node can be a node directly or indirectly connected to each NF, such as an EM (Element Management) directly connected to each NF. The execution entity NSSMF obtains the performance analysis results of each NF from the EM. Since the EM does not necessarily act as an intermediary node between NFs and common nodes on the management network—for example, on the core network side, each VNF does not depend on the EM, and the VNF can be directly connected to the NSSMF—it is preferable to choose the NSSMF as the common node.

[0032] It should be noted that since the same NF may belong to different network slice subnets, and the content and environment of different network slice subnets will be different, when the network slice subnet management method provided by the embodiments of the present invention is used to analyze different network slice subnets, the performance analysis results of the same NF in different network slice subnets may be different.

[0033] In a specific example, before S101, that is, before obtaining the performance analysis results of each NF in the network slice subnet, the following steps are also included:

[0034] The public node distributes model training tasks to each NF, so that the NF can train the model according to the model training tasks and report the model parameters obtained from the model training to the public node. The model training tasks include federated learning policy information, machine learning algorithm information and subscription data information.

[0035] The common nodes are used to aggregate the models based on the reported model parameters, and the aggregated model parameters are then distributed to each NF for iterative learning. Once the preset model accuracy is achieved, the analysis model is obtained.

[0036] The federated learning strategy can include: horizontal federated learning, vertical federated learning, or transfer federated learning. Machine learning algorithm information can include data parallelism, model parallelism, or graph parallelism, and iteration strategy information (single iteration, echelon-based iteration, or hybrid iteration, etc.). Subscribed data information refers to which data NF needs to subscribe to for model training, such as subscribing to MR (measurement reports) or Trace (tracking) data.

[0037] Specifically, NSSMF uses public nodes to distribute model training tasks to each NF in the network slice subnet. Upon receiving the training task, each NF subscribes to relevant data based on the subscription data information in the task, and performs model training and inference based on the federated learning strategy and machine learning algorithm information. The NF then reports the model parameters (such as model gradients) calculated during inference to the public nodes. After receiving the reported model parameters, NSSMF uses the public nodes to perform model aggregation based on these parameters and distributes the aggregated model parameters to each NF for iterative learning. After reaching a preset number of iterations or a preset model accuracy, the analysis model is obtained. The preset number of iterations or the preset model accuracy can be set according to actual needs and are not specifically limited here.

[0038] It should be understood that model training tasks and subscribed data information can be set and changed according to actual needs. When model training tasks and subscribed data information change, updated analysis models and performance analysis results can be obtained. In practical applications, model training tasks and subscribed data information can be changed according to performance analysis results and operation and maintenance management needs, thereby updating or further optimizing the analysis model and performance analysis results.

[0039] Please refer to Figure 2 This is a flowchart illustrating the federated learning process of the network slice subnet management method provided in this embodiment of the invention. Figure 2 In the process, when NSSMF issues model training tasks to each NF, it first issues the model training tasks to EM, and then EM issues the model training tasks to the NF. When the NF reports model parameters to NSSMF, it first sends the model parameters to EM, and then EM passes them through to NSSMF. When NSSMF issues the merged model parameters to the NF, it also issues the merged model parameters to the NF through EM. Figure 2 The other processes are basically the same as those described above, and will not be repeated here.

[0040] It should be understood that when the public node is NSSMF, using the public node to distribute model training tasks to each NF means that NSSMF directly distributes model training tasks to each NF, and using the public node to perform model aggregation based on the reported model parameters is similar.

[0041] S102: Perform operation and maintenance management on network slice subnets based on performance analysis results.

[0042] Optionally, corresponding operation and maintenance (O&M) rules can be set based on the performance analysis results. After obtaining the performance analysis results for each NF, O&M management can be implemented according to these rules. For example, if the performance analysis result of an NF is its own operational performance, the corresponding O&M rule is to adjust the NF's service parameters if the performance analysis result is poor. After obtaining the performance analysis results for each NF, NSSMF adjusts the service parameters of NFs with poor performance analysis results according to the corresponding O&M rules to improve the NF's own operational performance, thereby achieving the purpose of O&M management for each NF in the network slice subnet. As another example, if the corresponding O&M rule is to adjust the NF's instantiation parameters if the proportion of NFs with poor performance analysis results in the network slice subnet reaches 60%, NSSMF, after obtaining the performance analysis results for each NF, if statistically, the proportion of NFs with poor performance analysis results reaches 65%, then it adjusts the NF's instantiation parameters according to the O&M rules to improve the corresponding NF instantiation template. Specific O&M rules can be set according to actual needs; no specific restrictions are imposed here.

[0043] The network slice subnet management method provided by this invention obtains the performance analysis results of each NF in the network slice subnet and performs operation and maintenance management based on the performance analysis results. The performance analysis results are obtained through the analysis model of the NF, which is a model derived from several NFs and a common node based on federated learning. Because of the federated learning approach, each NF can analyze its own data using its own model. Therefore, during the operation and maintenance management of the network slice subnet, the NSSMF of the network slice subnet can directly obtain the analysis results from the corresponding physical sites or VNFs of the NFs, eliminating the need to obtain and process large amounts of periodic data from physical sites or VNFs, thus reducing the demand for network bandwidth and data storage costs. Simultaneously, by learning and managing the network slice subnet through the federated learning mechanism, the reliance on experienced professionals can be reduced, achieving intelligent and automated optimization and improvement of the network slice subnet's SLA, thereby further reducing the operation and maintenance management cost of the network slice subnet.

[0044] The second embodiment of the present invention relates to a network slice subnet operation and maintenance management method. The second embodiment is largely the same as the first embodiment, with the main difference being that: in this embodiment, the performance analysis result is the contribution of each NF's KPI to the network slice subnet's KPI; correspondingly, the operation and maintenance management of the network slice subnet based on the performance analysis result includes: determining the NFs to be optimized based on the contribution, and sending the NF information to be optimized to the NFVO (Network Function Virtual Orchestrator) of the network slice subnet, so that the NFVO can initiate the optimized deployment of NF instances or the creation of new NF instances to the VNFM (VNF Manager) based on the information.

[0045] The specific process of the network slice subnet operation and maintenance management method provided in the embodiments of the present invention is as follows: Figure 3 As shown, it includes the following steps:

[0046] S201: Based on the analysis model, obtain the contribution of each NF's KPI to the KPI of the network slice subnet.

[0047] When determining the contribution of each NF's KPIs to the network slice subnet's KPIs, KPIs common to both the NF and the network slice subnet can be selected, such as call latency, QoS (Quality of Service), and number of users. By assessing the contribution of each NF's KPIs to the network slice subnet's KPIs, the extent of the NF's impact on the network slice subnet's SLA (Service Level Agreement) can be determined.

[0048] Optionally, the training process of the analysis model can be as follows: NSSMF initiates a training request to EM targeting the network slice subnet (specifically, NSSI (Network Slice Subnet)). The model training task (identification of network slice subnets) is as follows: After receiving the model training task, the EM decomposes the task and initiates model training tasks for the contribution of the KPIs of the NFs and the network slice subnets to the list of NFs belonging to the same NSSI. After receiving the model training task, each NF subscribes to data (such as MR, Trace, KPI, etc.) according to the subscription data information of the model training task, and then performs model training and inference according to the federated learning strategy information (such as horizontal federated learning) and machine learning algorithm information (such as data parallelism) in the model training task, and reports the trained model parameters to the EM. The EM then passes the model parameters reported by the NFs to the NSSMF. After the NSSMF completes the collection of model parameters from each NF, it performs model aggregation (according to the model gradient descent method) to obtain the aggregated model. The NSSMF distributes the aggregated model (or model parameters) to each NF through the EM for iterative learning. After the number of iterations reaches a preset number or the aggregated model reaches a preset model accuracy, the analysis model is obtained.

[0049] S202: Determine the NF to be optimized based on the contribution, and obtain the NF information to be optimized.

[0050] Optionally, the contribution can include positive and negative contributions. It is understood that if the contribution of a certain NF is negative and deviates significantly from the KPI of the network slice subnet (e.g., the negative contribution exceeds 30%), it indicates that it is the "weak link" in the network slice subnet and needs to be optimized accordingly.

[0051] Optionally, a contribution threshold can be set. If the contribution of an NF is negative and less than the threshold, the NF is identified as an NF to be optimized. The contribution threshold can be set according to the actual situation, and no specific restrictions are imposed here.

[0052] After identifying all functional processes (NFs) to be optimized, the information about these NFs can be obtained. Optionally, an NF optimization list can be generated based on all the NFs to be optimized, and this list can be used as the information about the NFs to be optimized.

[0053] S203: Send the NF information to be optimized to the NFVO of the network slice subnet, so that the NFVO can initiate the optimized deployment of the NF instance or the creation of the NF instance to the VNFM based on the NF information to be optimized.

[0054] Optionally, when optimizing the deployment of NF instances, NFVO / VNFM can adjust the NSSD (Network Slice Subnet Description) or the network service description file according to the NF information to be optimized, thereby achieving the goal of optimizing the deployment of NF instances.

[0055] When VNFM optimizes and deploys NF instances or creates new NF instances based on the NF information to be optimized, it can set corresponding operation and maintenance management rules for the NF information to be optimized. This allows VNFM to optimize and deploy NF instances or create new NF instances according to the corresponding operation and maintenance management rules after obtaining the NF information to be optimized.

[0056] Please refer to Figure 4 This is a flowchart illustrating the implementation of the network slice subnet operation and maintenance management method provided in this invention on the RAN side. Specifically, the NSSMF distributes model training tasks to each NF through the EM for federated learning to obtain an analysis model. Then, based on the analysis model, it obtains the contribution of each NF's KPI to the network slice subnet's KPI. Based on the obtained contribution, it initiates the optimized deployment of the network slice subnet to the NFVO. The NFVO then initiates the optimized deployment of NFs to the VNFM to optimize the deployment of NF instances or create new NF instances.

[0057] As mentioned above, the network slice subnet operation and maintenance management method provided in this embodiment of the invention can also be applied to the core network (CN) side. Please refer to... Figure 5 This is a flowchart illustrating the implementation of the network slice subnet operation and maintenance management method provided in this invention on the CN side. Its specific process is similar to... Figure 4 The process is roughly the same, the difference being that the NSSMF on the CN side connects directly to the NF on the CN side, instead of connecting through EM.

[0058] The network slice subnet operation and maintenance management method provided by this invention obtains the contribution of each NF's KPI to the network slice subnet's KPI based on an analysis model. This contribution is then used as a performance analysis result to optimize the deployment of NF instances or create new NF instances. Since the contribution of an NF's KPI to the network slice subnet's KPI reflects the degree of impact of the NF on the network slice subnet's SLA, the points requiring optimization in the network slice subnet can be determined based on the contribution. Then, by optimizing the deployment of NF instances or creating new NF instances based on the corresponding issues, the overall performance of the network slice subnet can be improved, thus increasing the efficiency of network slice subnet operation and maintenance management.

[0059] The third embodiment of the present invention relates to a network slice subnet operation and maintenance management method. The third embodiment is largely the same as the first embodiment, with the main difference being that: in this embodiment, the correlation between the service parameters of each NF and the KPI of the network slice subnet is used as the performance analysis result; correspondingly, the operation and maintenance management of the network slice subnet based on the performance analysis result includes: obtaining the instantiated parameters corresponding to the service parameters in the network slice subnet, and optimizing the instantiated parameters based on the correlation.

[0060] The specific process of the network slice subnet operation and maintenance management method provided in the embodiments of the present invention is as follows: Figure 6 As shown, it includes the following steps:

[0061] S301: Obtain the correlation between the service parameters of each NF and the KPIs of the network slice subnet based on the analysis model.

[0062] When determining the correlation between the service parameters of each NF and the KPIs of the network slice subnet, service parameters with a high degree of correlation with the KPIs of the network slice subnet can be selected, such as cell selection algorithms, cell scenarios, directional coverage parameters, and cell power. By obtaining the correlation between the service parameters of each NF and the KPIs of the network slice subnet, the impact of the NF service parameter settings on the SLA of the network slice subnet can be determined.

[0063] Optionally, the training process of the analysis model can be as follows: NSSMF initiates a model training task for the Network Slice Subnet (NSSI) to EM. After receiving the model training task, EM decomposes the task and initiates model training tasks for the correlation between the business parameters of the NFs and the KPIs of the network slice subnet to the NF list belonging to the same NSSI. After receiving the model training task, each NF subscribes to data (such as business parameters, MR, Trace, KPI, etc.) according to the subscription data information of the model training task, and then performs model training and inference according to the federated learning strategy information (such as horizontal federated learning) and machine learning algorithm information (such as data parallelism) in the model training task, and reports the trained model parameters to EM. EM then passes the model parameters reported by the NFs to NSSMF. After collecting the model parameters of each NF, NSSMF performs model aggregation to obtain the aggregated model. NSSMF distributes the aggregated model (or model parameters) to each NF through EM for iterative learning. After the number of iterations reaches a preset number or the aggregated model reaches a preset model accuracy, the analysis model is obtained.

[0064] Optionally, during the training and analysis model, business parameters can be actively adjusted (e.g., increased or decreased) to analyze whether the business parameters significantly affect the KPIs of NF before and after the adjustment, thus providing a basis for modifying the business parameters when necessary.

[0065] S302: Obtain the instantiated parameters corresponding to the service parameters in the network slice subnet.

[0066] S303: Optimize instantiation parameters based on correlation.

[0067] For S302 and S303, the following explanation is provided:

[0068] After obtaining the correlation between the service parameters of each NF and the KPIs of the network slice subnet, NSSMF obtains the instantiation parameters corresponding to the service parameters, and then optimizes the instantiation parameters based on the correlation. Specifically, NSSMF can notify NFVO, and then NFVO can initiate the modification of instantiation parameters to VNFM. Alternatively, when optimizing instantiation parameters based on correlation, operation and maintenance management rules for instantiation parameters can be set based on the correlation, so that after obtaining the correlation, the instantiation parameters can be optimized according to the corresponding operation and maintenance management rules.

[0069] In a specific example, after S303, that is, after optimizing the instantiation parameters based on the correlation, such as Figure 7 As shown, it also includes the following steps:

[0070] S304: Determine the business parameters to be optimized and the target NF corresponding to the business parameters to be optimized based on the correlation.

[0071] S305: Send a request to the target NF to modify the business parameters, so that the target NF can modify the business parameters to be optimized according to the request. The request includes the modified business parameters.

[0072] Specifically, NSSMF determines the business parameters to be optimized and the target NF based on the correlation. It then sends a request to modify the business parameters to the target NF through EM. After receiving the request, the target NF modifies the corresponding business parameters according to the modified business parameters in the request, thereby optimizing the business parameters.

[0073] Please refer to Figure 8This is a flowchart illustrating the network slice subnet operation and maintenance management method provided in this embodiment of the invention. Specifically, the NSSMF distributes model training tasks to each NF through the EM for federated learning to obtain an analysis model. Then, based on the analysis model, it obtains the correlation between the service parameters of each NF and the KPIs of the network slice subnet. Based on the obtained correlation, it initiates the optimization deployment of the instantiation parameters of the network slice subnet to the VNFM through the NFVO. The NSSMF distributes a request to modify the service parameters to the target NF through the EM. After receiving the request, the target NF modifies the corresponding service parameters.

[0074] The network slice subnet operation and maintenance management method provided in this invention obtains the correlation degree of each NF's service parameters with the network slice subnet's KPIs based on an analysis model, and uses the correlation degree as a performance analysis result to optimize the instantiation parameters. Since the correlation degree of an NF's service parameters with the network slice subnet's KPIs reflects the degree of impact of that service parameter on the network slice subnet's SLA, the instantiation parameters corresponding to the service parameters can be determined based on the correlation degree. Optimizing the instantiation parameters based on the correlation degree can improve the overall performance of the network slice subnet and increase the efficiency of network slice subnet operation and maintenance management.

[0075] Furthermore, those skilled in the art will understand that the step divisions of the various methods described above are merely for clarity of description. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, without changing the core design of the algorithm and process, are also within the scope of protection of this patent.

[0076] The fourth embodiment of the present invention relates to a network slicing subnet operation and maintenance management device 400, such as... Figure 9 As shown, it includes an acquisition module 401 and a management module 402. The functions of each module are described in detail below:

[0077] Module 401 is used to obtain the performance analysis results of each NF in the network slice subnet. The performance analysis results are obtained through the NF.

[0078] The analytical model is obtained from the above, which is a model based on federated learning, consisting of several NFs and a common node;

[0079] Management module 402 is used to perform operation and maintenance management on network slice subnets based on performance analysis results.

[0080] Furthermore, the network slice subnet operation and maintenance management device 400 provided in this embodiment of the invention also includes a training module, wherein the training module is used for:

[0081] The public node distributes model training tasks to each NF, so that the NF can train the model according to the model training tasks and report the trained model parameters to the public node. The model training tasks include federated learning policy information, machine learning algorithm information and subscription data information.

[0082] The model is aggregated using the public nodes based on the reported model parameters, and the aggregated model parameters are then distributed to each NF for iterative learning. After reaching the preset number of iterations or the preset model accuracy, the analysis model is obtained.

[0083] Furthermore, the acquisition module 401 is also used for:

[0084] The contribution of each NF's KPI to the KPI of the network slice subnet is obtained based on the analysis model, and the contribution is used as the performance analysis result.

[0085] Furthermore, the management module 402 is also used for:

[0086] The NF to be optimized is determined based on the contribution, and the information of the NF to be optimized is obtained;

[0087] The NF information to be optimized is sent to the NFVO of the network slice subnet, so that the NFVO can initiate the optimized deployment of NF instances or the creation of new NF instances to the VNFM based on the information.

[0088] Furthermore, the acquisition module 401 is also used for:

[0089] The correlation between the service parameters of each NF and the KPIs of the network slice subnet is obtained based on the analysis model, and the correlation is used as the performance analysis result.

[0090] Furthermore, the management module 402 is also used for:

[0091] Retrieve the instantiated parameters corresponding to the business parameters in the network slice subnet;

[0092] Optimize instantiation parameters based on correlation.

[0093] Furthermore, the management module 402 is also used for:

[0094] Determine the business parameters to be optimized and the target NF corresponding to the business parameters to be optimized based on the correlation degree;

[0095] Send a request to the target NF to modify the business parameters, so that the target NF can modify the business parameters to be optimized according to the request. The request includes the modified business parameters.

[0096] Furthermore, the public node is NSSMF.

[0097] It is not difficult to see that this embodiment is a device embodiment corresponding to the aforementioned method embodiment, and this embodiment can be implemented in conjunction with the aforementioned method embodiment. The relevant technical details mentioned in the aforementioned method embodiment are still valid in this embodiment, and will not be repeated here to reduce repetition. Correspondingly, the relevant technical details mentioned in this embodiment can also be applied to the aforementioned method embodiment.

[0098] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by this invention; however, this does not mean that other units are absent from this embodiment.

[0099] The fifth embodiment of the present invention relates to a network slice subnet operation and maintenance management system, such as... Figure 10 As shown, it includes a public node 501 and several NF502s belonging to a network slice subnet;

[0100] The public node 501 is used to issue model training tasks to each NF502, perform model aggregation based on the model parameters reported by each NF502, issue the aggregated model parameters to each NF502 for iterative learning, obtain the performance analysis results of the analysis model obtained by NF502 based on iterative learning, and perform network slice subnet operation and maintenance management based on the performance analysis results. The model training task includes federated learning strategy information, machine learning algorithm information and subscription data information.

[0101] NF502 is used to receive model training tasks from public node 501, train the model according to the model training tasks, report the trained model parameters to public node 501, receive model parameters from public node 501, perform iterative learning according to the model parameters, and obtain the analysis model after reaching the preset number of iterations or the preset model accuracy.

[0102] Furthermore, the network slicing subnet operation and maintenance management system also includes EM;

[0103] NF502 connects to EM through the OAM channel of this NF, and EM connects to common node 501 through the NAF module of this EM.

[0104] Please refer to Figure 11This is a specific example diagram of the network slice subnet operation and maintenance management system provided in the embodiments of the present invention. Specifically, each NF acts as a participant (Federal Agent) in federated learning, while the public node 501 is the NSSMF, that is, the NSSMF acts as the coordinator (Federal Coordinator), and the EM acts as an intermediary channel. Multiple instances can exist, interacting with the NF and NSSMF respectively, to help the NSSMF coordinate information and converge models. Regarding... Figure 11 The following points need to be noted regarding the network slicing subnet operation and maintenance management system:

[0105] 1. The Federated Learning Agent and Coordinator modules are optional functional modules that can be dynamically deployed as needed without affecting the existing NF and NSSMF management models and capabilities.

[0106] 2. The NF management model introduces model learning modeling, aligning management model capabilities with NF's existing management capabilities such as alarm management, performance management, version management, and security management. When managing the model, parameter settings for model learning, training, and inference can be implemented based on a model-driven approach.

[0107] 3. NF and EM reuse the existing OAM channels of NF and EM. There are various specific protocols for OAM channels, such as NETCONF, RESTFUL, SNMPv3 or TLS.

[0108] 4. Internally, EM implements the northbound pass-through function for model learning and training through the NAF module, and provides open interfaces (such as RESTful interfaces) to the outside world;

[0109] 5. NSSMF uses the northbound open interface of EM to realize data and control interaction between NSSMF and EM;

[0110] 6. The capabilities of the Federal Agent module include: ① support for multiple machine learning algorithms; ② the ability to subscribe to various management data and network business data; ③ the ability to perform machine learning based on learning tasks; ④ the ability to train and infer models, output models, model parameters, and model gradients; ⑤ the ability to learn and infer based on model iterations; and ⑥ an independent model transmission channel.

[0111] 7. The capabilities of the Federal Coordinator module include: ① management of NF model training and inference, ② visualization of NF output models, ③ aggregation of NF output models, ④ management of iterative learning of NF output models, and ⑤ independent acquisition channels for NF output models.

[0112] Furthermore, NF502 employs distributed machine learning algorithms for model training.

[0113] Furthermore, public node 501 is also used for:

[0114] The contribution of each NF502 KPI to the KPI of the network slice subnet is obtained based on the analysis model, and the contribution is used as the performance analysis result.

[0115] Based on the contribution, the NF502 to be optimized is determined, and the NF information to be optimized is obtained;

[0116] The NF information to be optimized is sent to the NFVO of the network slice subnet, so that the NFVO can initiate the optimized deployment of NF instances or the creation of new NF instances to the VNFM based on the information.

[0117] Furthermore, public node 501 is also used for:

[0118] The correlation between the service parameters of each NF502 and the KPIs of the network slice subnet is obtained based on the analysis model, and the correlation is used as the performance analysis result.

[0119] Retrieve the instantiated parameters corresponding to the business parameters in the network slice subnet;

[0120] Optimize instantiation parameters based on correlation;

[0121] And / or,

[0122] Determine the business parameters to be optimized and the target NF502 corresponding to the business parameters to be optimized based on the correlation degree;

[0123] Send a request to modify service parameters to the target NF502, so that the target NF502 can modify the service parameters to be optimized according to the request. The request includes the modified service parameters.

[0124] It is not difficult to see that this embodiment is a system embodiment corresponding to the aforementioned method embodiment, and this embodiment can be implemented in conjunction with the aforementioned method embodiment. The relevant technical details mentioned in the aforementioned method embodiment are still valid in this embodiment, and will not be repeated here to reduce repetition. Correspondingly, the relevant technical details mentioned in this embodiment can also be applied to the aforementioned method embodiment.

[0125] The sixth embodiment of the present invention relates to an electronic device, such as... Figure 12 As shown, it includes: at least one processor 601; and a memory 602 communicatively connected to at least one processor 601; wherein the memory 602 stores instructions that can be executed by at least one processor 601, and the instructions are executed by at least one processor 601 to enable at least one processor 601 to execute the above-described network slice subnet operation and maintenance management method.

[0126] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0127] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0128] The seventh embodiment of the present invention relates to a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the above-described method embodiment.

[0129] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0130] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing the present invention, and in practical applications, various changes in form and detail may be made without departing from the spirit and scope of the present invention.

Claims

1. A method for network slice subnet operation and maintenance management, characterized in that, include: The performance analysis results of each NF in the network slice subnet are obtained. The performance analysis results are obtained through the analysis model on the NF. The analysis model is a model obtained by federated learning from several NFs and a common node. The network slice subnet is operated and managed based on the performance analysis results. The step of obtaining the performance analysis results of each NF in the network slice subnet includes: The correlation degree between the service parameters of each NF and the KPI of the network slice subnet is obtained according to the analysis model, and the correlation degree is used as the performance analysis result. The steps for operating and maintaining the network slice subnet based on the performance analysis results include: Obtain the instantiated parameters corresponding to the service parameters in the network slice subnet; The instantiation parameters are optimized based on the correlation.

2. The network slice subnet operation and maintenance management method according to claim 1, characterized in that, Before obtaining the performance analysis results of each NF in the network slice subnet, the following is also included: The public node is used to issue a model training task to each NF, so that the NF can train the model according to the model training task and report the trained model parameters to the public node. The model training task includes federated learning strategy information, machine learning algorithm information and subscription data information. The common node is used to aggregate the model based on the reported model parameters, and the aggregated model parameters are sent to each NF for iterative learning. After reaching a preset number of iterations or a preset model accuracy, the analysis model is obtained.

3. The network slice subnet operation and maintenance management method according to claim 1, characterized in that, The performance analysis results of each NF in the network slice subnet obtained include: The contribution of the KPI of each NF to the KPI of the network slice subnet is obtained according to the analysis model, and the contribution is used as the performance analysis result.

4. The network slice subnet operation and maintenance management method according to claim 3, characterized in that, The operation and maintenance management of the network slice subnet based on the performance analysis results includes: Based on the contribution level, the NF to be optimized is determined, and the NF information to be optimized is obtained; The NF information to be optimized is sent to the NFVO of the network slice subnet, so that the NFVO can initiate the optimized deployment of NF instances or the creation of new NF instances to the VNFM based on the information.

5. The network slice subnet operation and maintenance management method according to claim 1, characterized in that, The operation and maintenance management of the network slice subnet based on the performance analysis results includes: Based on the correlation, determine the business parameters to be optimized and the target NF corresponding to the business parameters to be optimized; A request to modify service parameters is sent to the target NF, so that the target NF can modify the service parameters to be optimized according to the request, wherein the request includes the modified service parameters.

6. The network slice subnet operation and maintenance management method according to any one of claims 1-5, characterized in that, The public node is NSSMF.

7. A network slicing subnet operation and maintenance management device, characterized in that, include: The acquisition module is used to acquire the performance analysis results of each NF in the network slice subnet. The performance analysis results are obtained through the analysis model on the NF. The analysis model is a model obtained by federated learning from several NFs and a common node. The management module is used to perform operation and maintenance management on the network slice subnet based on the performance analysis results; The acquisition module is specifically used to acquire the correlation degree of the service parameters of each NF with the KPI of the network slice subnet according to the analysis model, and use the correlation degree as the performance analysis result; The management module is specifically used to obtain the instantiation parameters corresponding to the service parameters in the network slice subnet, and optimize the instantiation parameters according to the correlation.

8. A network slicing subnet operation and maintenance management system, characterized in that, It includes a public node and several NFs belonging to a network slice subnet; The public node is used to issue model training tasks to each NF, perform model aggregation based on the model parameters reported by each NF, issue the aggregated model parameters to each NF for iterative learning, obtain the performance analysis results of the analysis model obtained by the NF based on the iterative learning, and perform operation and maintenance management of the network slice subnet based on the performance analysis results. The model training task includes federated learning strategy information, machine learning algorithm information and subscription data information. The NF is used to receive the model training task issued by the public node, perform model training according to the model training task, report the trained model parameters to the public node, receive the model parameters issued by the public node, perform iterative learning according to the issued model parameters, and obtain the analysis model after reaching a preset number of iterations or a preset model accuracy. The process of obtaining the performance analysis results of the NF based on the analysis model obtained through iterative learning includes: The correlation degree between the service parameters of each NF and the KPI of the network slice subnet is obtained according to the analysis model, and the correlation degree is used as the performance analysis result. The operation and maintenance management of the network slice subnet based on the performance analysis results includes: Obtain the instantiated parameters corresponding to the service parameters in the network slice subnet, and optimize the instantiated parameters based on the correlation.

9. The network slice subnet operation and maintenance management system according to claim 8, characterized in that, The network slice subnet operation and maintenance management system also includes EM; The NF connects to the EM through its OAM channel, and the EM connects to the public node through its NAF module.

10. The network slice subnet operation and maintenance management system according to claim 8, characterized in that, The public node is NSSMF, and the NF uses a distributed machine learning algorithm for model training and inference.

11. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the network slice subnet operation and maintenance management method as described in any one of claims 1 to 6.

12. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the network slice subnet operation and maintenance management method as described in any one of claims 1 to 6.

Images