WO2022110974A1

WO2022110974A1 - Method and apparatus for training data analysis model, and storage medium

Info

Publication number: WO2022110974A1
Application number: PCT/CN2021/117741
Authority: WO
Inventors: 夏海涛
Original assignee: 华为技术有限公司
Priority date: 2020-11-27
Filing date: 2021-09-10
Publication date: 2022-06-02
Also published as: CN114567559A

Abstract

The present application relates to the technical field of communications, and in particular, to a method and apparatus for training a data analysis model, and a storage medium. The method is applied to an MDA entity. The method comprises: receiving a notification message, the notification message carrying information required for a specified analysis topic related to an NS; and training a pre-trained basic data analysis model according to the notification message, to obtain a target data analysis model corresponding to the specified analysis topic, the basic data analysis model being obtained by training an original model according to configuration data of an NFV object. In embodiments of the present application, an MDA entity pre-trains a basic data analysis model, such that the basic data analysis model is reused during model training of a topic-specific target data analysis model, thereby shortening the duration of data analysis model training and increasing the model training efficiency.

Description

Training method, device and storage medium for data analysis model

This application claims the priority of the Chinese patent application with the application number 202011358153.5 and the application title "Data Analysis Model Training Method, Device and Storage Medium" filed with the China Patent Office on November 27, 2020, the entire contents of which are incorporated by reference in this application.

technical field

The present application relates to the field of communication technologies, and in particular, to a training method, device and storage medium for a data analysis model.

Background technique

The Management Data Analytics (MDA) entity refers to management services that use management analytics data in network and service management. Raw performance data for network functions can be analyzed together with other management data (eg, alarm data, configuration data) and form management analysis data for one or more network functions, sub-networks, or network slices/sub-network slice instances. MDA provides the ability to process and analyze raw data related to network and service events and status to provide analytical reports to support necessary operations for network and service management.

Currently, MDA can integrate artificial intelligence (AI) and/or machine learning (ML) capabilities to bring intelligence and automation to network service management and orchestration. For example, for a specific analysis topic, the MDA entity collects relevant network information, and analyzes the collected network information according to the built-in data analysis model corresponding to the analysis topic (such as the analysis model implemented by AI algorithm) to obtain the analysis result. , and return the analysis results to the Network Functions Virtualization Orchestrator (NFVO) that initiated the topic analysis, that is, the service consumer, to enhance NFVO to make closed-loop decisions in the Network Function Virtualization (NFV) management domain Ability.

In the above method, the data analysis models used by different analysis topics are different, and the training of data analysis models corresponding to different analysis topics needs to derive various associations required for data analysis from scratch, and the model training efficiency is low. .

SUMMARY OF THE INVENTION

In view of this, a training method, device and storage medium for a data analysis model are proposed. The MDA entity pre-trains a basic data analysis model according to the configuration data of the NFV object, so that the MDA entity receives the data carrying the network service (Network Service). , NS) related to the notification information of the information required by the specified analysis topic, the target data analysis model corresponding to the specified analysis topic can be further trained on the basis of the basic data analysis model, that is, the target data analysis models corresponding to different analysis topics are in The pre-trained basic data analysis model can be reused during training, which can improve the training efficiency of the data analysis model.

In the first aspect, an embodiment of the present application provides a training method for a data analysis model, which is used in an MDA entity, and the method includes:

Receive a notification message, which carries the information required by the specified analysis topic related to the NS;

According to the notification information, the pre-trained basic data analysis model is trained to obtain the target data analysis model corresponding to the specified analysis topic. The basic data analysis model is obtained by training the original model according to the configuration data of the NFV object. The NFV object is managed by NFV Managed objects within the domain.

In this implementation, the MDA entity is pre-trained to complete a basic data analysis model, and on the basis of the basic data analysis model, the MDA entity receives the notification information that carries the information required by the specified analysis topic related to the NS, Further training completes the target data analysis model corresponding to the specified analysis topic, that is, the basic data analysis model is reused by the topic-specific target data analysis model during model training, avoiding the training of data analysis models corresponding to different analysis topics in related technologies. When it is necessary to derive various associations required for data analysis from scratch, the training time of the data analysis model is shortened and the model training efficiency is improved.

With reference to the first aspect, in a first possible implementation manner of the first aspect, a basic data analysis model is used to indicate an attribute of an NFV object and an association relationship between the NFV object.

In this implementation manner, the basic data analysis model pre-trained by the MDA entity can maintain the attributes of the NFV objects and the association relationship between the NFV objects.

With reference to the first aspect, in a second possible implementation manner of the first aspect, the method further includes: performing an analysis on the original model according to the descriptor template information of the imported NFV object and/or the image information after the instantiation of the NFV object. The basic data analysis model is obtained by training.

In this implementation, the MDA entity takes the information model of the NFV object in the design state (that is, the descriptor template information of the NFV object) and the information model of the NFV object in the running state (that is, the image information after the instantiation of the NFV object) as the basic data Analyze the input data of model training, so as to train a basic data analysis model irrelevant to the analysis topic, so as to enhance the ability of the MDA entity to dynamically obtain the association relationship between NFV objects in the NFV management domain, and improve the subsequent analysis model based on the basic data. Efficiency of training the target data analysis model.

With reference to the second possible implementation manner of the first aspect, in the third possible implementation manner of the first aspect, the descriptor template information includes network service descriptor (Network Service Descriptor, NSD) template information and/or virtualization Network function descriptor (Virtualized Network Function Descriptor, VNFD) template information; and/or, the image information includes NS instance image information and/or Virtualized Network Function (Virtualized Network Function, VNF) instance image information.

In this implementation, the input data for training the basic data analysis model may further include at least one of NSD template information, VNFD template information, NS instance mirroring information, and VNF instance mirroring information, which further ensures the training effect of the basic data analysis model.

In combination with the third possible implementation manner of the first aspect, in the fourth possible implementation manner of the first aspect, the method further includes: during the training process of the basic data analysis model, establishing NSD template information and members of the NS The relationship between the object's descriptor template information.

In this implementation, in the stage of putting NSD and/or VNF data packets on the shelf, the MDA entity trains the basic data analysis model, and establishes an association relationship between the NSD template information and the descriptor template information of the member objects of the NS, so that the The trained basic data analysis model can indicate relatively static associations between object classes, which further improves the efficiency of subsequent training of the target data analysis model based on the basic data analysis model.

With reference to the third possible implementation manner of the first aspect, in the fifth possible implementation manner of the first aspect, the method further includes: in the training process of the basic data analysis model, establishing NS instance mirror information and NS instance image information The association relationship between the mirror information of the member object instance.

In this implementation, in the NS and/or VNF instantiation stage, the MDA entity performs basic data analysis model training, and establishes an association relationship between the mirror information of the NS instance and the mirror information of the member object instances of the NS instance, so that pre-training The completed basic data analysis model can indicate relatively dynamic associations between object instances, which further improves the efficiency of subsequent training of the target data analysis model based on the basic data analysis model.

With reference to the third possible implementation manner of the first aspect, in the sixth possible implementation manner of the first aspect, the method further includes:

The basic data analysis model is updated according to the modified NS instance image information and/or the modified VNF instance image information.

In this implementation, in the running state after the NS and/or VNF are instantiated, the MDA entity updates the basic data analysis model according to the modified NS instance image information and/or the modified VNF instance image information, so as to dynamically Adjust the basic data analysis model.

In combination with the first aspect and any possible implementation manner of the first aspect, in a seventh possible implementation manner of the first aspect, the method further includes:

Input the NS-related performance data and/or alarm data into the target data analysis model corresponding to the NS alarm event analysis theme, and output to obtain a first analysis result, where the first analysis result includes the fundamental alarm and/or the root cause of the NS failure; or,

Input the information required for the NS health degree analysis into the target data analysis model corresponding to the NS health degree analysis subject, and output the second analysis result, where the second analysis result includes the health state of the NS and the description information corresponding to the health state; or ,

The information required for the NS resource utilization analysis is input into the target data analysis model corresponding to the NS resource utilization analysis theme, and a third analysis result is output, and the third analysis result includes an indication of the resource utilization of the NS.

In this implementation, three possible implementations are provided for the process of MDA entity invoking the target data analysis model corresponding to the specified analysis topic for data analysis. The specified analysis topic includes but is not limited to NS alarm event analysis topic, NS health degree analysis Either the topic or the NS resource utilization analysis topic, realizes the effective analysis of the data under the specified analysis topic, and ensures the accuracy of the data analysis.

In a second aspect, an embodiment of the present application provides a training device for a data analysis model, which is used in an MDA entity, and the device includes:

a receiving unit, configured to receive a notification message, where the notification message carries the information required by the specified analysis topic related to the NS;

The processing unit is used to train the pre-trained basic data analysis model according to the notification information to obtain the target data analysis model corresponding to the specified analysis topic. The basic data analysis model is obtained by training the original model according to the configuration data of the NFV object, NFV objects are managed objects in the NFV management domain.

With reference to the second aspect, in a first possible implementation manner of the second aspect, the basic data analysis model is used to indicate the attributes of the NFV object and the association relationship between the NFV objects.

With reference to the second aspect, in a second possible implementation manner of the second aspect, the processing unit is further configured to:

According to the descriptor template information of the imported NFV object and/or the image information after the instantiation of the NFV object, the original model is trained to obtain a basic data analysis model.

With reference to the second possible implementation manner of the second aspect, in a third possible implementation manner of the second aspect, the descriptor template information includes NSD template information and/or VNFD template information; and/or, the mirroring information includes NS Instance image information and/or VNF instance image information.

With reference to the third possible implementation manner of the second aspect, in the fourth possible implementation manner of the second aspect, the processing unit is further configured to:

During the training process of the basic data analysis model, the association relationship between the NSD template information and the descriptor template information of the member objects of the NS is established.

With reference to the third possible implementation manner of the second aspect, in a fifth possible implementation manner of the second aspect, the processing unit is further configured to:

In the training process of the basic data analysis model, an association relationship between the mirror information of the NS instance and the mirror information of the member object instances of the NS instance is established.

With reference to the third possible implementation manner of the second aspect, in the sixth possible implementation manner of the second aspect, the processing unit is further configured to:

In combination with the second aspect and any possible implementation manner of the second aspect, in a seventh possible implementation manner of the second aspect, the processing unit is further configured to:

In a third aspect, embodiments of the present application provide a data analysis model training apparatus for use in an MDA entity, the apparatus comprising: a processor; a memory for storing executable instructions of the processor; wherein the processor is configured A training method for a data analysis model provided for implementing the first aspect or any one of the possible implementation manners of the first aspect when executing an instruction.

In a fourth aspect, embodiments of the present application provide a computer program product, comprising computer-readable codes, or a non-volatile computer-readable storage medium carrying computer-readable codes, when the computer-readable codes are stored in an electronic When running in the device, the processor in the electronic device executes the data analysis model training method provided by the first aspect or any one of the possible implementations of the first aspect.

In a fifth aspect, embodiments of the present application provide a non-volatile computer-readable storage medium on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the first aspect or the first aspect is implemented. The training method of the data analysis model provided by any possible implementation.

Description of drawings

FIG. 1 shows a schematic diagram of the basic framework of 5G network automation in the related art.

FIG. 2 shows a schematic structural diagram of an MDA entity provided by an exemplary embodiment of the present application.

FIG. 3 shows a flowchart of a training method for a data analysis model provided by an exemplary embodiment of the present application.

FIG. 4 shows a flowchart of the training and use process of the data analysis model provided by an exemplary embodiment of the present application.

FIG. 5 shows a flowchart of a training method for a data analysis model provided by another exemplary embodiment of the present application.

FIG. 6 shows a flowchart of a training method for a data analysis model provided by another exemplary embodiment of the present application.

FIG. 7 shows a flowchart of a training method for a data analysis model provided by another exemplary embodiment of the present application.

FIG. 8 shows a flowchart of a training method for a data analysis model provided by another exemplary embodiment of the present application.

FIG. 9 shows a flowchart of a training method for a data analysis model provided by another exemplary embodiment of the present application.

FIG. 10 shows a flowchart of the training and use process of the data analysis model provided by another exemplary embodiment of the present application.

FIG. 11 shows a block diagram of an apparatus for training a data analysis model provided by an exemplary embodiment of the present application.

Detailed ways

Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numbers in the figures denote elements that have the same or similar functions. While various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise indicated.

The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

In addition, in order to better illustrate the present application, numerous specific details are given in the following detailed description. It should be understood by those skilled in the art that the present application may be practiced without certain specific details. In some instances, methods, means, components and circuits well known to those skilled in the art have not been described in detail so as not to obscure the subject matter of the present application.

In the research of network function automation for 5th generation mobile networks (5G), the network data analysis function (Network Data Analytics Function, SBA) is introduced into the 5G core network service-based architecture (SBA). NWDAF), as shown in Figure 1. The basic principle of network function automation includes: the NWDAF entity 10 subscribes the input data required for network data analysis to the network function (or service) entity of the surrounding control plane, and the analysis algorithm is executed by the NWDAF entity 10 itself (usually, the analysis algorithm is considered as artificial intelligence algorithm), distribute the analysis results to other network function (or service) entities, such as Operation Administration and Maintenance (OAM) entity 11, Application Function (AF) entity 12, network function (Network Function, NF) entity 13. Another example is the Policy Control Function (PCF) entity to help it achieve higher-order policy issuance and execution. The use case description and analysis in the related art focus on the input data and output data of the NWDAF entity, and the analysis algorithms within the NWDAF entity are not within the scope of standardization.

Currently, the Management Data Analytics Service (MDAS) has been introduced in the 5G management plane. Its function is similar to that of NWDAF. The core function of MDAS is to manage data analysis. According to the built-in artificial intelligence and/or machine learning data analysis of this function The model analyzes the collected network information related to a specific analysis topic, returns the analysis result to the service consumer who initiated the analysis of the specific topic, and assists it in closed-loop decision-making on the management level.

Before calling the data analysis model, the data analysis model also needs to be trained. The goal of the training process of the data analysis model is not to provide analysis results for the topic to be analyzed at the moment, but to make the data analysis model provide analysis results that are more accurate and more in line with the expectations of service consumers in subsequent use, and enhance the data through training input data. Analytical accuracy of the analytical model. In the related art, there are many studies on the input/output information of management data analysis, but there are few studies on the input/output information flow in the training process of the data analysis model.

At present, the function of Management Data Analytics (MDA) has been introduced into the NFV management domain, and its function is similar to that of MDAS. NFVO is enhanced by the interaction of the NFVO entity (that is, the consumer of the MDA function) and the data analysis process of the MDA entity. The ability to make closed-loop decisions within the NFV management domain.

In the related art, the input data set used in the training process of the data analysis model usually includes historical performance data, alarm data and configuration data generated by the network. Regardless of whether it is historically generated or real-time collected performance data or alarm data, these data are used as incremental input information to reinforce the original model in the training of the data analysis model, that is, the data are processed based on probability statistics or knowledge. The training and analysis of the model is based on the original model to add new relationships of managed objects or relationships of management events. For the model training of different analysis topics, the set of performance or alarm events subscribed by MDA entities from the NFV management domain is also different.

The establishment of the original model mainly depends on the configuration data of managed objects, and performance data or alarm data are usually attached to these managed objects. For example, the alarm data "The IP address of the virtual machine A is unreachable" is attached to the managed object virtual machine A.

In order to improve the training efficiency of the data analysis model, shorten the model training time, and avoid the data analysis model training of different analysis topics to derive various associations required for data analysis from scratch, the embodiment of the present application provides a training method for the data analysis model In the method, device and storage medium, the MDA entity pre-trains a basic data analysis model, and on the basis of the basic data analysis model, the MDA entity receives the notification information that carries the information required by the designated analysis topic related to the NS , and further train to complete the target data analysis model corresponding to the specified analysis topic, that is, the basic data analysis model is reused by the topic-specific target data analysis model during model training, avoiding the training of data analysis models corresponding to different analysis topics in related technologies. It is necessary to derive various associations required for data analysis from scratch, which shortens the training time of the data analysis model and improves the model training efficiency.

For the convenience of understanding, the terms involved in the embodiments of the present application are explained below.

The information model of the NFV object in the design state: including the descriptor template information of the NFV object. Optionally, the descriptor template information of the NFV object in the NFV management domain includes, but is not limited to: at least one of NSD template information, VNFD template information, and Virtualization Resource Descriptor (VR Descriptor) template information.

NSD template information: used to describe the virtualized resources used by the NS and define the behavior requirements of the NS during the deployment and operation phases. Usually based on the descriptor definitions of the constituent members of the NS (eg, VNFs).

VNFD template information: It is used to describe the virtualized resources used by the VNF and define the behavior requirements of the VNF during the deployment and operation phases. The description of the requirements of the VNF for virtualized resources can be mapped to the descriptor definitions of the virtualized resources. For example, the VNFD template information includes descriptor template information of virtual computing resources, virtual storage resources, virtual links, connection points or other virtual network resources. This embodiment of the present application does not limit this.

Virtualization resource descriptor template information: used to describe the characteristics of a single virtualized resource. For example, the virtualized resource descriptor template information includes description information of virtual computing resources, virtual storage resources, virtual network resources and/or the virtualized resource descriptor. Description information of the demand of the physical resources on the underlying physical resources.

The information model of the NFV object in the running state: including the image information after the instantiation of the NFV object, and the image information includes but is not limited to: VNF instance image information (VNFInfo) and/or NS instance image information (NSInfo).

VNF instance image information: The image information of the VNF instance after the VNF is instantiated. Optionally, the VNF instance image information includes basic information of the VNF instance in the running state for lifecycle management.

NS instance image information: The image information of the NS instance after NS is instantiated. Optionally, the NS instance image information includes basic information when the NS instance performs life cycle management in the running state.

It should be noted that, for some related terms involved in the embodiments of the present application, reference may be made to the corresponding related descriptions in the 3GPP protocol or the ETSI NFV protocol, which will not be repeated herein.

Please refer to FIG. 2 , which shows a schematic structural diagram of an MDA entity provided by an exemplary embodiment of the present application. The MDA entity includes: a processor 21, a receiver 22, a transmitter 23, a memory 24, and a bus 25.

The processor 21 includes one or more processing cores, and the processor 21 executes various functional applications and information processing by running software programs and modules.

The receiver 22 and the transmitter 23 can be implemented as a communication component, which can be a communication chip, and the communication chip can include a receiving module, a transmitting module, a modulation and demodulation module, etc., for modulating and demodulating the information, and This information is received or transmitted via wireless signals.

The memory 24 is connected to the processor 21 through a bus 25 . The memory 54 stores program instructions and data necessary for the terminal.

The processor 51 is configured to execute the program instructions and data in the memory 54 to implement the functions of each step in each method embodiment of the present application.

The processor 21 controls the receiver 22 to implement the following step 401 and the implicit receiving function of the MDA entity side in each step by running at least one program instruction in the memory 24; the processor 21 runs at least one program instruction in the memory 24. , the transmitter 23 is controlled to implement the sending function on the MDA entity side implicit in each step of the embodiment of the present application.

Additionally, memory 24 may be implemented by any type of volatile or non-volatile storage device or combination thereof, such as static anytime access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable Except programmable read only memory (EPROM), programmable read only memory (PROM), read only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

It will be appreciated that Figure 2 only shows a simplified design of the MDA entity. In other embodiments, the MDA entity may include any number of transmitters, receivers, processors, controllers, memories, communication units, etc., and all MDA entities that can implement the present application are within the protection scope of the present application .

Please refer to FIG. 3 , which shows a flowchart of a training method for a data analysis model provided by an exemplary embodiment of the present application, and the method is used in the MDA entity shown in FIG. 2 . The method includes the following steps.

Step 301: Receive a notification message, where the notification message carries information required for a specified analysis topic related to the NS.

The MDA entity receives notification messages that carry the information required for the specified analysis topic related to the NS.

Optionally, for the specified analysis topic, the MDA entity subscribes the target entity to the information required by the specified analysis topic related to the NS; the target entity sends a notification message carrying the information required by the specified analysis topic related to the NS to the MDA entity. Correspondingly, the MDA entity receives the notification message.

Optionally, the target entity includes NFVO, Virtualised Network Function Manager (Virtualised Network Function Manager, VNFM), Virtualised Infrastructure Manager (Virtualised Infrastructure Manager, VIM), Container Infrastructure Service Management (Container Infrastructure Service Management, CISM) ) any of the entities.

Optionally, the specified analysis topic includes any one of an NS alarm event analysis topic, an NS health degree analysis topic, and an NS resource utilization analysis topic.

The information required for the specified analysis topic is the input data to be trained, that is, the input data required for training the target data analysis model corresponding to the specified analysis topic. For example, the specified analysis topic is an NS alarm event analysis topic, and the information required by the specified analysis topic includes NS-related performance data and/or alarm data.

It should be noted that the embodiments of the present application do not limit the type of the specified analysis topic and the information content required for the specified analysis topic.

Step 302: According to the notification information, the pre-trained basic data analysis model is trained to obtain the target data analysis model corresponding to the specified analysis topic. The basic data analysis model is obtained by training the original model according to the configuration data of the NFV object. It is the managed object in the NFV management domain.

The MDA entity obtains the information required by the NS-related designated analysis topic carried in the notification information, and retrains the pre-trained basic data analysis model according to the information required by the NS-related designated analysis topic, and obtains the corresponding data of the designated analysis topic. Target data analysis model.

Among them, the MDA entity pre-trains the basic data analysis model before training the basic data analysis model to obtain the target data analysis model, that is, the MDA entity trains the original model according to the configuration data of the NFV object. NFV objects are managed objects in the NFV management domain.

The configuration data of the NFV object is used to indicate the configuration of the NFV object. The original model is the initialized model.

The basic data analysis model is a general model obtained by pre-training the original model based on the configuration data of the NFV object. This basic data analysis model is independent of the subject of analysis.

Optionally, the basic data analysis model is an AI model or an ML model. This embodiment of the present application does not limit this.

The target data analysis model is a data analysis model obtained by retraining the basic data analysis model based on the information required by the designated analysis topic related to the NS. The target data analysis model is related to the analysis topic.

A target data analysis model is a model that has the ability to perform data analysis on the information needed to analyze the subject. The target data analysis model is used to convert the information required by the input analysis subject into data analysis results. The target data analysis model is used to represent the correlation between the information needed to analyze the subject and the results of the data analysis.

The target data analysis model is a preset mathematical model, and the target data analysis model includes model coefficients between the information required for analyzing the subject and the data analysis result.

Optionally, after the MDA entity has trained and completed the target data analysis model corresponding to the specified analysis topic, after acquiring the information required by the specified analysis topic to be analyzed, the information required by the specified analysis topic is input into the target data analysis. model, and output the data analysis results.

In a schematic example, as shown in Figure 4, the training and use process of the data analysis model includes but is not limited to the following steps: 1. The MDA entity obtains the configuration data of the NFV object to be trained, and configures the NFV object The data is input into the original model; 2. The basic data analysis model is obtained by pre-training the original model; 3. The training sample set is obtained, and the training sample set includes the designated analysis topic to be trained carried in the notification message received by the MDA entity To obtain the required information, input the training sample set into the basic data analysis model; 4. Retrain the basic data analysis model to obtain the target data analysis model. After training the target data analysis model, 5. When the MDA entity receives the information required by the specified analysis topic to be analyzed, input the information required by the specified analysis topic into the trained target data analysis model; 6. Output Get data analysis results.

To sum up, the embodiment of the present application completes a basic data analysis model by pre-training the MDA entity. On the basis of the basic data analysis model, the MDA entity carries the information required by the designated analysis topic related to the NS through the received The target data analysis model corresponding to the specified analysis topic is further trained, that is, the basic data analysis model is reused by the topic-specific target data analysis model during model training, avoiding the data corresponding to different analysis topics in related technologies. The analysis model training needs to derive various associations required for data analysis from scratch, which shortens the training time of the data analysis model and improves the model training efficiency.

It should be noted that, before the MDA entity trains the pre-trained basic data analysis model according to the notification information to obtain the target data analysis model corresponding to the specified analysis topic, it needs to train the original model according to the configuration data of the NFV object to obtain the basic data analysis. Model. In a possible implementation manner, the MDA entity trains the original model according to the configuration data of the NFV object to obtain a basic data analysis model. The model is trained to obtain a basic data analysis model. The information model of the NFV object in the design state includes descriptor template information of the NFV object, and the information model of the NFV object in the running state includes image information after the NFV object is instantiated. That is, based on the above embodiment, before step 301, the training method of the data analysis model further includes the following steps, as shown in FIG. 5:

Step 501: According to the descriptor template information of the imported NFV object and/or the image information after the instantiation of the NFV object, the original model is trained to obtain a basic data analysis model.

The MDA entity receives the descriptor template information of the imported NFV object and/or the image information after the instantiation of the NFV object, and trains the original model according to the descriptor template information of the NFV object and/or the image information after the instantiation of the NFV object. Basic data analysis model.

Optionally, the descriptor template information includes NSD template information and/or VNFD template information; and/or, the image information includes NS instance image information and/or VNF instance image information.

Among them, the basic data analysis model is used to indicate the attributes of the NFV object and the association relationship between the NFV objects.

Optionally, the attributes of the NFV object include attributes in NSD template information, attributes in VNFD template information, attributes in NS instance image template information, and attributes in VNF instance image information.

Illustratively, the attributes in the NSD template information include NSD identifier, provider, NSD name, NSD version, auto-scale rule used by NS, Deployment Flavor used by NS, and at least one of the security signature. A sort of.

Illustratively, the attributes in the VNFD template information include at least one of the VNFD identifier, the VNF provider, the VNF product name, the VNF software version, the VNFD version, the auto-scale rule used by the VNF, and the deployment template used by the VNF. kind.

Illustratively, the attributes in the NS instance image template information include at least one of the NS instance identifier, the NS instance name, the NS instance state, the performance indicators monitored by the NS instance, the NSD used by the NS instance, and the deployment template.

Illustratively, the attributes in the VNF instance image template information include at least one of the VNF instance identifier, the VNF instance name, the VNF instance state, the performance indicators monitored by the VNF instance, the VNFD used by the VNF instance, and the connection information with the VIM.

It should be noted that the embodiments of the present application do not limit the types of attributes of the NFV object.

Optionally, the association relationship between the NFV objects includes the association relationship between the upper and lower layer NFV objects in the vertical direction and/or the association relationship between the NFV objects connected at the same layer in the horizontal direction. For example, the virtual machine port alarm of the member object "VNF_1" in the NS instance is strongly correlated with the unreachable alarm of the member object "VL_1", that is, the basic data analysis model is used to indicate the virtual machine port of the member object "VNF_1" and the member object" VL_1" has an association relationship.

It should be noted that, for relevant definitions of descriptor template information, mirror information, original model and basic data analysis model, reference may be made to the relevant descriptions in the foregoing embodiments, which will not be repeated here.

Optionally, the training process of the basic data analysis model includes but is not limited to the following stages:

The first stage: the onboard stage of the NSD and/or VNF data package (package), the MDA entity performs basic data analysis model training based on the NSD template information and/or VNFD template information, and creates a relatively static model in the training of the basic data analysis model. Object class associations.

The second stage: the NS and/or VNF instantiation stage, the MDA entity performs basic data analysis model training based on the NS instance image information and/or VNF instance image information, and creates relatively dynamic object instance associations in the basic data analysis model training.

The third stage: In the running state after the instantiation of NS and/or VNF, the MDA entity analyzes the data in the basic data analysis model according to the NS instance image information modified during the NS update process and/or the modified VNF instance image information. The association relationship of the object instance is updated.

In the following, for the convenience of description, the above three stages are introduced and described in sequence through several exemplary embodiments.

The first stage is the stage of uploading NSD and/or VNF data packets. Taking the descriptor template information including NSD template information as an example, the training process of the basic data analysis model includes but is not limited to the following steps, as shown in Figure 6:

In step 601, the operations support system (OSS)/business support system (BSS) initiates an NSD onboard process to the NFVO, and the specified NSD is listed in the NFV management domain.

Step 602, the NFVO imports the NSD template information on the shelves into the MDA entity.

That is, NFVO sends the NSD template information to the MDA entity as the input data for training the basic data analysis model.

Optionally, NFVO imports the listed NSD template information into the MDA entity, including but not limited to the following two possible implementations:

In a possible implementation manner, the NSD import process adopts an on-path process, that is, along with the NSD listing process in step 601, the OSS/BSS directly imports the NSD template information into the MDA entity through NFVO.

In another possible implementation manner, the NFVO sends a model training input request to the MDA entity, where the model training input request carries the information of the NSD template to be imported.

Step 603, the MDA entity performs basic data analysis model training according to the imported NSD template information, and establishes an association relationship between the NSD template information and the descriptor template information of the member objects of the NS.

MDA trains the basic data analysis model according to the imported NSD template information. During the training of the basic data analysis model, the MDA entity establishes the association relationship between the NSD template information and the descriptor template information of the member objects of the NS. That is, the trained basic data analysis model is used to indicate the association relationship between the NSD template information and the descriptor template information of the member objects of the NS.

Optionally, the MDA entity establishes an association relationship between the NSD template information and the descriptor template information of the member object of the NS, and the attribute of the member object of the NS corresponds to the attribute of the NSD template information.

Optionally, in the training process of the basic data analysis model, the MDA entity establishes an inclusion relationship between the NS object class and the object class corresponding to the descriptor template information of the member objects of the NS. That is, the trained basic data analysis model is used to indicate the inclusion relationship between the NS object class and the object class corresponding to the descriptor template information of the member objects of the NS.

The attribute of the object class of the member object corresponds to the attribute of the descriptor template information of the member object.

Illustratively, the object class of the member object includes: the object class of the member nesting NS of the NS, the object class of the member VNF, the object class of the member PNF, the object class of the Service Access Point (SAP), the NS At least one of the object class of the external virtual link and the object class of the used VNF forwarding graph. This embodiment of the present application does not limit this.

In a possible implementation manner, the descriptor template information includes VNFD template information, and the above steps 601 to 603 can be replaced and implemented as the following steps: OSS/BSS initiates a VNF data package listing process to NFVO, and puts them on the shelf in the NFV management domain The specified VNF packet. NFVO imports the listed VNFD template information into the MDA entity. The MDA entity performs basic data analysis model training based on the imported VNFD template information.

In another possible implementation, the descriptor template information includes NSD template information and VNFD template information, and the above steps 601 to 603 can be replaced and implemented as the following steps: OSS/BSS initiates the NSD listing process to NFVO, and the NFV manages The specified NSD is listed in the domain. NFVO imports the listed NSD template information into the MDA entity. The OSS/BSS initiates the VNF data package listing process to the NFVO, and the specified VNF data package is listed in the NFV management domain. NFVO imports the listed VNFD template information into the MDA entity. The MDA entity performs basic data analysis model training according to the imported NSD template information and VNFD template information, and establishes the association relationship between the NSD template information and the descriptor template information of the member objects of the NS. It should be noted that the descriptor template information includes VNFD template information, or, when including NSD template information and VNFD template information, the training process of the corresponding basic data analysis model can be referred to by analogy with the relevant descriptions in the above embodiments, which will not be repeated here. .

The second stage is the NS and/or VNF instantiation stage. Taking the image information including NS instance image information and VNF instance image information as an example, the training process of the basic data analysis model includes but is not limited to the following steps, as shown in Figure 7 :

Step 701, the OSS/BSS initiates a process of creating an NS instance identifier to the NFVO, and creates an identifier of the NS instance and the NS instance image information corresponding to the NS instance.

Step 702, NFVO imports the created NS instance image information into the MDA entity.

That is, NFVO sends the NS instance image information to the MDA entity as the input data for training the basic data analysis model.

Optionally, NFVO imports the created NS instance image information into the MDA entity, including but not limited to the following two possible implementations:

In a possible implementation manner, the process of importing the NS instance image information adopts an on-path process, that is, along with the NS instance identifier creation process in step 701, the OSS/BSS directly imports the NS instance image information into the MDA entity through NFVO.

In another possible implementation manner, the NFVO sends a model training input request to the MDA entity, where the model training input request carries the image information of the NS instance to be imported.

Step 703, the NFVO initiates a process of creating a VNF instance identifier to the VNFM, and creates an identifier of the VNF instance and the VNF instance image information corresponding to the VNF instance.

Step 704, the VNFM imports the created VNF instance image information into the MDA entity.

That is, the VNFM sends the VNF instance image information to the MDA entity as the input data for training the basic data analysis model.

Optionally, the VNFM imports the created VNF instance image information into the MDA entity, including but not limited to the following two possible implementations:

In a possible implementation manner, the process of importing VNF instance image information adopts an on-path process, that is, along with the VNF instance identifier creation process in step 703, the NFVO directly imports the VNF instance image information into the MDA entity through the VNFM.

In another possible implementation manner, the VNFM sends a model training input request to the MDA entity, where the model training input request carries the image information of the VNF instance to be imported.

It should be noted that the import process of the NS instance image information introduced in steps 701 to 702 and the import process of the VNF instance image information introduced in steps 703 to 704 can be executed in parallel, or the NS instance image information can be executed first The import process of the VNF instance image information is performed after the import process of the VNF instance image information. The import process of the VNF instance image information information can also be performed first and then the import process of the NS instance image information information process is performed. This embodiment does not limit the execution order.

Step 705, the MDA entity performs basic data analysis model training according to the imported NS instance mirror information and VNF instance mirror information, and establishes an association relationship between the NS instance mirror information and the mirror information of member object instances of the NS instance.

The MDA entity performs basic data analysis model training according to the imported NS instance image information and VNF instance image information. During the training process of the basic data analysis model, the MDA entity establishes the relationship between the NS instance image information and the image information of the member object instances of the NS instance. association relationship. That is, the trained basic data analysis model is used to indicate the association relationship between the mirror information of the NS instance and the mirror information of the member object instances of the NS instance.

Optionally, the MDA entity establishes an association relationship between the mirror information of the NS instance and the mirror information of the member object instances of the NS instance, and the attributes of the member object instances of the NS instance correspond to the attributes in the mirror template information of the NS instance.

Optionally, in the training process of the basic data analysis model, the MDA entity establishes an inclusion relationship between the NS instance and the object instance corresponding to the descriptor template information of the member objects of the NS. That is, the trained basic data analysis model is used to indicate the inclusion relationship between the NS instance and the object instance corresponding to the descriptor template information of the member object of the NS.

The attribute of the object instance of the member object corresponds to the attribute of the mirror information of the member object instance.

Illustratively, the object instance of the member object includes: at least one of the member nested NS instance, the member VNF instance, the member PNF instance, the SAP instance, the virtual link instance outside the NS, and the used VNF forwarding graph instance that constitute the NS instance. kind. This embodiment of the present application does not limit this.

In a possible implementation manner, the image information includes the image information of the NS instance, and the above steps 701 to 705 can be replaced by the following steps: OSS/BSS initiates a process of creating an NS instance identifier to NFVO, and the identifier of the NS instance and the identifier of the NS instance are created. NS instance image information corresponding to the NS instance. NFVO imports the created NS instance image information into the MDA entity. The MDA entity conducts basic data analysis model training according to the imported NS instance mirror information, and establishes an association relationship between the NS instance mirror information and the mirror information of the member object instances of the NS instance.

In another possible implementation manner, the image information includes VNF instance image information, and the above steps 701 to 705 can be replaced and implemented as the following steps: NFVO initiates a VNF instance identifier creation process to the VNFM, and creates the VNF instance identifier and the VNF instance identifier. Information about the VNF instance image corresponding to the instance. The VNFM imports the created VNF instance image information into the MDA entity. The MDA entity performs basic data analysis model training based on the imported VNF instance image information.

It should be noted that the training process of the corresponding basic data analysis model when the mirroring information includes the mirroring information of the NS instance or the mirroring information of the VNF instance can be referred to the relevant description in the above embodiment by analogy, and details are not repeated here.

The third stage is the running state after the NS and/or VNF are instantiated. Taking the image information of the modified instance as the VNF instance image information during the update process of the NFV object instance as an example, the training process of the basic data analysis model includes but is not limited to the following Several steps, as shown in Figure 8:

Step 801, the OSS/BSS initiates an NS update process to the NFVO.

Optionally, the OSS/BSS sends a request message to the NFVO, where the request message carries the NS update type, and the NS update type is used to instruct to modify the image information of the VNF instance that constitutes the NS instance.

Step 802, the NFVO initiates a VNF instance information modification process to the VNFM, and modifies the specified VNF instance image information.

After receiving the request message carrying the NS update type, the NFVO initiates the VNF instance information modification process to the VNFM, and modifies the VNF instance image information indicated by the NS update type to obtain the modified VNF instance image information.

Step 803, the VNFM imports the modified VNF instance image information into the MDA entity.

That is, the VNFM sends the modified VNF instance image information to the MDA entity as the input data of the basic data analysis model training to instruct the basic data analysis model to be updated.

Optionally, the VNFM imports the modified VNF instance image information into the MDA entity, including but not limited to the following two possible implementations:

In a possible implementation manner, the process of importing the modified VNF instance image information adopts an on-path process, that is, along with the VNF instance information modification process in step 802, the NFVO directly imports the modified VNF instance image information into the MDA entity through the VNFM .

In another possible implementation manner, the VNFM sends a model training input request to the MDA entity, where the model training input request carries the modified VNF instance image information.

Step 804, the MDA entity updates the basic data analysis model according to the modified VNF instance image information.

The MDA entity updates the basic data analysis model according to the modified VNF instance image information to obtain the updated basic data analysis model. For related details, refer to the related description of the MDA entity training the basic data analysis model according to the VNF instance image information, and will not be repeated here.

In a possible implementation manner, the above steps 801 to 804 can also be replaced and implemented as the following steps: the OSS/BSS initiates an NS update process to the NFVO to modify the NS instance image information. NFVO imports the modified NS instance image information into the MDA entity. The MDA entity updates the basic data analysis model according to the modified NS instance image information. In another possible implementation manner, the above steps 801 to 804 can also be replaced and implemented as the following steps: the OSS/BSS initiates an NS update process to the NFVO. The NFVO initiates the VNF instance information modification process to the VNFM to modify the specified VNF instance image information. The VNFM imports the modified VNF instance image information into the MDA entity. OSS/BSS initiates the NS update process to NFVO to modify the NS instance image information. NFVO imports the modified NS instance image information into the MDA entity. The MDA entity updates the basic data analysis model according to the modified NS instance image information and the modified VNF instance image information.

It should be noted that the MDA entity updates the basic data analysis model according to the modified NS instance image information, or the MDA entity updates the basic data analysis model according to the modified NS instance image information and the modified VNF instance image information. For the update process, reference can be made to the relevant descriptions in the foregoing embodiments by analogy, and details are not repeated here.

Based on the above trained basic data analysis model, please refer to FIG. 9 , which shows a flowchart of a training method for a data analysis model provided by another exemplary embodiment of the present application, and the method is used for the MDA entity shown in FIG. 2 . middle. The method includes the following steps.

Step 901, the MDA entity subscribes to the NFVO for the information required by the specified analysis topic related to the NS.

Optionally, the information required for the specified analysis topic related to the NS includes information related to the specified NS instance. For example, the information required when specifying an analysis topic as an NS alarm event analysis topic includes NS-related performance data and/or alarm data.

Step 902, the NFVO sends a notification message to the MDA entity that carries the information required by the specified analysis topic related to the NS.

The NFVO sends a notification message to the MDA entity, which carries the information required for the specified analysis topic related to the NS.

In a possible implementation manner, the above steps 901 and 902 are replaced by the following steps: the MDA entity subscribes to the VNFM for the information required by the specified analysis topic related to the NS. The VNFM sends a notification message to the MDA entity that carries the information required for the specified analysis topic related to the NS.

Optionally, in this implementation manner, the information required for the specified analysis topic related to the NS includes information related to all member VNF instances of the specified NS instance.

In another possible implementation manner, the above-mentioned steps 901 and 902 are replaced by the following steps: the MDA entity subscribes to the VIM for information required by the specified analysis topic related to the NS. The VIM sends a notification message to the MDA entity that carries the information needed for the specified analysis topic related to the NS.

Optionally, in this implementation manner, the information required for the specified analysis topic related to the NS includes information related to virtualized resources used by all member VNF instances and virtual link instances of the specified NS instance.

In another possible implementation manner, the above steps 901 and 902 are replaced by the following steps: the MDA entity subscribes to the CISM entity for information required by the specified analysis topic related to the NS. The CISM entity sends a notification message to the MDA entity that carries the information required for the specified analysis topic related to the NS.

Optionally, in this implementation manner, the information required for the specified analysis topic related to the NS includes information related to the managed container infrastructure object (Managed Container Infrastructure Object, MCIO) used by all the member VNF instances of the specified NS instance. information.

Step 903, the MDA entity receives a notification message that carries the information required for the specified analysis topic related to the NS.

The MDA entity receives a notification message that carries the information required for the specified analysis topic related to the NS.

Step 904, the MDA entity trains the pre-trained basic data analysis model according to the notification information to obtain the target data analysis model corresponding to the specified analysis topic.

The basic data analysis model is the above-trained basic data analysis model.

Optionally, the MDA entity establishes an association relationship between NFV objects and/or an association relationship between management events according to the notification message, and expands the basic data analysis model based on the established association relationship to form a target data analysis model. For example, if the analysis topic is specified as the NS alarm event analysis topic, the MDA entity receives the notification message carrying the NS-related alarm data, and analyzes the received multiple alarm data to determine the virtual machine of the member object "VNF_1" in the NS instance. The port alarm is strongly correlated with the unreachable alarm of the member object "VL_2", thus establishing the correlation between the virtual machine port alarm of the member object "VNF_1" in the NS instance and the member object "VL_2". does not have this relationship.

In a possible implementation manner, the MDA entity trains the pre-trained basic data analysis model according to the notification information to obtain the target data analysis model corresponding to the specified analysis topic, and then invokes the trained target data analysis model for data analysis. Based on the above embodiment, after step 302 or step 904, the use process of the target data analysis model includes but is not limited to the following steps, as shown in FIG. 10 :

Step 1001: Obtain information required by a specified analysis topic to be analyzed.

The MDA entity obtains the information required for the specified analysis subject to be analyzed. Optionally, the specified analysis topic includes any one of an NS alarm event analysis topic, an NS health degree analysis topic, and an NS resource utilization analysis topic. For example, the specified analysis topic is an NS alarm event analysis topic, and the information required by the specified analysis topic includes NS-related performance data and/or alarm data.

Step 1002: Input the information required by the designated analysis topic into the target data analysis model, and output the data analysis result.

The MDA entity inputs the information required by the specified analysis subject into the target data analysis model, and outputs the data analysis result.

In a possible implementation manner, the MDA entity inputs the NS-related performance data and/or alarm data into the target data analysis model corresponding to the NS alarm event analysis topic, and outputs a first analysis result, where the first analysis result includes The underlying alarm and/or root cause of the NS failure.

Optionally, the performance data and/or alarm data related to the NS includes performance data and/or alarm data related to the NS and its constituent components/infrastructure resources. NS-related performance data and/or alarm data are provided in NFVO. NFVO sends NS related performance data and/or alarm data to the MDA entity for data analysis. The MDA entity invokes the target data analysis model corresponding to the NS alarm event analysis topic, performs data analysis on the NS-related performance data and/or alarm data, and outputs the first analysis result, where the first analysis result includes the fundamental alarm and/or the NS failure. or root cause.

Optionally, the MDA entity returns the first analysis result to NFVO. The NFVO confirms the first analysis result, and groups the NS-related performance data and/or alarm data corresponding to the first analysis result.

In another possible implementation manner, the MDA entity inputs the information required for NS health degree analysis into a target data analysis model corresponding to the subject of NS health degree analysis, and outputs a second analysis result, where the second analysis result includes NS health status and the description information corresponding to the health status.

Optionally, the information required for NS health degree analysis includes NS status, NS fault management (Fault Management, FM)/performance management (Performance Management, PM) information, VNF indicators, and configuration information related to the analyzed NS. At least one, which is not limited in this embodiment of the present application.

NFVO decided to request the MDA entity to analyze the health of the NS in its management domain. The NFVO sends a data analysis request to the MDA entity, where the data analysis request is used to instruct the MDA entity to perform health analysis on the specified NS. The MDA entity interacts with other NFV-MANO functional entities such as NFVO, VNFM and/or VIM to gather information required for NS health analysis. The MDA entity invokes the target data analysis model corresponding to the NS health degree analysis topic, and performs data analysis and output on the information required for the NS health degree analysis to obtain a second analysis result. The second analysis result includes the health state of the NS and the corresponding health state. Description information. The health status of the NS is used to indicate whether the NS is healthy or unhealthy.

Optionally, the MDA entity returns the second analysis result to NFVO. The NFVO receives the second analysis result, and obtains the operating status of the NS.

In another possible implementation manner, the MDA entity inputs the information required for the NS resource utilization analysis into the target data analysis model corresponding to the subject of the NS resource utilization analysis, and outputs a third analysis result, which is the third analysis result. Include an indication of resource utilization of the NS.

Optionally, the information required for the NS resource utilization analysis includes virtual computing-related measurement information of each constituent VNF of the NS, network data volume-related measurement information of the SAP belonging to the NS, and network data of the external CPs belonging to the NS that form the VNF. At least one of quantity-related measurement information, an indicator of the NS constituting the VNF. This embodiment of the present application does not limit this.

NFVO decided to perform data analysis on the resource utilization of NS. The NFVO sends a data analysis request to the MDA entity, where the data analysis request is used to instruct the MDA entity to perform resource utilization analysis on the specified NS. The MDA interacts with the corresponding NFV-MANO functional entities to collect the information required for NS resource utilization analysis over a period of time. The MDA entity invokes the target data analysis model corresponding to the NS resource utilization analysis theme, and performs data analysis and output on the information required for the NS resource utilization analysis to obtain a third analysis result. The third analysis result includes an indication of the resource utilization of the NS .

Optionally, the third analysis result includes resource utilization of at least one type of NS resources and corresponding suggestion information, so as to solve the resource utilization problem determined in the analysis report. For example, the types of NS resources include at least one of computing resources, storage resources, and network resources.

Optionally, the MDA entity returns the analysis results to NFVO. NFVO obtains the resource utilization of NS over a period of time.

It should be noted that, the embodiment of the present application does not limit the type of the specified analysis topic and the information content required for the specified analysis topic.

To sum up, in the training method of the data analysis model provided by the embodiment of the present application, the basic data analysis model dynamically maintains the attributes of the NFV object and the association relationship between the NFV objects based on the configuration data of the NFV object. On the basis of the basic data analysis model, further training is performed to complete the analysis theme-specific data analysis model by collecting the analysis theme-specific NS-related performance data and/or alarm data.

In this embodiment of the present application, the information model of the NFV object in the design state (that is, the descriptor template information of the NFV object) and the information model of the NFV object in the running state (that is, the mirror information after the instantiation of the NFV object) are used as basic data through the MDA entity. The input data of model training is analyzed, and the basic data analysis model irrelevant to the analysis topic is obtained by training, so as to enhance the ability of the MDA entity to dynamically obtain the association relationship between NFV objects in the NFV management domain, and further improve the subsequent training target data analysis model. efficiency.

Please refer to FIG. 11 , which shows a block diagram of an apparatus for training a data analysis model provided by an exemplary embodiment of the present application. The training device of the data analysis model can be implemented by software, hardware or a combination of the two to become all or a part of the MDA entity shown in FIG. 2 . The training device for the data analysis model may include: a receiving unit 1110 and a processing unit 1120 .

The receiving unit 1110 is used to realize the functions of the

above steps

301, 903 and 1001 and the receiving function of the MDA entity side implicit in each step;

The processing unit 1120 is configured to implement the functions of the

above steps

302, 501, 603, 705, 804, 904 and 1002 and the processing functions on the MDA entity side implicit in each step.

Relevant details can be combined with the method embodiments described with reference to FIG. 3 to FIG. 10 , and details are not repeated here.

An embodiment of the present application provides a training device for a data analysis model, which is used in an MDA entity. The device includes: a processor and a memory for storing instructions executable by the processor; wherein, when the processor is configured to execute the instructions Implement the methods described above to be performed by the MDA entity.

Embodiments of the present application provide a computer program product, including computer-readable codes, or a non-volatile computer-readable storage medium carrying computer-readable codes, when the computer-readable codes are executed in a processor of an electronic device , the processor in the electronic device executes the above-mentioned method executed by the MDA entity.

Embodiments of the present application provide a non-volatile computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the above-mentioned method executed by an MDA entity is implemented.

A computer-readable storage medium may be a tangible device that can hold and store instructions for use by the instruction execution device. The computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read-only memory (Electrically Programmable Read-Only-Memory, EPROM or flash memory), static random access memory (Static Random-Access Memory, SRAM), portable compact disk read-only memory (Compact Disc Read-Only Memory, CD - ROM), Digital Video Disc (DVD), memory sticks, floppy disks, mechanically encoded devices, such as punch cards or raised structures in grooves on which instructions are stored, and any suitable combination of the foregoing .

Computer readable program instructions or code described herein may be downloaded to various computing/processing devices from a computer readable storage medium, or to an external computer or external storage device over a network such as the Internet, a local area network, a wide area network and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer-readable program instructions from a network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing/processing device .

The computer program instructions used to perform the operations of the present application may be assembly instructions, Instruction Set Architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or in one or more source or object code written in any combination of programming languages, including object-oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server implement. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or, may be connected to an external computer (eg, use an internet service provider to connect via the internet). In some embodiments, electronic circuits, such as programmable logic circuits, Field-Programmable Gate Arrays (FPGA), or Programmable Logic Arrays (Programmable Logic Arrays), are personalized by utilizing state information of computer-readable program instructions. Logic Array, PLA), the electronic circuit can execute computer readable program instructions to implement various aspects of the present application.

Aspects of the present application are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the present application. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.

These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer or other programmable data processing apparatus to produce a machine that causes the instructions when executed by the processor of the computer or other programmable data processing apparatus , resulting in means for implementing the functions/acts specified in one or more blocks of the flowchart and/or block diagrams. These computer readable program instructions can also be stored in a computer readable storage medium, these instructions cause a computer, programmable data processing apparatus and/or other equipment to operate in a specific manner, so that the computer readable medium on which the instructions are stored includes An article of manufacture comprising instructions for implementing various aspects of the functions/acts specified in one or more blocks of the flowchart and/or block diagrams.

Computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other equipment to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other equipment to produce a computer-implemented process , thereby causing instructions executing on a computer, other programmable data processing apparatus, or other device to implement the functions/acts specified in one or more blocks of the flowcharts and/or block diagrams.

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of apparatuses, systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more functions for implementing the specified logical function(s) executable instructions. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.

It is also noted that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented in hardware (eg, circuits or ASICs (Application) that perform the corresponding functions or actions. Specific Integrated Circuit, application-specific integrated circuit)), or can be implemented by a combination of hardware and software, such as firmware.

Although the application is described herein in conjunction with the various embodiments, those skilled in the art will understand and understand from a review of the drawings, the disclosure, and the appended claims in practicing the claimed application. Other variations of the disclosed embodiments are implemented. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that these measures cannot be combined to advantage.

Various embodiments of the present application have been described above, and the foregoing descriptions are exemplary, not exhaustive, and not limiting of the disclosed embodiments. Numerous modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the various embodiments, the practical application or improvement over the technology in the marketplace, or to enable others of ordinary skill in the art to understand the various embodiments disclosed herein.

Claims

A training method for a data analysis model, characterized in that it is used to manage a data analysis MDA entity, the method comprising:

receiving a notification message carrying information required for a specified analysis topic related to the network service NS;

According to the notification information, the pre-trained basic data analysis model is trained to obtain the target data analysis model corresponding to the specified analysis topic, and the basic data analysis model is based on the configuration data of the network function virtualization NFV object. Obtained through training, the NFV object is a managed object in the NFV management domain.
The method according to claim 1, wherein the basic data analysis model is used to indicate an attribute of the NFV object and an association relationship between the NFV object.
The method according to claim 1, wherein the method further comprises:

According to the imported descriptor template information of the NFV object and/or the image information after the instantiation of the NFV object, the original model is trained to obtain the basic data analysis model.
The method of claim 3, wherein:

The descriptor template information includes network service descriptor NSD template information and/or virtualized network function descriptor VNFD template information; and/or,

The image information includes NS instance image information and/or virtualized network function VNF instance image information.
The method according to claim 4, wherein the method further comprises:

During the training process of the basic data analysis model, an association relationship between the NSD template information and the descriptor template information of the member objects of the NS is established.
The method according to claim 4, wherein the method further comprises:

During the training process of the basic data analysis model, an association relationship is established between the mirror information of the NS instance and the mirror information of the member object instances of the NS instance.
The method according to claim 4, wherein the method further comprises:

The basic data analysis model is updated according to the modified image information of the NS instance and/or the modified image information of the VNF instance.
The method according to any one of claims 1 to 7, wherein the method further comprises:

Input NS-related performance data and/or alarm data into the target data analysis model corresponding to the NS alarm event analysis theme, and output to obtain a first analysis result, where the first analysis result includes the fundamental alarm of the NS failure and/or root cause; or,

Input the information required for NS health degree analysis into the target data analysis model corresponding to the NS health degree analysis subject, and output a second analysis result, where the second analysis result includes the health state of the NS and the Description information corresponding to the health status; or,

Input the information required for the NS resource utilization analysis into the target data analysis model corresponding to the NS resource utilization analysis theme, and output a third analysis result, where the third analysis result includes the resource utilization of the NS indication of the situation.
A training device for a data analysis model, characterized in that it is used to manage a data analysis MDA entity, the device comprising:

a receiving unit, configured to receive a notification message, the notification message carrying information required for a specified analysis topic related to the network service NS;

The processing unit is configured to train the pre-trained basic data analysis model according to the notification information to obtain the target data analysis model corresponding to the specified analysis topic, and the basic data analysis model is a virtualized NFV object according to the network function. The configuration data is obtained by training the original model, and the NFV object is a managed object in the NFV management domain.
The apparatus according to claim 9, wherein the basic data analysis model is used to indicate an attribute of the NFV object and an association relationship between the NFV object.
The apparatus according to claim 9, wherein the processing unit is further configured to:

According to the imported descriptor template information of the NFV object and/or the image information after the instantiation of the NFV object, the basic data analysis model is obtained by training the original model.
The device of claim 11, wherein:

The descriptor template information includes network service descriptor NSD template information and/or virtualized network function descriptor VNFD template information; and/or,

The image information includes NS instance image information and/or virtualized network function VNF instance image information.
The apparatus according to claim 12, wherein the processing unit is further configured to:

During the training process of the basic data analysis model, an association relationship between the NSD template information and the descriptor template information of the member objects of the NS is established.
The apparatus according to claim 12, wherein the processing unit is further configured to:

During the training process of the basic data analysis model, an association relationship is established between the mirror information of the NS instance and the mirror information of the member object instances of the NS instance.
The apparatus according to claim 12, wherein the processing unit is further configured to:

The basic data analysis model is updated according to the modified NS instance image information and/or the modified VNF instance image information.
The device according to any one of claims 9 to 15, wherein the processing unit is further configured to:

Input NS-related performance data and/or alarm data into the target data analysis model corresponding to the NS alarm event analysis theme, and output to obtain a first analysis result, where the first analysis result includes the fundamental alarm of the NS failure and/or root cause; or,

Input the information required for the NS health degree analysis into the target data analysis model corresponding to the NS health degree analysis subject, and output a second analysis result, where the second analysis result includes the health state of the NS and the Description information corresponding to the health status; or,

Input the information required for NS resource utilization analysis into the target data analysis model corresponding to the NS resource utilization analysis theme, and output to obtain a third analysis result, where the third analysis result includes the resource utilization of the NS indication of the situation.
A device for data analysis model training, characterized in that it is used to manage a data analysis MDA entity, the device comprising:

processor;

memory for storing processor-executable instructions;

Wherein, the processor is configured to implement the method of any one of claims 1-8 when executing the instructions.
A non-volatile computer-readable storage medium on which computer program instructions are stored, characterized in that, when the computer program instructions are executed by a processor, the method described in any one of claims 1 to 8 is implemented.