Network performance optimization method, device and storage medium

Abstract

The embodiment of the present application discloses a method for optimizing network performance and related equipment, which can realize online expansion and contraction of VNF and VNFC levels. The method includes: determining the CPU utilization corresponding to the current virtual resources of the network cloud; if the CPU utilization reaches the preset CPU utilization upper limit threshold, then determining the required number of virtual network element components VNFC corresponding to the virtual resources according to the relationship model and the CPU utilization, the relationship model is a model obtained by training network management performance data through a machine learning algorithm, and the relationship model is a relationship model between predicted user behavior data and virtual resource CPU utilization; determining the optimization configuration model of the virtual network element VNF ​​corresponding to the virtual resource according to the required number of VNFC; comparing the optimization configuration model with the current configuration model of the VNF to obtain a comparison result; adjusting the VNF according to the comparison result.

Description

Technical Field

[0001] The present application relates to the field of communications, and in particular to a method, device and storage medium for optimizing network performance. Background Art

[0002] As a new type of telecommunication network architecture, network cloud can provide a new service environment for traditional telecommunication services with elastically scalable resources, globally schedulable traffic, and fully open capabilities. Through virtualization (OpenStack as VIM, OpenStack is an open source cloud computing management platform project, a combination of a series of software open source projects, VIM is a text editor on the Linux system), cloud computing, network function virtualization (NFV) and other technologies to realize the cloudification of telecommunication services, and based on the intelligent management and orchestration system, service resources can be deployed on demand and flexibly scheduled in the resource pool. Network cloud provides bearer for telecommunication basic communication networks, adopts the NFV architecture defined by the ETSI standardization organization, and consists of three layers: hardware layer, virtual layer and virtual network element, as well as NFV (VNF is a virtualized network element that implements a certain network function unit) management and orchestration system (Management and Orchestration, MANO). MANO is an architecture framework for managing and coordinating VNFs and other software components, mainly including NFV orchestrator (NFVO), VNF manager (VNFManager, VNFM) and virtual resource manager (VIM). NFVO implements network services, virtual network lifecycle management and global resource management, and is the decision maker of cloud management; VNFM implements virtual network element lifecycle management and is the executor of VNF management; VIM is a virtualized infrastructure management system and is the executor of virtual resource and hardware resource management.

[0003] In the existing model, the main function of the cloud computing orchestrator (NFVO+) is the life cycle management of virtual network elements (VNFs) and network services, including VNFs, virtual resources, physical resources related resources, performance, alarms, three-layer alarm correlation and performance threshold alarm monitoring, etc. When managing the virtual resources required for the VNF life cycle, NFVO+ adopts an indirect mode to manage the virtual resources related to the VNF, and requests the VIM to allocate and recycle virtual computing, virtual storage, and virtual network resources.

[0004] Ideally, for an instantiated VNF / NS, NFVO+ can initiate elastic scaling control, and send elastic scaling commands to the VNFM corresponding to each VNF, which is responsible for execution, to achieve the goal of promoting network efficiency and reducing cost operations. However, in order to ensure the stable operation of the telecommunications network and the limitations of existing technologies, comprehensive risk and cost analysis, NFV cannot achieve online scaling operations at the VNF and VNFC levels. Summary of the invention

[0005] The embodiments of the present application provide a method, device and storage medium for optimizing network performance, which can realize online expansion and contraction operations at the VNF and VNFC levels, thereby achieving the purpose of dynamically optimizing network performance.

[0006] The first aspect of the present application provides a method for optimizing network performance, which may include:

[0007] Determine the CPU utilization corresponding to the current virtual resources of the network cloud;

[0008] Determine whether the CPU utilization reaches a preset CPU utilization upper limit threshold;

[0009] If so, the required number of virtual network element components VNFC corresponding to the virtual resource is determined according to the relationship model and the CPU utilization, wherein the relationship model is a model obtained by training the network management performance data through a machine learning algorithm, and the relationship model is a relationship model between the predicted user behavior data and the CPU utilization occupied by the virtual resource;

[0010] Determine an optimization configuration model of the virtual network element VNF ​​corresponding to the virtual resource according to the required number of the VNFC;

[0011] Comparing the optimized configuration model with the current configuration model of the VNF to obtain a comparison result;

[0012] The VNF is adjusted according to the comparison result.

[0013] In one possible design, the method further includes:

[0014] Acquire network management performance data within a preset time period, wherein the network management performance data includes user behavior data and CPU utilization corresponding to multiple VNFCs;

[0015] Preprocessing the network management performance data to obtain training data;

[0016] Iteratively training the initial relationship model according to the training data until a preset iteration termination condition is reached;

[0017] The initial relationship model that meets the preset iteration termination condition is determined as the relationship model.

[0018] In one possible design, adjusting the VNF according to the comparison result includes:

[0019] Determine whether the conditions for adjusting resource allocation are met;

[0020] If so, the capacity of the VNF is expanded or reduced according to the comparison result.

[0021] In a possible design, after the number of the current VNFCs is expanded or reduced according to the optimization strategy corresponding to the VNF, the method includes:

[0022] Collecting network management performance data corresponding to the VNFC after expansion or reduction;

[0023] Analyze the network management performance data corresponding to the VNFC after expansion or reduction based on the relationship model to obtain analysis results;

[0024] Execute corresponding operations according to the analysis results.

[0025] In one possible design, the method further includes:

[0026] Determine whether the number of iterations reaches a preset value, and if so, determine whether the preset iteration termination condition is met;

[0027] or,

[0028] It is determined whether the model parameters of the initial relationship model converge, and if so, whether the preset iteration termination condition is satisfied.

[0029] In a possible design, the comparison of the optimized configuration model with the current configuration model of the VNF to obtain a comparison result includes:

[0030] Determining a first virtual resource type corresponding to the optimization configuration model and an adjustment range corresponding to the first virtual resource type;

[0031] Determine a second virtual resource type corresponding to the current configuration model of the VNF;

[0032] The first virtual resource type and the adjustment range corresponding to the first virtual resource type are respectively compared with the second virtual resource type to obtain the comparison result.

[0033] A second aspect of the present application provides a network performance optimization device, including:

[0034] A determination unit, used to determine the CPU utilization corresponding to the current virtual resources of the network cloud;

[0035] A determination unit, configured to determine whether the CPU utilization reaches a preset CPU utilization upper limit threshold;

[0036] The determining unit is further configured to determine, if the CPU utilization reaches a preset CPU utilization upper limit threshold, the required number of virtual network element components VNFC corresponding to the virtual resource according to the relationship model and the CPU utilization, wherein the relationship model is a model obtained by training network management performance data through a machine learning algorithm, and the relationship model is a relationship model between predicted user behavior data and CPU utilization occupied by virtual resources;

[0037] The determining unit is further used to determine an optimized configuration model of the virtual network element VNF ​​corresponding to the virtual resource according to the required number of the VNFC;

[0038] A comparison unit, configured to compare the optimized configuration model with the current configuration model of the VNF to obtain a comparison result;

[0039] An adjusting unit is used to adjust the VNF according to the comparison result.

[0040] In one possible design, the device further includes:

[0041] A model training unit, wherein the model training unit is used to:

[0042] Acquire network management performance data within a preset time period, wherein the network management performance data includes user behavior data and CPU utilization corresponding to multiple VNFCs;

[0043] Preprocessing the network management performance data to obtain training data;

[0044] Iteratively training the initial relationship model according to the training data until a preset iteration termination condition is reached;

[0045] The initial relationship model that meets the preset iteration termination condition is determined as the relationship model.

[0046] In a possible design, the adjustment unit is specifically used for:

[0047] Determine whether the conditions for adjusting resource allocation have been met;

[0048] If so, the capacity of the VNF is expanded or reduced according to the comparison result.

[0049] In a possible design, the adjustment unit is further used for:

[0050] Collecting network management performance data corresponding to the VNFC after expansion or reduction;

[0051] Analyze the network management performance data corresponding to the VNFC after expansion or reduction based on the relationship model to obtain analysis results;

[0052] Execute corresponding operations according to the analysis results.

[0053] In a possible design, the model training unit is further used for:

[0054] Determine whether the number of iterations reaches a preset value, and if so, determine whether the preset iteration termination condition is met;

[0055] or,

[0056] It is determined whether the model parameters of the initial relationship model converge, and if so, whether the preset iteration termination condition is satisfied.

[0057] In a possible design, the comparison unit is specifically used for:

[0058] Determining a first virtual resource type corresponding to the optimization configuration model and an adjustment range corresponding to the first virtual resource type;

[0059] Determine a second virtual resource type corresponding to the current configuration model of the VNF;

[0060] The first virtual resource type and the adjustment range corresponding to the first virtual resource type are respectively compared with the second virtual resource type to obtain the comparison result.

[0061] A third aspect of the present application provides a server, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other through the communication bus;

[0062] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute the operation of the network performance optimization method.

[0063] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores at least one executable instruction. When the executable instruction is executed on a computing device, the computing device executes the network performance optimization method as described in the first aspect of the present application.

[0064] In a fifth aspect, the present application discloses a computer program product. When the computer program product is run on a computer, the computer is enabled to execute the network performance optimization method described in the first aspect of the present application.

[0065] The sixth aspect of the present application discloses an application publishing platform, which is used to publish a computer program product. When the computer program product runs on a computer, the computer executes the network performance optimization method described in the first aspect of the present application.

[0066] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0067] By judging the CPU utilization corresponding to the current virtual resources of the network cloud, and when the CPU utilization reaches the preset CPU utilization upper limit threshold, the required number of virtual network element components is determined according to the pre-trained relationship model between the predicted user behavior data and the CPU utilization occupied by virtual resources, and the optimization configuration model of the virtual network element is determined according to the required number of virtual network element components, and the current configuration model of the virtual network element is determined according to the optimization configuration model for comparison, and the comparison result is obtained, and the virtual network element is adjusted according to the comparison result. In this way, online expansion and contraction operations at the VNF and VNFC levels can be realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The accompanying drawings are only used to illustrate the embodiments and are not to be considered as limiting the present application. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings. In the accompanying drawings:

[0069] Figure 1 A network architecture diagram of a 5G core network provided in an embodiment of the present application;

[0070] Figure 2 A flowchart of a method for optimizing network performance in an embodiment of the present application;

[0071] Figure 3 A schematic diagram of a virtual structure of a network performance optimization device in an embodiment of the present application;

[0072] Figure 4 This is a schematic diagram of the structure of the server in the embodiment of the present application. DETAILED DESCRIPTION

[0073] In order to make the technical personnel in the technical field better understand the scheme of the present application, the technical scheme in the embodiment of the present application will be described below in conjunction with the drawings in the embodiment of the present application. Obviously, the described embodiment is only a part of the embodiment of the present application, not all of the embodiments. Based on the embodiments in the present application, they should all fall within the scope of protection of the present application.

[0074] In the early stage of NFV commercialization, most operators and manufacturers believed that elastic scaling of capacity was an important feature and the biggest driving force of NFV. Ideally, for an instantiated VNF / NS, NFVO+ can initiate elastic scaling control, and the elastic scaling command is sent to the VNFM corresponding to each VNF, which is responsible for execution, so as to achieve the purpose of promoting network efficiency and reducing cost operations. However, in order to ensure the stable operation of the telecommunications network and the limitations of existing technologies, comprehensive risk and cost analysis, NFV cannot achieve online scaling operations at the VNF and VNFC levels.

[0075] Since the construction of the cloud resource pool is not synchronized with the construction of VNF applications, the cloud resource pool needs to predict the demand and deploy it first. In addition, there are multiple cloud resource pool projects corresponding to the deployment of VNFs in multiple projects, which leads to differences in VNF resource demand predictions across time and projects, and the deployment of resources cannot be optimized.

[0076] In view of this, the embodiment of the present application provides a method for optimizing network performance and related equipment, by judging the CPU utilization corresponding to the current virtual resources of the network cloud, and when the CPU utilization reaches the preset CPU utilization upper limit threshold, the required number of virtual network element components is determined according to the relationship model between the predicted user behavior data and the CPU utilization occupied by the virtual resources that has been trained in advance, and the optimization configuration model of the virtual network element is determined according to the required number of virtual network element components, and the current configuration model of the virtual network element is determined according to the optimization configuration model for comparison, and the comparison result is obtained, and the virtual network element is adjusted according to the comparison result. This can solve the existing problem that online expansion and contraction operations at the VNF and VNFC levels cannot be realized.

[0077] The network performance optimization method and related equipment provided in this application can be applied to the optimization of 2G, 3G, 4G, 5G and IP Multimedia System (IMS) resource configuration models after cloudification, and of course can also be extended to the resource performance optimization of other cloud computing systems.

[0078] The following uses the 5G core network as an example for explanation. Figure 1 , Figure 1 A network architecture diagram of a 5G core network provided in an embodiment of the present application, wherein:

[0079] The 5G core network includes: control plane authentication service (Authentication Server Function, AUSF), access and mobility management function (access and mobility management function, AMF), policy control (PCF), session management (SMF), unified data management (Unified Data Management, UDM), charging management network element (CHF), network slice selection network element (Network Slice Selection Function, NSSF), unstructured data storage function (Unstructured Data Storage Function, UDSF), network storage function (NF Repository Function, NRF), unified data repository function (Unified Data Repository, UDR) and user plane network element (UserPlane Function, UPF) management, as follows:

[0080] 1. AUSF is used to receive AMF's request for identity authentication of the user equipment (UE), request a key from UDM, and then forward the key issued by UDM to AMF for authentication processing.

[0081] 2. AMF is the termination point of the RAN signaling interface (N2) and the termination point of the Network Attached Storage (NAS) (N1) signaling (MM message). It is responsible for the encryption and security of NAS messages, registration, access, mobility, authentication, and transparent transmission of SMS messages.

[0082] 3. PCF supports a unified policy framework to manage network behavior, provides policy rules to network entities for implementation, and accesses subscription information of the Universal Data Repository (UDR).

[0083] 4. SMF is the termination point of the SM message of the NAS message, manages the establishment, modification, release of the session, the allocation management of the UEIP, and implements the DHCP function, etc.;

[0084] 5. UDM is responsible for the management of user identification, contract data, authentication data, and user service network element registration management (such as AMF, SMF, etc. that currently provide services to the terminal. For example, when the user switches the AMF to access, UDM will also initiate a deregistration message to the old AMF, requiring the old AMF to delete user-related information);

[0085] 6. CHF is mainly composed of AGF, CDF and CGF, and supports three scenarios: online billing, offline billing and converged billing.

[0086] 7. The NSSF selects the set of network slice instances that serve the UE; determines the allowed NSSAI and, if necessary, the mapping to the subscribed S-NSSAI; determines the configured NSSAI and, if necessary, the mapping to the subscribed S-NSSAI; determines the set of AMFs that may be used to query the UE, or determines a list of candidate AMFs based on the configuration (possibly by querying the NRF);

[0087] 8. UDSF, any NF can store and retrieve its unstructured data;

[0088] 9. NRF is used for NF registration, management, and status detection to achieve automated management of all NFs;

[0089] 10. UDR stores and retrieves contract data, PCF stores and retrieves policy data, structured services stores and retrieves, NEF application data;

[0090] 11. UPF is responsible for data packet routing, forwarding, and Quality of Service (QoS) flow mapping.

[0091] The following is a detailed description of the network performance optimization method provided in this application from the perspective of the network performance optimization device.

[0092] See also Figure 2 , Figure 2 The schematic diagram of an embodiment of the method for optimizing network performance provided in the embodiment of the present application includes:

[0093] 201. Determine the CPU utilization corresponding to the current virtual resources of the network cloud.

[0094] In this embodiment, when optimizing the network performance of the network cloud, the network performance optimization device can determine the CPU utilization corresponding to the current virtual resources of the network cloud. The method of determining the CPU utilization is not specifically limited here. For example, the CPU utilization of each virtual resource in the network cloud can be monitored.

[0095] 202. Determine whether the CPU utilization reaches a preset CPU utilization upper limit threshold. If so, execute step 203.

[0096] In this embodiment, after determining the CPU utilization corresponding to the current virtual resources of the network cloud, the network performance optimization device can determine whether the CPU utilization reaches the preset CPU utilization upper limit threshold, and if the CPU utilization reaches the preset CPU utilization upper limit threshold, step 203 is executed. Among them, the preset CPU utilization upper limit threshold is the value at which the CPU utilization of the server corresponding to the network cloud reaches the bottleneck, and the value can be set according to a test or an empirical value, that is, if the CPU utilization corresponding to the virtual resources is used to determine whether the CPU utilization of the server corresponding to the network cloud has reached the bottleneck, if the CPU utilization has reached the bottleneck, step 203 is executed.

[0097] 203. Determine the required quantity of the virtual network element component VNFC corresponding to the virtual resource according to the relationship model and the CPU utilization.

[0098] In this embodiment, when the network performance optimization device determines that the CPU utilization reaches the preset CPU utilization upper limit threshold, it can determine the required number of virtual network element components VNFC corresponding to the virtual resources through the pre-trained relationship model between the predicted user behavior data and the CPU utilization occupied by the virtual resources, that is, determine the number of VNFCs that currently need to be expanded. The training method of the relationship model is described below:

[0099] Step 1: Obtain network management performance data within a preset time period, where the network management performance data includes user behavior data and CPU utilization corresponding to multiple VNFCs.

[0100] In this step, the collected network management performance data is described by taking the 5G core network as an example. For the 5G core network, the network management performance data includes but is not limited to those shown in Table 1. In addition to the network management performance data in Table 1, it also includes the CPU utilization statistics of each VNFC, such as PBU and OMU, etc. In addition, the collected network management performance data should be data for a longer period of time, such as more than 3 months or longer, for subsequent machine learning and modeling.

[0101] Table 1

[0102]

[0103]

[0104] Step 2: Preprocess the network management performance data to obtain training data.

[0105] Step 3: Iteratively train the initial relationship model according to the training data until a preset iteration termination condition is reached.

[0106] In this step, after obtaining the network management performance data, the network performance optimization device can pre-process the network management performance data, and iteratively train the pre-processed data through a machine learning algorithm until the preset iteration termination condition is reached. In this application, the network performance optimization device uses a long short-term memory (LSTM) neural network algorithm to model and predict network management performance data. LSTM is mainly used to model and predict the future of time-related events, and its application areas include text generation, machine translation, speech recognition, fault diagnosis, and traffic prediction. In this application, the performance data collected by the network management of the 5G core network, that is, user behavior statistics and CPU utilization data counted by the virtual network element component (VNFC), are used. The LSTM neural network algorithm is used for modeling and prediction, and a relationship model between user behavior data and resource occupancy CPU utilization is established, and historical data is used to predict the future to form CPU utilization requirements.

[0107] Step 4: Determine the initial relational model that meets the preset iteration termination condition as the relational model.

[0108] In this step, during the model training process through LSTM, after each iteration, it can be determined whether the number of iterations reaches the preset value. If so, it is determined that the preset iteration termination condition is met; or, it is determined whether the model parameters of the initial relationship model converge. If so, it is determined that the preset iteration termination condition is met, and the initial relationship model that meets the preset iteration termination condition is determined as the relationship model.

[0109] 204. Determine an optimization configuration model of the virtual network element VNF ​​corresponding to the virtual resources according to the required number of VNFCs.

[0110] In this embodiment, after determining the required number of VNFCs, the network performance optimization device can determine the optimal configuration model of the VNF corresponding to the virtual resource according to the required number of VNFCs, and determine the required number of VNFCs by CPU requirements to achieve the purpose of optimizing the VNF configuration. Among them, NFV is a virtualization technology or concept that solves the problem of deploying network functions on general hardware. VNF refers to a specific virtual network function that provides a certain network service. It is software and is deployed in a virtual machine, container or bare-metal physical machine using the infrastructure provided by NFVI.

[0111] 205. Compare the optimized configuration model with the current configuration model of the VNF to obtain a comparison result.

[0112] In this embodiment, after obtaining the optimized configuration model, the network performance optimization device can compare the optimized configuration model with the current configuration model of the VNF to obtain a comparison result. When performing the comparison, the first virtual resource type corresponding to the optimized configuration model and the adjustment range corresponding to the first virtual resource type can be determined, and the second virtual resource type corresponding to the current configuration model of the VNF can be determined, and the first virtual resource type and the adjustment range corresponding to the first virtual resource type can be compared with the second virtual resource type to obtain the comparison result, which is also the comparison of resource types and their adjustment ranges. That is, the resource types that need to be adjusted and the adjustment ranges corresponding to the resource types are determined.

[0113] 206. Adjust the VNF according to the comparison result.

[0114] In this embodiment, after obtaining the comparison results, the network performance optimization device can adjust the VNF according to the comparison results. Specifically, it can determine whether the resource configuration adjustment conditions are met. The resource configuration adjustment conditions can be manually determined whether to adjust, or automatically determined whether to adjust the resource configuration. If so, the VNF is expanded or reduced in capacity according to the comparison results. Of course, the VNFC can also be expanded or reduced in capacity.

[0115] It should be noted that when the decision is to adopt the optimized configuration model, the optimized configuration model is issued through the cloud resource management pool network management system (such as NFVO+) and instantiated and deployed in the cloud resource pool.

[0116] In addition, the network performance optimization device can also collect the network management performance data corresponding to the VNFC after expansion or reduction; analyze the network management performance data corresponding to the VNFC after expansion or reduction based on the relationship model to obtain the analysis result; and perform corresponding operations according to the analysis result. That is, after the optimized NFV network instance is deployed, the corresponding network management performance management data is collected, and the network management performance management data is analyzed and predicted based on the network management performance management data to form a closed-loop process, ensure the effectiveness of the optimization configuration model, and make dynamic adjustments as the business volume and user behavior change.

[0117] In summary, in the embodiment provided by the present application, by judging the CPU utilization rate corresponding to the current virtual resources of the network cloud, and when the CPU utilization rate reaches the preset CPU utilization rate upper limit threshold, the required number of virtual network element components is determined according to the relationship model between the pre-trained predicted user behavior data and the CPU utilization rate occupied by the virtual resources, and the optimization configuration model of the virtual network element is determined according to the required number of virtual network element components, and the current configuration model of the virtual network element is determined according to the optimization configuration model for comparison, and the comparison result is obtained, and the virtual network element is adjusted according to the comparison result. In this way, online expansion and contraction operations at the VNF and VNFC levels can be realized.

[0118] The above describes the embodiment of the present application from the perspective of the network performance optimization method. The following describes the embodiment of the present application from the perspective of the network performance optimization device:

[0119] See also Figure 3 , Figure 3 This is a schematic diagram of an embodiment of a network performance optimization device provided in an embodiment of the present application. The network performance optimization device 300 includes:

[0120] The determining unit 301 is used to determine the CPU utilization corresponding to the current virtual resources of the network cloud;

[0121] The judging unit 302 is used to judge whether the CPU utilization reaches a preset CPU utilization upper limit threshold;

[0122] The determining unit 301 is further configured to determine the required number of virtual network element components VNFC corresponding to the virtual resource according to the relationship model and the CPU utilization if the CPU utilization reaches a preset CPU utilization upper limit threshold, wherein the relationship model is a model obtained by training network management performance data through a machine learning algorithm, and the relationship model is a relationship model between predicted user behavior data and CPU utilization occupied by virtual resources;

[0123] The determining unit 301 is further configured to determine an optimized configuration model of the virtual network element VNF ​​corresponding to the virtual resource according to the required number of the VNFC;

[0124] A comparing unit 303 is used to compare the optimized configuration model with the current configuration model of the VNF to obtain a comparison result;

[0125] The adjusting unit 304 is used to adjust the VNF according to the comparison result.

[0126] In one possible design, the device further includes:

[0127] The model training unit 305 is used for:

[0128] Acquire network management performance data within a preset time period, wherein the network management performance data includes user behavior data and CPU utilization corresponding to multiple VNFCs;

[0129] Preprocessing the network management performance data to obtain training data;

[0130] Iteratively training the initial relationship model according to the training data until a preset iteration termination condition is reached;

[0131] The initial relationship model that meets the preset iteration termination condition is determined as the relationship model.

[0132] In a possible design, the adjustment unit 304 is specifically used for:

[0133] Determine whether the conditions for adjusting resource allocation have been met;

[0134] If so, the capacity of the VNF is expanded or reduced according to the comparison result.

[0135] In a possible design, the adjusting unit 304 is further configured to:

[0136] Collecting network management performance data corresponding to the VNFC after expansion or reduction;

[0137] Analyze the network management performance data corresponding to the VNFC after expansion or reduction based on the relationship model to obtain analysis results;

[0138] Execute corresponding operations according to the analysis results.

[0139] In a possible design, the model training unit 305 is further used for:

[0140] Determine whether the number of iterations reaches a preset value, and if so, determine whether the preset iteration termination condition is met;

[0141] or,

[0142] It is determined whether the model parameters of the initial relationship model converge, and if so, whether the preset iteration termination condition is satisfied.

[0143] In a possible design, the comparison unit 303 is specifically used for:

[0144] Determining a first virtual resource type corresponding to the optimization configuration model and an adjustment range corresponding to the first virtual resource type;

[0145] Determine a second virtual resource type corresponding to the current configuration model of the VNF;

[0146] The first virtual resource type and the adjustment range corresponding to the first virtual resource type are respectively compared with the second virtual resource type to obtain the comparison result.

[0147] The embodiment of the present application also provides a server, which is used to deploy a network performance optimization device and execute the above-mentioned network performance optimization method. The server includes a processor, a memory, a communication interface and a communication bus. The processor, the memory and the communication interface communicate with each other through the communication bus; the memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute the operation of the above-mentioned relay node selection method. For details, please refer to Figure 4 , Figure 4 4 is a schematic diagram of the structure of a server provided in an embodiment of the present application. The server 400 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 422 (for example, one or more processors) and a memory 432, and one or more storage media 430 (for example, one or more mass storage devices) storing application programs 442 or data 444. Among them, the memory 432 and the storage medium 430 may be temporary storage or permanent storage. The program stored in the storage medium 430 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the server. Furthermore, the central processing unit 422 may be configured to communicate with the storage medium 430 to execute a series of instruction operations in the storage medium 430 on the server 400.

[0148] The server 400 may also include one or more power supplies 426, one or more wired or wireless network interfaces 450, one or more input and output interfaces 458, and / or one or more operating systems 441, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0149] The steps performed by the network performance optimization device in the above embodiment can be based on the Figure 4 The server structure shown.

[0150] The present application also provides a computer-readable storage medium, in which at least one executable instruction is stored. When the executable instruction is executed on a computing device, the computing device executes the network performance optimization method described in any of the above embodiments.

[0151] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0152] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, a computer, a server, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or a data center that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

[0153] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0154] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0155] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0156] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0157] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0158] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for optimizing network performance, characterized in that: include: Determine the CPU utilization corresponding to the current virtual resources of the network cloud; Determine whether the CPU utilization reaches a preset CPU utilization upper limit threshold; If so, the required number of virtual network element components VNFC corresponding to the virtual resource is determined according to the relationship model and the CPU utilization, wherein the relationship model is a model obtained by training the network management performance data through a machine learning algorithm, and the relationship model is a relationship model between the predicted user behavior data and the CPU utilization occupied by the virtual resource; the network management performance data includes the user behavior data and the CPU utilization corresponding to the plurality of VNFCs; Determine an optimization configuration model of the virtual network element VNF ​​corresponding to the virtual resource according to the required number of the VNFC; Comparing the optimized configuration model with the current configuration model of the VNF to obtain a comparison result, including: determining a first virtual resource type corresponding to the optimized configuration model and an adjustment range corresponding to the first virtual resource type; determining a second virtual resource type corresponding to the current configuration model of the VNF; comparing the first virtual resource type and the adjustment range corresponding to the first virtual resource type with the second virtual resource type, respectively, to obtain the comparison result; the comparison result is a comparison of resource types and their adjustment ranges; The VNF is adjusted according to the comparison result, including: determining whether a resource configuration adjustment condition is met, the resource configuration adjustment condition being manually determined whether to make an adjustment, or automatically determined whether to make an adjustment to the resource configuration, and if so, the VNF is expanded or reduced in capacity according to the comparison result.

2. The method according to claim 1, characterized in that The method further comprises: Acquire network management performance data within a preset time period, wherein the network management performance data includes user behavior data and CPU utilization corresponding to multiple VNFCs; Preprocessing the network management performance data to obtain training data; Iteratively training the initial relationship model according to the training data until a preset iteration termination condition is reached; The initial relationship model that meets the preset iteration termination condition is determined as the relationship model.

3. The method according to claim 1, characterized in that After the current number of VNFCs is expanded or reduced according to the optimization strategy corresponding to the VNF, the method includes: Collecting network management performance data corresponding to the VNFC after expansion or reduction; Analyze the network management performance data corresponding to the VNFC after expansion or reduction based on the relationship model to obtain analysis results; Execute corresponding operations according to the analysis results.

4. The method according to claim 2, characterized in that: The method further comprises: Determine whether the number of iterations reaches a preset value, and if so, determine whether the preset iteration termination condition is met; or, It is determined whether the model parameters of the initial relationship model converge, and if so, whether the preset iteration termination condition is satisfied.

5. A network performance optimization device, characterized in that: include: A determination unit, used to determine the CPU utilization corresponding to the current virtual resources of the network cloud; A determination unit, configured to determine whether the CPU utilization reaches a preset CPU utilization upper limit threshold; The determining unit is further configured to determine the required number of virtual network element components VNFC corresponding to the virtual resource according to the relationship model and the CPU utilization if the CPU utilization reaches a preset CPU utilization upper limit threshold, wherein the relationship model is a model obtained by training the network management performance data through a machine learning algorithm, and the relationship model is a relationship model between the predicted user behavior data and the CPU utilization occupied by the virtual resource; the network management performance data includes the user behavior data and the CPU utilization corresponding to the plurality of VNFCs; The determining unit is further used to determine an optimized configuration model of the virtual network element VNF ​​corresponding to the virtual resource according to the required number of the VNFC; A comparison unit, used for comparing the optimized configuration model with the current configuration model of the VNF to obtain a comparison result, including: determining a first virtual resource type corresponding to the optimized configuration model and an adjustment range corresponding to the first virtual resource type; determining a second virtual resource type corresponding to the current configuration model of the VNF; comparing the first virtual resource type and the adjustment range corresponding to the first virtual resource type with the second virtual resource type respectively to obtain the comparison result; the comparison result is a comparison of resource types and their adjustment ranges; An adjustment unit is used to adjust the VNF according to the comparison result, including: judging whether a resource configuration adjustment condition is met, the resource configuration adjustment condition is manually determining whether to adjust, or automatically deciding whether to adjust the resource configuration, and if so, expanding or reducing the capacity of the VNF according to the comparison result.

6. The device according to claim 5, characterized in that The device also includes: A model training unit, wherein the model training unit is used to: Acquire network management performance data within a preset time period, wherein the network management performance data includes user behavior data and CPU utilization corresponding to multiple VNFCs; Preprocessing the network management performance data to obtain training data; Iteratively training the initial relationship model according to the training data until a preset iteration termination condition is reached; The initial relationship model that meets the preset iteration termination condition is determined as the relationship model.

7. A server, characterized in that: include: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute the operation of the network performance optimization method as described in any one of claims 1-4.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one executable instruction. When the executable instruction is executed on a computing device, the computing device executes the method for optimizing network performance according to any one of claims 1 to 4.

Images