Network topology graph generation method, apparatus, electronic device and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for generating network topology maps. The method includes: acquiring XDR data, core network configuration data, and B-domain data corresponding to a target enterprise; performing data filtering processing on the XDR data, core network configuration data, and B-domain data to generate a network data field table; constructing a network device service path table corresponding to the network devices within the target enterprise based on the network data field table; and drawing a network topology map corresponding to the target enterprise based on the core network node configuration conversion table and the network device service path table. This application can realize real-time updated full-path visualization applications, fuzzy queries by slice, enterprise name, or DNN, presentation of key indicators and alarms, and achieve functions such as visibility, manageability, and evaluation of private network services, adapting to the needs of refined management of private network resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to a network topology graph generation method and device, electronic equipment and storage medium. BACKGROUND

[0002] As an important product for operators ToB business expansion and transformation, 5G private network uses network slicing, edge computing MEC (Mobile Electronic Commerce) and other technologies to provide customized network solutions for various vertical industries. Whether it is an operator or an industry customer, the demand for operation and maintenance of 5G private network and its industry application is becoming more and more urgent. How to solve the problem that the current core network topology does not have the function of querying and generating the network topology corresponding to a specific slice business or enterprise business. How to change the status quo of manually maintaining relevant information by static statistical node parameters. How to realize real-time extraction of node configuration, real-time conversion of logic, and real-time drawing of topology. How to realize the generation of enterprise-level, slice-level, and business-level resources and link topology across manufacturers to provide atomic capabilities for business management and fault definition, thereby reducing user complaints and improving user satisfaction. This has always been a difficult problem to solve.

[0003] The existing network topology generation method and device, the conventional means classification has: 1, protocol discovery type, based on a topology discovery starting IP, set the discovery range, discovery device type limit and other information, automatically discover the target network and generate 2 layer and 3 layer network topology; 2, network management input type, in the comprehensive network management system, through the network device new integrated input, device type, device relationship logic preset input, record the device connection topology of this professional and this manufacturer.

[0004] The main problems existing in the prior art are: 1, the protocol discovery type means is simple and easy to operate, but only suitable for automatic discovery of general TCP / IP protocol data network devices; 2, the network management input type means is professional and accurate, which belongs to single professional topology and does not have certain business topology continuity and does not support cross-manufacturer. SUMMARY

[0005] The embodiments of the present application provide a network topology graph generation method and device, electronic equipment and storage medium, to realize real-time updating of full-path visualization application, slice / enterprise name / DNN fuzzy query, key indicator and alarm presentation, realize private network business visualization management and evaluation, and adapt to the demand of private network resource fine management.

[0006] In a first aspect, the embodiments of the present application provide a network topology graph generation method, the method comprising:

[0007] obtaining XDR data, core network configuration data and B domain data corresponding to a target enterprise;

[0008] performing data filtering processing on the XDR data, the core network configuration data and the B domain data to generate a network data field table;

[0009] based on the network data field table, constructing a network device service path table corresponding to network devices in the target enterprise;

[0010] based on a core network node configuration conversion table and the network device service path table, drawing a network topology graph corresponding to the target enterprise.

[0011] Optionally, the obtaining of the XDR data, the core network configuration data and the B domain data corresponding to the target enterprise comprises:

[0012] based on a set period, collecting the XDR data, the core network configuration data and the B domain data corresponding to the target enterprise.

[0013] Optionally, the performing of the data filtering processing on the XDR data, the core network configuration data and the B domain data to generate a network data field table comprises:

[0014] based on the correspondence between base station ID, base station IP, TAC, DNN, slice and AMF IP, performing filtering processing on the XDR data to obtain filtered target XDR data;

[0015] performing filtering operations on AMF data, SMF data and UPF data in the core network configuration data to obtain filtered target core network configuration data;

[0016] based on the target XDR data, the target core network configuration data and the B domain data, generating the network data field table.

[0017] Optionally, the constructing of the network device service path table corresponding to network devices in the target enterprise based on the network data field table comprises:

[0018] based on 5G private network service logic, filtering key factors from the network data field table;

[0019] based on the information gain of features, performing verification processing on the key factors to obtain multi-domain features;

[0020] According to a maximum feature class method of a Naive Bayes model, the multi-domain features are subjected to probability analysis processing to obtain a probability of association between the multi-domain features and the enterprise topology class;

[0021] According to the probability, target multi-domain features are screened out from the multi-domain features to construct the network device service path table.

[0022] Optionally, the network topology graph corresponding to the target enterprise is drawn based on the node configuration conversion table of the core network and the network device service path table, and includes:

[0023] Based on the node configuration conversion table of the core network and the network device service path table, service logic for drawing a whole network topology graph of the core network is obtained;

[0024] The original data in the network data field table is converted into a data model through a data algorithm;

[0025] Based on the service logic, the data model is drawn as an initial network topology graph;

[0026] The initial network topology graph is post-processed to generate the network topology graph corresponding to the target enterprise.

[0027] In a second aspect, an embodiment of the present application provides a network topology graph generation device, and the device includes:

[0028] A data acquisition module is configured to acquire XDR data, core network configuration data, and B-domain data corresponding to a target enterprise;

[0029] A field table generation module is configured to perform data filtering processing on the XDR data, the core network configuration data, and the B-domain data to generate a network data field table;

[0030] A path table construction module is configured to construct a network device service path table corresponding to a network device in the target enterprise based on the network data field table;

[0031] A topology graph drawing module is configured to draw the network topology graph corresponding to the target enterprise based on a node configuration conversion table of a core network and the network device service path table.

[0032] Optionally, the data acquisition module includes:

[0033] A data collection unit is configured to collect the XDR data, the core network configuration data, and the B-domain data corresponding to the target enterprise based on a set period.

[0034] Optionally, the field table generation module includes:

[0035] An XDR data screening unit is configured to perform screening processing on the XDR data based on a correspondence relationship between a base station ID, a base station IP, a TAC, a DNN, a slice, and an AMF IP, to obtain screened target XDR data.

[0036] A configuration data screening unit is configured to perform screening operation on AMF data, SMF data, and UPF data in the core network configuration data, to obtain screened target core network configuration data.

[0037] A field table generating unit is configured to generate the network data field table based on the target XDR data, the target core network configuration data, and the B domain data.

[0038] Optionally, the path table constructing module comprises:

[0039] A key factor screening unit is configured to screen a key factor from the network data field table based on 5G private network service logic.

[0040] A multi-domain feature acquiring unit is configured to perform verification processing on the key factor based on information gain of features, to obtain multi-domain features.

[0041] A probability acquiring unit is configured to perform probability analysis processing on the multi-domain features according to a maximum feature class method of a naive Bayes model, to obtain a probability of association between the multi-domain features and an enterprise topology category.

[0042] A path table constructing unit is configured to screen target multi-domain features from the multi-domain features according to the probability, to construct a network device service path table.

[0043] Optionally, the topology graph drawing module comprises:

[0044] A service logic acquiring unit is configured to obtain service logic of drawing a whole network topology graph of the core network based on a node configuration conversion table of the core network and the network device service path table.

[0045] A data model converting unit is configured to convert original data in the network data field table into a data model through a data algorithm.

[0046] An initial graph drawing unit is configured to draw the data model into an initial network topology graph based on the service logic.

[0047] A topology graph generating unit is configured to perform post-processing on the initial network topology graph, to generate a network topology graph corresponding to the target enterprise.

[0048] In a third aspect, an electronic device is provided, comprising:

[0049] A processor, a memory, and a computer program stored on the memory and executable on the processor, the processor implementing the network topology graph generation method of any of the preceding items when executing the program.

[0050] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the network topology graph generation method of any of the preceding items.

[0051] Compared with the prior art, the embodiments of the present application have the following advantages:

[0052] In the embodiments of the present application, the XDR data, the core network configuration data and the B domain data corresponding to the target enterprise are obtained. The XDR data, the core network configuration data and the B domain data are subjected to data filtering processing to generate a network data field table. Based on the network data field table, a network device service path table corresponding to the network devices in the target enterprise is constructed. Based on the node configuration conversion table of the core network and the network device service path table, a network topology graph corresponding to the target enterprise is drawn. The embodiments of the present application can automatically output cross-manufacturer network resource topology at the enterprise level and the slice level based on industry customers, and help to realize topology visualization, so as to realize real-time updating of full-path visualization application, slice / enterprise name / DNN fuzzy query, key indicator and alarm presentation, and realize functions such as special network business visualization, management and evaluation, and adapt to the fine management demand of special network resources.

[0053] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 A step flowchart of a network topology graph generation method provided by the embodiments of the present application;

[0055] Figure 2 An input and output schematic diagram of a device for enterprise-level and slice-level 5G special network topology generation based on configuration data provided by the embodiments of the present application.

[0056] Figure 3 A process schematic diagram of enterprise-level and slice-level 5G special network topology generation based on configuration data provided by the embodiments of the present application.

[0057] Figure 4 An automatic data conversion module schematic diagram of enterprise-level and slice-level 5G special network topology generation based on configuration data provided by the embodiments of the present application;

[0058] Figure 5 An automatic generation path information module schematic diagram provided by the embodiments of the present application;

[0059] Figure 6 A 5G private network opening information extraction field diagram provided for an embodiment of the present application.

[0060] Figure 7 An AMF configuration extraction field diagram provided for an embodiment of the present application;

[0061] Figure 8 Another AMF configuration extraction field diagram provided for an embodiment of the present application;

[0062] Figure 9 Still another AMF configuration extraction field diagram provided for an embodiment of the present application;

[0063] Figure 10 An information gain measurement method provided for an embodiment of the present application verifies 11 features as a feature subset diagram;

[0064] Figure 11 A core network DNN / slice topology diagram automatically drawn for an embodiment of the present application;

[0065] Figure 12 A business association DNN diagram provided for an embodiment of the present application;

[0066] Figure 13 A core network configuration information field extraction diagram provided for an embodiment of the present application;

[0067] Figure 14 A network topology diagram generation device structure diagram provided for an embodiment of the present application;

[0068] Figure 15 An electronic device structure diagram provided for an embodiment of the present application. DETAILED DESCRIPTION

[0069] In order to make the above objectives, features and advantages of the present application more apparent and easy to understand, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0070] The terms used in the embodiments of the present application are merely for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "an" and "the" used in the embodiments of the present application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0071] This application addresses the following problems in existing technologies: 1. Protocol discovery methods are simple and easy to implement, but only applicable to the automatic discovery of data communication network devices using the general TCP / IP protocol. 2. Network management input methods are professional and accurate, but belong to single-discipline topology, lacking certain business topology continuity and not supporting cross-vendor applications. The proposed solution is to use the PyOD algorithm to perform anomaly detection and normalization on core network configuration data and interface XDR collection data. It utilizes feature information gain measurement to filter reasonable feature subsets and combines the probabilistic relationship between user-related enterprises, DNN, UPF, TAC, AMF, base station IP, and other features with network element topology node paths to determine the probability of enterprise association for nodes, achieving cross-discipline node classification and prediction. Based on business process transaction association, it synthesizes key business scenario private network business path trajectories and automatically generates core network element slice resource topology. This enables real-time updated full-path visualization applications, slice / enterprise name / DNN fuzzy search, key indicator and alarm presentation, and realizes functions such as visibility, manageability, and evaluation of private network services, adapting to the needs of refined private network resource management.

[0072] The technical solutions of the embodiments of this application will be described in detail below with reference to specific examples.

[0073] Reference Figure 1 The flowchart illustrates the steps of a network topology graph generation method provided in an embodiment of this application, as follows: Figure 1 As shown, the network topology map generation method may include the following steps:

[0074] Step 101: Obtain the XDR data, core network configuration data, and B domain data corresponding to the target enterprise.

[0075] The embodiments of this application can be applied to scenarios where a network topology diagram of a target enterprise is constructed.

[0076] This application provides an apparatus for generating enterprise-level slice-level 5G private network topology based on configuration data. For example... Figure 2 As shown, the device for generating enterprise-level slice-level 5G private network topology based on configuration data only requires input of 5G private network activation information, XDR collection, and core network AMF / SMF / UPF configuration data. Through three built-in modules—automatic data conversion, automatic path information generation, and automatic slice panoramic topology drawing—the device automatically outputs an enterprise-level slice-level core network resource topology based on industry customers. This provides the foundation for supporting upper-layer applications to associate alarms and performance data with corresponding devices, facilitating rapid end-to-end business location analysis, reducing network operation and maintenance costs, and improving customer business perception. Figure 2 As shown, 5G cross-domain data can be input into the 5G private network topology generation device to output an enterprise-level slice-level core network topology.

[0077] The target enterprise refers to the enterprise used to construct the network topology map. In this example, the target enterprise can be one or more types of enterprises, such as power companies, internet companies, and vehicle networking companies.

[0078] In practical implementation, when constructing the network topology map of the target enterprise, the corresponding XDR data, core network configuration data, and B-domain data of the target enterprise can be obtained. In practical implementation, the XDR data, core network configuration data, and B-domain data of the target enterprise can be obtained periodically, such as every 3 days, 5 days, or once a week. Specifically, the data collection cycle can be determined according to business needs, and this embodiment does not impose any restrictions on it.

[0079] XDR is a feature of open network computing environments. XDR provides an architecture-independent representation of data, resolving differences in data byte ordering, byte size, data representation, and data alignment. Applications using XDR can exchange data on heterogeneous hardware systems. The RPC / XDR specification and source code are freely available from SunSoft.

[0080] Core network configuration data may include: data such as base stations and core network used by various network devices within the target enterprise.

[0081] B-domain refers to a feature of SunSoft's Open Network Computing Environment. XDR provides an architecture-independent representation of data, resolving differences in data byte ordering, byte size, data representation, and data alignment. Applications using XDR can exchange data on heterogeneous hardware systems. The RPC / XDR specification and source code are freely available from SunSoft.

[0082] After obtaining the XDR data, core network configuration data and B domain data corresponding to the target enterprise, proceed to step 102.

[0083] Step 102: Perform data filtering processing on the XDR data, the core network configuration data, and the B domain data to generate a network data field table.

[0084] After obtaining the XDR data, core network configuration data, and B-domain data corresponding to the target enterprise, data filtering processing can be performed on the XDR data, core network configuration data, and B-domain data to generate a network data field table. In specific implementation, the PyOD algorithm can be used to perform anomaly detection on the parameters in the periodically obtained XDR data, core network configuration data, and B-domain data, filtering dirty data, normalizing outliers in IPv6 addresses and slice format data, filtering and automatically extracting key valid fields, and outputting a key field data table. The specific filtering process can be as follows: 1. Based on the correspondence between base station ID, base station IP, TAC, DNN, slice, and AMFIP, the XDR data is filtered to obtain the target XDR data. 2. Filtering operations are performed on the AMF data, SMF data, and UPF data in the core network configuration data to obtain the target core network configuration data. 3. Based on the target XDR data, target core network configuration data, and B-domain data, a network data field table is generated.

[0085] The process of automatically extracting rules and filtering key valid fields may include the following steps:

[0086] Step A1: For the 5G private network activation information, filter the correspondence between enterprise project information and the attributes of the customized network; wherein, the attributes of the customized network include DNN (Data Network Name), slicing, and private network deployment method.

[0087] Step A2: For the XDR acquisition, filter the correspondence between base station ID, base station IP, TAC (Tracking Area Code), DNN, slice, and AMFIP.

[0088] Step A3: Configure the core network AMF (Access and Mobility Management Function), SMF (Service Management Function), and UPF (User Port Function). Extract slices, various interface service IPs, and the IPs of peer NRF (Network Repository Function) and NSSF (Network Slice Selection Function) network elements from the AMF. Extract slices, DNNs, and their corresponding UPF and TAC information, various interface service IPs, and the IPs of peer NRF and other network elements from the SMF. Extract slices, DNNs, and various interface service IPs from the UPF.

[0089] The implementation process for step A3 above can be as follows: 1. For nodes from different manufacturers, set rules for extracting the corresponding information fields in the configuration based on the operator's data configuration specifications of AMF / SMF / UPF. 2. Use Java and regular expressions to parse the TXT file and convert it into an Excel file. 3. After obtaining the data line by line, use regular expressions to match the matching data and store it in a list. Write the data in the list into an Excel file to automatically generate the data table (i.e., the network data field table).

[0090] After filtering and processing the XDR data, core network configuration data, and B domain data to generate a network data field table, proceed to step 103.

[0091] Step 103: Based on the network data field table, construct the network device service path table corresponding to the network devices within the target enterprise.

[0092] After filtering and processing XDR data, core network configuration data, and B-domain data to generate a network data field table, a network device service path table corresponding to the network devices within the target enterprise can be constructed based on the network data field table. This implementation process can be described in detail below with reference to the specific implementation methods.

[0093] In one specific implementation of this application, step 103 may include:

[0094] Sub-step B1: Based on the 5G private network service logic, select key factors from the network data field table.

[0095] In this embodiment, after generating the network data field table, key factors can be selected from the network data field table based on the 5G private network service logic. Specifically, based on the core network office data configuration combined with XDR collection information fields, the SPARK ML algorithm can be used to perform multi-domain association modeling of service paths, and key factors can be selected from several key fields according to the 5G private network service logic.

[0096] After selecting key factors from the network data field table based on the 5G private network business logic, sub-step B2 is executed.

[0097] Sub-step B2: Verify the key factors based on the information gain of the features to obtain multi-domain features.

[0098] After selecting key factors from the network data field table based on 5G private network service logic, the selected key factors can be validated based on the information gain of the features to obtain multi-domain features. Specifically, the information gain of the features can be used to validate 11 key feature attributes of 5G private network services as feature subsets, outputting a training dataset. The top 11 features with the largest information gain are then used as feature subsets, and the feature attributes selected from these feature subsets are used as multi-domain features.

[0099] After verifying the key factors based on feature-based information gain to obtain multi-domain features, sub-step B3 is executed.

[0100] Sub-step B3: Based on the feature class maximization method of the Naive Bayes model, perform probability analysis on the multi-domain features to obtain the probability of the multi-domain features being associated with the enterprise topology category.

[0101] After verifying key factors based on feature-based information gain to obtain multi-domain features, the multi-domain features can be subjected to probability analysis using the feature class maximization method of the Naive Bayes model to obtain the probability of association between multi-domain features and enterprise topology categories.

[0102] Sub-step B4: Select target multi-domain features from the multi-domain features based on the probability to construct the network device service path table.

[0103] Then, target multi-domain features can be selected from the multi-domain features based on the obtained probabilities to construct the network device service path table. Specifically, the accuracy of service path identification can be improved by combining multi-domain features: combining the feature maximization method based on the Naive Bayes model, the probability of whether a node feature value is associated with an enterprise topology category can be inferred by using the probabilistic relationship between features such as user-related enterprises, DNN, UPF, TAC, AMF, and base station ID and network element topology node paths, thus realizing cross-professional node classification and identification prediction, and outputting the "Service Path Association Wide Table", i.e., the network device service path table.

[0104] After constructing the network device service path table corresponding to the network devices within the target enterprise based on the network data field table, proceed to step 104.

[0105] Step 104: Based on the configuration conversion table of each node in the core network and the service path table of the network devices, draw the network topology diagram corresponding to the target enterprise.

[0106] After constructing the network device service path table corresponding to the network devices within the target enterprise based on the network data field table, the network topology diagram corresponding to the target enterprise can be drawn based on the configuration translation table of each node in the core network and the network device service path table. This implementation process can be described in detail below with reference to the specific implementation method.

[0107] In one specific implementation of this application, step 104 may include:

[0108] Sub-step C1: Based on the configuration conversion table of each node in the core network and the service path table of the network devices, obtain the service logic for drawing the overall network topology of the core network.

[0109] In this embodiment of the application, after obtaining the network device service path table, the service logic for drawing the overall network topology map of the core network can be obtained based on the configuration conversion table of each node of the core network and the network device service path table.

[0110] After obtaining the business logic for drawing the overall network topology map of the core network by configuring the conversion table and network device service path table for each node of the core network, sub-step C2 is executed.

[0111] Sub-step C2: Convert the raw data in the network data field table into a data model using a data algorithm.

[0112] The raw data in the network data field table can be converted into a data model using data algorithms.

[0113] Sub-step C3: Based on the business logic, draw the data model as an initial network topology diagram.

[0114] After obtaining the data model, it can be used to draw an initial network topology diagram based on business logic.

[0115] After drawing the data model into an initial network topology based on business logic, execute sub-step C4.

[0116] Sub-step C4: Post-process the initial network topology map to generate the network topology map corresponding to the target enterprise.

[0117] After drawing an initial network topology diagram based on the business logic, the initial network topology diagram can be post-processed to generate the network topology diagram corresponding to the target enterprise.

[0118] In practical implementation, data is collected and stored in the database through the configuration transformation table of each core network node and the wide table associated with business paths. This establishes a data model of the core network element slice topology related to enterprises / slices / DNNs, yielding the business logic for drawing the overall core network topology map on the view end. Data algorithms transform the raw data into a data model. SVG technology is applied to draw the data model into a network topology map, HTML5 is used to build the box model, and CSS3 is used for rendering styles. JavaScript is used to implement human-computer interaction functions, enabling adaptive icons and connections. AJAX is used in the interaction to achieve asynchronous data updates.

[0119] During the above process, the AMF / SMF pools in DC1 and DC2 can be drawn based on the corresponding AMF / SMF information in the "Service Path Association Wide Table" (i.e., the network device service path table).

[0120] Based on the corresponding UPF / private network deployment information in the "Business Path Association Wide Table", draw different UPF network topologies.

[0121] Based on the corresponding peer NSSF / NRF information in the "Business Path Association Wide Table", draw the NSSF / NRF in DC1 and DC2.

[0122] Then, by combining business logic and data models to draw the network topology diagram, and after further processing, the network topology diagram corresponding to the target enterprise can be obtained.

[0123] This application embodiment can automatically output cross-vendor network resource topologies based on enterprise-level and slice-level data for industry customers, facilitating topology visualization. This enables real-time updated full-path visualization applications, slice / enterprise name / DNN fuzzy search, key indicator and alarm presentation, and other functions such as visibility, manageability and evaluation of private network services, adapting to the needs of refined management of private network resources.

[0124] Next, the technical solutions of the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0125] Reference Figure 3 The diagram illustrates a process for generating an enterprise-level slice-level 5G private network topology based on configuration data, as provided in an embodiment of this application.

[0126] like Figure 3 As shown, the device includes: an automatic data conversion module, an automatic path information generation module, and an automatic slice panoramic topology drawing module.

[0127] The automatic data conversion module can be used to filter and automatically extract key effective fields from 5G cross-domain data, and output several key field data tables; wherein, the 5G cross-domain data includes 5G private network activation information, XDR data, and core network AMF / SMF / UPF configuration from B-domain information, such as... Figure 4 As shown.

[0128] The automatic path information generation module can be used to combine core network office data configuration with XDR-collected information fields, select key feature factors from several key fields according to 5G private network service logic, refine correlation models to improve the efficiency of network device service path analysis, use multi-domain feature combination judgment to improve the accuracy of service path identification, and automatically generate network devices used in the network, such as... Figure 5 As shown.

[0129] The automatic topology drawing module for slices can collect and store data from the core network node configuration transformation table and the business path association wide table, establish a data model of the core network element slice topology related to enterprises / slices / DNN, and obtain the business logic for drawing the overall core network topology map on the view end. Then, data algorithms are used to transform the raw data into a data model. The view end can use SVG technology to draw the data model into a network topology map, and implement human-computer interaction functions through JavaScript, outputting the automatically drawn core network DNN / slice topology.

[0130] The method for generating enterprise-level slice-level 5G private network topology based on configuration data provided in this application embodiment may include steps S1, S2 and S3.

[0131] Step S1: Obtain the key field data table through anomaly detection. Specifically, this includes: using the PyOD algorithm to perform anomaly detection on parameters in periodically obtained XDR data, core network configuration data, and B-domain data; filtering dirty data; normalizing outlier values ​​in IPv6 addresses and slice format data; filtering and automatically extracting key valid fields; and outputting the key field data table.

[0132] This step specifically includes setting up automatic extraction rules and filtering key valid fields:

[0133] S11. For the 5G private network activation information, filter the correspondence between enterprise project information and the attributes of the customized network; wherein, the attributes of the customized network include DNN, slicing, and private network deployment method, such as... Figure 6 As shown.

[0134] S12. For the XDR acquisition, the correspondence between base station ID, base station IP, TAC, DNN, slice and AMFIP is filtered. For specific fields, refer to cell id, gnb ip add, Tac, pdu session1dnn, Allowed NSSAI 1SD, amf ip add in the interface data dictionary intf_5g_n1n2.

[0135] S13. Configure the core network AMF / SMF / UPF. Extract slices, various interface service IPs, and the IPs of peer NRF and NSSF network elements from the AMF; extract slices, DNN and their corresponding UPF and TAC information, various interface service IPs, and the IPs of peer NRF and other network elements from the SMF; extract slices, DNN, and various interface service IPs from the UPF. For nodes from different manufacturers, the extraction logic rules for the corresponding information fields in the configuration are set based on the operator's data configuration specifications for AMF / SMF / UPF. Use Java and regular expressions to parse the TXT file into an Excel file, extract data line by line, match matching data with regular expressions and store them in a list, and write the data in the list into Excel to achieve automated data table generation. Figures 7-9 As shown.

[0136] Step S2: Based on the core network bureau data configuration and XDR collected information fields, the SPARK ML algorithm is used to perform multi-domain association modeling of service paths. According to the 5G private network service logic, key factors are selected from key fields. The information gain measurement method of features is used to verify key feature attributes with high information gain, such as enterprise name, network type, slice, DNN, base station, AMF, SMF, UPF, TAC, A device, and B device, as feature subsets. The association model is refined to improve the service path analysis efficiency of network devices. Multi-domain feature combination judgment is used to improve the accuracy of service path identification and automatically generate the network devices used in the network.

[0137] Step S2 above may include:

[0138] S21. Based on the core network bureau data configuration and combined with XDR collected information fields, the SPARK ML algorithm is used to perform multi-domain association modeling of business paths, and key factors are selected from key fields according to the 5G private network business logic.

[0139] S22. Using the information gain metric, verify key feature attributes with high information gain, such as enterprise name, network type, slice, DNN, base station, AMF, SMF, UPF, TAC, device A, and device B, as a feature subset. The feature attributes selected from this subset include, for example: Figure 10 As shown;

[0140] S23. Improve the accuracy of business path identification by combining multi-domain features: Combine the feature class maximization method based on the Naive Bayes model, and infer the probability of whether the node feature value is associated with the enterprise topology category by using the probability relationship between features such as user-related enterprises, DNN, UPF, TAC, AMF, base station ID and network element topology node path, so as to realize cross-professional node classification identification and prediction, and output the "Business Path Association Wide Table".

[0141] In the specific scheme, the feature subset verification steps are as follows: Calculate the total information entropy of the given sample: H(C)=-∑P(Ci)log(2,P(Ci)) (i=1,2,..n), and calculate the conditional entropy of the fixed feature T: T represents a feature (e.g., general DNN and custom DNN), and t represents the occurrence of T. This means T does not appear; the difference between the two is the information gain. The larger the gain value, the more information that feature T provides for classification, and the less uncertainty there is in classification after selecting T. Information gain = original entropy of the system - conditional entropy of fixed feature T, as shown in the following formula:

[0142]

[0143] Select the top feature T with the largest information gain as the feature subset and output the training samples.

[0144] Furthermore, the accuracy of business path identification is improved by using multi-domain feature combination judgment: combining the feature class maximization method based on the Naive Bayes model, the probability of whether the node feature value is associated with the enterprise topology category is inferred by the probabilistic relationship between features such as user-related enterprises, DNN, UPF, TAC, AMF, base station ID and network element topology node path, so as to realize cross-professional node classification identification and prediction, and output the "Business Path Association Wide Table".

[0145] In the specific scheme, the classification and recognition algorithm based on the Naive Bayes model is used: P(B|A) represents the probability that B (belonging to an enterprise base station) occurs when A (base station ID) has occurred;

[0146] Taking the identification of whether "customized DNN" features belong to "enterprise base stations" as an example:

[0147] The following steps can be used to identify the class that maximizes P(custom DNN | belongs to enterprise base stations), thereby determining whether a base station belongs to the enterprise base stations in the TOP:

[0148] P (belongs to enterprise base station | customized DNN) =

[0149] P(Custom DNN | Belongs to Enterprise Base Station) P(Belongs to Enterprise Base Station) / P(Custom DNN)

[0150] P (not belonging to enterprise base stations | customized DNN) =

[0151] P(Custom DNN | Not belonging to enterprise base station) P(Not belonging to enterprise base station) / P(Custom DNN)

[0152] Comparing the two probabilities, if P(belongs to enterprise base station | customized DNN) > P(does not belong to enterprise base station | customized DNN), then the node belongs to the enterprise base station in the service topology; substitute the aforementioned 11 features, including enterprise, DNN, UPF, TAC, AMF, and base station ID, into the Ci variable: P(X|C i (i = 1, 2, ..., m), find P(X|C) i )P(C i By maximizing the class, we can infer whether the node feature value is related to the enterprise topology category; similarly, we can deduce the maximized feature class for categories such as enterprise bearer A, enterprise bearer B, and enterprise core network.

[0153] In the specific scheme, the classification and recognition method is to obtain training samples through data preprocessing, estimate the probability of each category, estimate the prior probability of feature attributes, calculate the probability of each attribute combination belonging to each category, and select the class C_i with the highest posterior probability as the inferred result of the data.

[0154] Step S3 involves collecting and storing data from the core network node configuration conversion table and the business path association wide table to establish a data model of the core network element slice topology related to enterprises / slices / DNNs. This provides the business logic for drawing the overall core network topology diagram on the view end. The raw data is then transformed into a data model using data algorithms. SVG technology is applied to draw the data model into a network topology diagram. HTML5 is used to build the box model, CSS3 is used for rendering styles, and finally, JavaScript is used to implement human-computer interaction functions, enabling adaptive icons and connections. AJAX is used in the interaction to achieve asynchronous data updates.

[0155] Step S3 above may include:

[0156] S31. By collecting and storing data from the configuration conversion table of each node in the core network and the wide table associated with the business path, a data model of the core network element slice topology related to the enterprise / slice / DNN is established, and the business logic for drawing the overall network topology of the core network is obtained on the view end.

[0157] S32. Transform the raw data into a data model using data algorithms.

[0158] S33. Use SVG technology to draw the data model into a network topology diagram, use HTML5 to build the box model, and use CSS3 to render the style.

[0159] S34. Implement human-computer interaction functions using JavaScript, achieve adaptive icons and lines, and use AJAX in the interaction to achieve asynchronous data updates.

[0160] like Figure 11 As shown, it can output the automatically drawn core network DNN / slice topology.

[0161] The specific implementation process is as follows:

[0162] 1. Examples of specific data transformation

[0163] 5G private network activation information filtering key fields are converted into customer information and logical business-related DNN and SNSSAI information, see [link / reference]. Figure 12 For example, the DNN and slice corresponding to a company name.

[0164] The configuration of core network professional function nodes (AMF / SMF / UPF) is extracted from a business dimension. The specific steps are as follows: A scheduled task is set up to pull the latest files from the SFTP server where the configuration is stored to the local server. After obtaining the file buffered stream using Java, regular expressions are used to collect data. The collected data is then processed through Cartesian product operations to obtain the final data table result, and the data is stored in the database using DataTable and QueryAgent. Through the above process, automatic configuration transformation associates the DNN in the core network configuration with the TAC information in the configuration, such as... Figure 13 As shown.

[0165] 2. Specific examples of automatically generating path information

[0166] Based on the 5G private network service logic, 11 key factors were selected from 45 key fields. The information gain metric was used to verify these 11 key feature attributes, including enterprise name, network type, slice, DNN, base station, AMF, SMF, UPF, TAC, A device, and B device, as a feature subset. The information gain algorithm was then used to verify the 11 key features of the 5G private network service. The top 11 features T with the highest information gain were verified and met the criteria for being selected as a feature subset. Training samples were then output. The feature attributes after feature subset selection are shown in Table 1 below.

[0167] Table 1:

[0168]

[0169]

[0170]

[0171] Improving Business Path Identification Accuracy Through Multi-Domain Feature Combination Judgment: Combining the feature-maximization method based on the Naive Bayes model, the probability of whether a node feature value is associated with an enterprise topology category is inferred by using the probabilistic relationship between features such as user-related enterprises, DNN, UPF, TAC, AMF, and base station ID and network element topology node paths. This enables cross-professional node classification and prediction, outputting a "Business Path Association Wide Table". In a practical use case, taking the category attribute of belonging to an enterprise base station as an example, the specific classification and identification process can be as follows:

[0172] 1) In the XDR message and configuration file information containing 100 training samples, according to the classification algorithm of this patent, the class prior probability P (belonging to enterprise base station) and the conditional probability P(x|Ci) of each attribute feature under each class condition are calculated using the sample category attribute (taking belonging to enterprise base station as an example): taking P (customized DNN|belonging to enterprise base station) as an example.

[0173] The prior probability P (belonging to an enterprise base station) obtained from the training samples is as follows, that is, the probability of belonging to an enterprise base station in the training samples is:

[0174] Enterprise base station count = select count(distinct phy_cell_id)from dic_wypt_wy_db_nr_ccdu where is_twob_cell = 'is' is 764.

[0175] Non-enterprise base station count = select count(distinct phy_cell_id)from dic_wypt_wy_db_nr_ccdu where is_twob_cell='No' is 89486.

[0176] Total 90,250

[0177] P(belonging to enterprise base stations) = 0.0085, p(not belonging to enterprise base stations) = 0.991.

[0178] 2) The probability obtained in step 1 is used as prior knowledge to calculate the likelihood probability P(X|Ci) (i=1,2,...,m) of the feature condition X under the category condition Ci for each node with each “customized DNN” feature, taking Ci as an example of belonging to an enterprise base station.

[0179] P (customized DNN | belongs to enterprise base station) = select count (distinct case when if_2bdnn = '1' then msisdn end) / count (distinct msisdn) from st_cl_user_day where day = 20221022and phy_cell_id in (select phy_cell_id from dic_wypt_wy_db_nr_ccduwhere is_twob_cell='is')=0.901.

[0180] P(Custom DNN|Not belonging to enterprise base station) = select count(distinct case when if_2bdnn='1'then msisdn end) / count(distinct msisdn)from st_cl_user_day where day=20221022and phy_cell_id in(select phy_cell_id from dic_wypt_wy_db_nr_ccduwhere is_twob_cell='No') = 0.00313.

[0181] 3) Based on the conditional independence assumption, we only need to find the class that maximizes P(X|Ci)P(Ci).

[0182] P(Custom DNN | Belongs to Enterprise Base Station) p(Belongs to Enterprise Base Station) = 0.901 × 0.0085 = 0.0076585.

[0183] P(Custom DNN | Not belonging to enterprise base station) p(Not belonging to enterprise base station) = 0.0031 × 0.991 = 0.0030721.

[0184] Therefore, for a base station X with customized DNN features, its posterior probability of belonging to class 1 (0.0076585) is greater than its posterior probability of belonging to class 0 (0.0030721), meaning that the class of X is □=1, which means it belongs to an enterprise base station.

[0185] Based on the method described above, cross-professional node classification, identification, and prediction are achieved, and a "Business Path Association Wide Table" is output.

[0186] 3. Specific Topology Drawing Examples

[0187] Based on the configuration conversion table and service path association wide table of each node in the core network, data is collected and stored in the database to establish a data model of the core network element slice topology related to enterprises / slices / DNNs. In a specific example:

[0188] Based on the corresponding AMF / SMF information in the "Business Path Association Wide Table", draw the AMF / SMF pool in DC1 and DC2.

[0189] Based on the corresponding UPF / private network deployment information in the "Business Path Association Wide Table", draw different UPF network topologies.

[0190] Based on the corresponding peer NSSF / NRF information in the "Business Path Association Wide Table", draw the NSSF / NRF in DC1 and DC2.

[0191] The data model is drawn into a network topology diagram using SVG technology, the box model is built using HTML5, and the style is rendered using CSS3.

[0192] In terms of resource data collection, this implementation method uses access event detection to obtain full core network configuration data, judges the time and data type of the data based on thresholds learned by machine learning, selects an appropriate time for processing, and automatically filters key valid fields and automatically converts them into the database, thereby improving the real-time performance of topology drawing. Regarding automatic path information generation: The PyOD algorithm is used to perform anomaly detection and normalization on core network configuration data and interface XDR collection data. Feature information gain measurement is used to verify the rationality of the selected feature subset. Innovatively, the probability relationship between user-related enterprise, DNN, UPF, TAC, AMF, base station IP, and other features and network element topology node paths is combined to determine the enterprise association probability of a node, achieving cross-professional node classification and prediction. Regarding front-end view presentation: SVG technology is used to draw the data model into a network topology diagram. HTML5 is used to build the box model, CSS3 is used for rendering styles, and finally, JavaScript is used to implement human-computer interaction functions, enabling adaptive icons and connections. AJAX is used in the interaction to achieve asynchronous data updates, flexibly adapting to three deployment methods: private network Seeyon, Bilin, and Ruyi. Regarding enterprise-level topology: The current core network topology lacks the function of querying and generating network topologies corresponding to a specific slice service or enterprise service, while also addressing the pain point of core network topologies not being cross-vendor.

[0193] The network topology map generation method provided in this application embodiment obtains XDR data, core network configuration data, and B-domain data corresponding to the target enterprise. Data filtering processing is performed on the XDR data, core network configuration data, and B-domain data to generate a network data field table. Based on the network data field table, a network device service path table corresponding to the network devices within the target enterprise is constructed. Based on the core network node configuration conversion table and the network device service path table, the network topology map corresponding to the target enterprise is drawn. This application embodiment can automatically output cross-vendor network resource topologies at the enterprise level and slice level based on industry customers, facilitating topology visualization. This enables real-time updated full-path visualization applications, slice / enterprise name / DNN fuzzy search, key indicator and alarm presentation, and achieves functions such as visibility, manageability, and evaluation of private network services, adapting to the needs of refined management of private network resources.

[0194] Reference Figure 14 The diagram illustrates a network topology graph generation apparatus according to an embodiment of this application. Figure 14 As shown, the network topology map generation device 1400 may include the following modules:

[0195] Data acquisition module 1410 is used to acquire XDR data, core network configuration data and B domain data corresponding to the target enterprise;

[0196] The field table generation module 1420 is used to perform data filtering processing on the XDR data, the core network configuration data and the B domain data to generate a network data field table.

[0197] The path table construction module 1430 is used to construct a network device service path table corresponding to the network devices within the target enterprise based on the network data field table.

[0198] The topology drawing module 1440 is used to draw the network topology diagram corresponding to the target enterprise based on the configuration conversion table of each node in the core network and the service path table of the network devices.

[0199] Optionally, the data acquisition module includes:

[0200] The data acquisition unit is used to collect XDR data, core network configuration data and B domain data corresponding to the target enterprise based on a set period.

[0201] Optionally, the field table generation module includes:

[0202] The XDR data filtering unit is used to filter the XDR data based on the correspondence between base station ID, base station IP, TAC, DNN, slice and AMFIP to obtain the filtered target XDR data.

[0203] A data filtering unit is configured to perform filtering operations on AMF data, SMF data, and UPF data in the core network configuration data to obtain the target core network configuration data to be filtered.

[0204] The field table generation unit is used to generate the network data field table based on the target XDR data, the target core network configuration data, and the B domain data.

[0205] Optionally, the path table construction module includes:

[0206] The key factor filtering unit is used to filter key factors from the network data field table based on 5G private network service logic.

[0207] A multi-domain feature acquisition unit is used to verify the key factors based on the information gain of the features to obtain multi-domain features;

[0208] The probability acquisition unit is used to perform probability analysis on the multi-domain features according to the feature class maximization method of the Naive Bayes model, and obtain the probability of the multi-domain features being associated with the enterprise topology category.

[0209] The path table construction unit is used to filter target multi-domain features from the multi-domain features according to the probability, so as to construct the network device service path table.

[0210] Optionally, the topology drawing module includes:

[0211] The business logic acquisition unit is used to obtain the business logic for drawing the overall network topology of the core network based on the configuration conversion table of each node of the core network and the service path table of the network device.

[0212] A data model conversion unit is used to convert the original data in the network data field table into a data model using a data algorithm;

[0213] An initial graph drawing unit is used to draw the data model into an initial network topology graph based on the business logic;

[0214] The topology graph generation unit is used to post-process the initial network topology graph to generate the network topology graph corresponding to the target enterprise.

[0215] The network topology generation device provided in this application embodiment acquires XDR data, core network configuration data, and B-domain data corresponding to the target enterprise. It then performs data filtering processing on the XDR data, core network configuration data, and B-domain data to generate a network data field table. Based on the network data field table, it constructs a network device service path table corresponding to the network devices within the target enterprise. Based on the core network node configuration conversion table and the network device service path table, it draws the network topology diagram corresponding to the target enterprise. This application embodiment can automatically output cross-vendor network resource topologies at the enterprise level and slice level based on industry customers, facilitating topology visualization. This enables real-time updated full-path visualization applications, slice / enterprise name / DNN fuzzy search, key indicator and alarm presentation, and achieves functions such as visibility, manageability, and evaluation of private network services, adapting to the needs of refined management of private network resources.

[0216] This application also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the above-described network topology graph generation method.

[0217] Figure 15 A schematic diagram of the structure of an electronic device 1500 according to an embodiment of the present invention is shown. Figure 15 As shown, the electronic device 1500 includes a central processing unit (CPU) 1501, which can perform various appropriate actions and processes according to computer program instructions stored in read-only memory (ROM) 1502 or loaded from storage unit 1508 into random access memory (RAM) 1503. The RAM 1503 may also store various programs and data required for the operation of the electronic device 1500. The CPU 1501, ROM 1502, and RAM 1503 are interconnected via bus 1504. An input / output (I / O) interface 1505 is also connected to bus 1504.

[0218] Multiple components in electronic device 1500 are connected to I / O interface 1505, including: input unit 1506, such as keyboard, mouse, microphone, etc.; output unit 1507, such as various types of monitors, speakers, etc.; storage unit 1508, such as disk, optical disk, etc.; and communication unit 1509, such as network card, modem, wireless transceiver, etc. Communication unit 1509 allows electronic device 1500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0219] The various processes and handling described above can be executed by processing unit 1501. For example, the methods of any of the above embodiments can be implemented as computer software programs tangibly contained in a computer-readable medium, such as storage unit 1508. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 1500 via ROM 1502 and / or communication unit 1509. When the computer program is loaded into RAM 1503 and executed by CPU 1501, one or more actions of the methods described above can be performed.

[0220] Additionally, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned network topology graph generation method.

[0221] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0222] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0223] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminals (systems), and computer program products according to embodiments of this application. It should 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 program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0224] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxesFigure 1 The function specified in one or more boxes.

[0225] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal, causing a series of operational steps to be executed on the computer or other programmable terminal to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0226] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0227] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal that includes said element.

[0228] The foregoing has provided a detailed description of a network topology diagram generation method, a network topology diagram generation device, an electronic device, and a computer-readable storage medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A network topology map generation method, characterized by, The method comprises: obtaining XDR data, core network configuration data and B domain data corresponding to a target enterprise; performing data filtering processing on the XDR data, the core network configuration data and the B domain data to generate a network data field table; based on the network data field table, constructing a network equipment service path table corresponding to network equipment in the target enterprise; based on the node configuration conversion table of the core network and the network equipment service path table, drawing a network topology graph corresponding to the target enterprise; wherein, based on the network data field table, constructing a network equipment service path table corresponding to network equipment in the target enterprise, comprising: based on the 5G private network service logic, screening out key factors from the network data field table; based on the information gain of the characteristics, verifying the key factors to obtain multi-domain characteristics; according to the maximum feature class method of the naive Bayes model, performing probability analysis processing on the multi-domain characteristics to obtain the probability of the association between the multi-domain characteristics and the enterprise topology category; according to the probability, screening out target multi-domain characteristics from the multi-domain characteristics to construct the network equipment service path table.

2. The method of claim 1, wherein, The method comprises: based on a set period, collecting XDR data, core network configuration data and B domain data corresponding to the target enterprise.

3. The method of claim 1, wherein, The method comprises: based on the correspondence between the base station ID, the base station IP, the TAC, the DNN, the slice and the AMF IP, performing filtering processing on the XDR data to obtain filtered target XDR data; performing filtering operation on the AMF data, the SMF data and the UPF data in the core network configuration data to obtain filtered target core network configuration data; based on the target XDR data, the target core network configuration data and the B domain data, generating the network data field table.

4. The method of claim 1, wherein, The method comprises: based on the node configuration conversion table of the core network and the network equipment service path table, obtaining the service logic of drawing the overall network topology graph of the core network; by data algorithm, converting the original data in the network data field table into a data model; based on the service logic, drawing the data model into an initial network topology graph; performing post-processing on the initial network topology graph to generate the network topology graph corresponding to the target enterprise.

5. A network topology map generation apparatus characterized by comprising: The device comprises: a data acquisition module for obtaining XDR data, core network configuration data and B domain data corresponding to a target enterprise; a field table generation module for performing data filtering processing on the XDR data, the core network configuration data and the B domain data to generate a network data field table; a path table construction module for constructing a network equipment service path table corresponding to network equipment in the target enterprise based on the network data field table; A topology graph drawing module is configured to draw a network topology graph corresponding to the target enterprise based on the node configuration conversion table of the core network and the network device service path table. The path table construction module comprises: A key factor screening unit is configured to screen key factors from the network data field table based on 5G private network service logic; A multi-domain feature acquisition unit is configured to verify the key factors based on information gain of features to obtain multi-domain features; A probability acquisition unit is configured to perform probability analysis on the multi-domain features according to a maximum feature class method of a naive Bayes model to obtain a probability of association between the multi-domain features and enterprise topology categories; A path table construction unit is configured to screen target multi-domain features from the multi-domain features according to the probability to construct the network device service path table.

6. The apparatus of claim 5, wherein, The data acquisition module comprises: A data acquisition unit is configured to acquire XDR data, core network configuration data and B-domain data corresponding to the target enterprise based on a set period.

7. The apparatus of claim 5, wherein, The field table generation module comprises: An XDR data screening unit is configured to screen the XDR data based on a corresponding relationship between base station IDs, base station IPs, TACs, DNNs, slices and AMF IPs to obtain target XDR data; A configuration data screening unit is configured to perform a screening operation on AMF data, SMF data and UPF data in the core network configuration data to obtain target core network configuration data; A field table generation unit is configured to generate the network data field table based on the target XDR data, the target core network configuration data and the B-domain data.

8. An electronic device, comprising: The computer program is stored in the memory and executable on the processor, and the processor executes the program to implement the network topology graph generation method in any one of claims 1 to 4. When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device can execute the network topology graph generation method in any one of claims 1 to 4.

9. A computer-readable storage medium, characterized in that, ​

Images