A method, apparatus and processing method for locating network problems based on big data

Abstract

This invention discloses a network problem localization method, apparatus, and processing method based on big data. The network problem localization method is applied to the network side and includes the following steps: acquiring user network problem data; determining, based on the user's network problem data, whether the user's network problem is an individual user problem or a group user network problem; if it is an individual user network problem, then localizing the problem based on an expert database model; if it is a group user network problem, then localizing the problem based on the expert database model for the entire group of users; the expert database model is a pre-constructed network problem localization model based on big data, and the expert database model carries historical network problems and historical network problem localization results. This localization method can achieve problem localization based on big data, thereby improving the efficiency and accuracy of network problem localization.

Description

Technical Field

[0001] This invention relates to the field of network technology, and in particular to a method, apparatus and processing method for locating network problems based on big data. Background Technology

[0002] With the rapid development of mobile networks, mobile phone use has become ubiquitous, and almost everyone relies on them in their daily work and life. People use their phones daily for internet access, shopping, social networking, and work. However, users may experience dissatisfaction with mobile network services, such as difficulty making voice calls, inability to access the internet, or slow internet speeds. Dissatisfied users typically file customer complaints, making the frequency of customer complaints a crucial performance indicator for telecom operators. For telecom operators, focusing on customer needs, reducing complaint rates, and improving customer satisfaction are key aspects of daily operations. As living standards improve, customer expectations for telecom service quality are also rising. The increasing number and frequency of customer complaints pose challenges to the service quality and efficiency of telecom operators.

[0003] Currently, when troubleshooting network issues, after a user files a complaint, customer service personnel guide the user step-by-step through a series of problems based on the user's prompts. However, this approach has significant drawbacks. 1. Due to the varying skill levels of customer personnel and the differing experience and tools available to cloud network technicians, the efficiency and accuracy of network problem localization can be low. Consequently, the efficiency and effectiveness of resolving these network issues may not meet user satisfaction, and sometimes this can even lead to escalating complaints. 2. This method of locating network problems is relatively inefficient and passive, which has a significant negative impact on the service reputation of operators. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to address the above-mentioned shortcomings of the prior art by proposing a network problem localization method, device and processing method based on big data. This localization method can realize problem localization based on big data, thereby improving the efficiency and accuracy of network problem localization.

[0005] In a first aspect, the present invention provides a network problem localization method based on big data, which is applied to the network side and includes the following steps: Step S1: Obtain the user's network problem data; Step S2: Based on the user's network problem data, determine whether the user's network problem is an isolated issue or a problem affecting a group of users. If the network problem is an individual user's network issue, the problem will be located based on the expert database model; if the network problem is a group of users' network issues, the problem will be located based on the expert database model. The expert database model is a network problem localization model pre-built based on big data, and the expert database model carries historical network problems and historical network problem localization results.

[0006] Furthermore, step S1 specifically includes: By using the Kafka tool to connect to the real-time network data system, we can obtain complaint order data, signaling data, MRO data, and B-domain data to obtain network problem data.

[0007] Furthermore, the step of locating network problems for individual users based on the expert database model specifically includes the following steps: A1: Obtain the cell coverage performance of the user's service trajectory and determine whether there are cell coverage problems based on the expert database model: If a cell coverage issue exists, the initial location of the network problem for an individual user is obtained; if no cell coverage issue exists, the initial location of the network problem for an individual user is not obtained. A2: Obtain the core network signaling success rate, which includes the attachment success rate, handover success rate, and PDN request success rate; A3: Based on the expert database model, one or more core network signaling success rates that are lower than expected or abnormal are identified, thus obtaining a secondary location of the network problem for individual users; A4: By combining the first and second localizations of the individual user's network problem, the localization of the individual user's network problem is obtained.

[0008] Furthermore, the step of locating network problems for group users based on the expert database model specifically includes: Based on the expert database model and user package clustering, network problems for specific user groups can be identified. And / or, Based on the expert database model and clustering by user terminal brand and model, network problems among group users are located. And / or, Based on the expert database model and user service performance clustering, network problems of group users are located.

[0009] Furthermore, step S0 is included before step S1. Step S0: Construct an expert database model for network problem localization, which includes the following steps: Step S01: Based on historical data, obtain training and test samples for network problem localization; Step S02: Label the training samples, including network problems and their corresponding localization. Step S03: Train the initial expert database model using labeled training samples to obtain a training model, wherein the initial expert database model is a selected machine learning or deep learning model; Step S04: Use test samples to evaluate and optimize the trained model, thereby constructing an expert database model for network problem localization.

[0010] Secondly, the present invention provides a network problem localization device based on big data, which is applied to the network side and includes: The acquisition unit is used to acquire network problem data from users. The judgment unit, connected to the acquisition unit, is used to determine whether the user's network problem is an individual user's network problem or a group of users' network problems based on the user's network problem data. The first positioning unit, connected to the judgment unit, is used to locate the network problem of the individual user based on the expert database model after the judgment unit determines that the user's network problem is an individual user's network problem. The second positioning unit, connected to the judgment unit, is used to locate the network problem of the group users based on the expert database model after the judgment unit determines that the user's network problem is a group user network problem. The expert database model is a network problem localization model pre-built based on big data, and the expert database model carries historical network problems and historical network problem localization results.

[0011] Furthermore, the acquisition unit includes: The interface module is used to interface with a real-time network data system via the Kafka tool; The acquisition module, connected to the docking module, is used to acquire complaint order data, signaling data, MRO data, and B-domain data from the real-time network data system to obtain network problem data.

[0012] Furthermore, the first positioning unit includes: The first acquisition module is used to acquire the cell coverage performance of the user's service trajectory; The first judgment module, connected to the first acquisition module, is used to determine whether a cell coverage problem exists based on the expert database model. If a cell coverage issue exists, the initial location of the network problem for an individual user is obtained; if no cell coverage issue exists, the initial location of the network problem for an individual user is not obtained. The second acquisition module is used to acquire the core network signaling success rate, which includes the attachment success rate, the handover success rate, and the PDN request success rate. The second judgment module, connected to the second acquisition module, is used to determine, based on the expert database model, one or more core network signaling success rates that are lower than expected or abnormal, and to obtain the second location of network problems for individual users. The module is connected to the first judgment module and the second judgment module respectively, and is used to combine the first location of the network problem of an individual user with the second location of the network problem of an individual user to obtain the location of the network problem of an individual user.

[0013] Furthermore, the second positioning unit includes a first positioning module, a second positioning module, and a third positioning module connected in parallel. The first positioning module is used to locate network problems of a group of users based on the expert database model and user package clustering. The second positioning module is used to locate network problems of a group of users based on the expert database model and clustering based on user terminal brand and model. The third positioning module is used to locate network problems of a group of users based on the expert database model and user service performance clustering.

[0014] Thirdly, the present invention provides a network problem handling method based on big data, the method comprising the following steps: Based on the big data-based network problem localization method described in the first aspect, the user's network problem is located, and the localization analysis results are obtained; Determine whether a user's network problem constitutes a network problem complaint: If a user's network problem is a reported network problem, then the complaint will be processed based on the location analysis results; if a user's network problem is not a reported network problem, then an early warning will be issued based on the location analysis results.

[0015] The beneficial effects of this invention are: 1. This invention enables problem localization based on big data, thereby improving the efficiency and accuracy of network problem localization.

[0016] 2. This invention constructs an analytical method from three dimensions: point-based problem analysis, group user problem analysis, and pre-complaint analysis. It considers user packages, user terminals, and network performance, and solidifies the model into a pre-warning module. This improves processing efficiency, provides early warnings of complaints, reduces the scale of user complaints, and overall enhances user satisfaction with complaint handling, thereby improving the company's reputation. Details are as follows: 2.1. For point-based complaints, the time, location, and user information of the complaint can be combined to quickly trace the user's historical signaling plane and user plane data. The cause of the problem can be located by considering factors such as the success or failure of the signaling process, the end-to-end establishment latency of the business, the uplink and downlink speeds of the business, and whether the BO domain contract information is consistent. 2.2. For the problem localization of a certain type of user complaints, such as a set of complaints about slow internet speed within a month, we can correlate the complaint timestamp with the time window of O-domain data, and perform cluster analysis from dimensions such as package profile, terminal profile, service profile, and geographic profile. If there are obvious clustering characteristics, we can further conduct detailed analysis on a certain type of characteristic, thereby achieving the ability to localize problems from the surface to the line and then to the point. 2.3. For certain problems that can be fixed, analysis models can be formed into machine learning models and monitoring tools, thereby monitoring the user's signaling status in real time, discovering potential experience problems before user complaints, and investigating network problems through SMS care, network optimization, etc., thereby improving user business satisfaction and service reputation. Attached Figure Description

[0017] Figure 1 This is an overall schematic diagram of an embodiment of the present invention; Figure 2 This is a schematic diagram of a network problem localization method based on big data in an embodiment of the present invention; Figure 3 This is a schematic diagram of a network problem localization device based on big data in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the distribution of complaint user packages in an embodiment of the present invention; Figure 5 This is a schematic diagram showing the distribution of 4G terminal brands among all network users in this embodiment of the invention; Figure 6 This is a schematic diagram showing the distribution of 5G terminal brands among all users in this embodiment of the invention; Figure 7 This is a schematic diagram showing the distribution of 4G terminal brands among complaining users in an embodiment of the present invention; Figure 8 This is a schematic diagram showing the distribution of 5G terminal brands among complaining users in an embodiment of the present invention; Figure 9 This is a schematic diagram comparing the brand distribution of terminals of users across the entire network and those who have filed complaints, as described in this embodiment of the invention.

[0018] Reference numerals: 10, acquisition unit; 20, judgment unit; 30, first positioning unit; 40, second positioning unit. Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0020] It is understood that the specific embodiments and accompanying drawings described herein are merely for explaining the invention and are not intended to limit the invention.

[0021] It is understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0022] It is understood that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, while the parts unrelated to the present invention are not shown in the drawings.

[0023] It is understood that each unit or module involved in the embodiments of the present invention may correspond to only one entity structure, or may be composed of multiple entity structures, or multiple units or modules may be integrated into one entity structure.

[0024] It is understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of this invention may occur in a different order than that marked in the accompanying drawings.

[0025] It is understood that the flowcharts and block diagrams of this invention illustrate the possible architecture, functions, and operations of systems, apparatuses, devices, and methods according to various embodiments of this invention. Each block in the flowchart or block diagram may represent a unit, module, program segment, or code, containing executable instructions for implementing the specified function. Furthermore, each block or combination of blocks in the block diagram and flowchart can be implemented using a hardware-based system to achieve the specified function, or using a combination of hardware and computer instructions.

[0026] It is understood that the units and modules involved in the embodiments of the present invention can be implemented by software or by hardware. For example, the units and modules can be located in a processor.

[0027] Example 1: like Figure 1 As shown, this embodiment provides a method for locating network problems of complaining users based on big data analysis. Starting from three dimensions, namely point problem analysis, group user problem analysis, and pre-complaint analysis, it constructs a set of analysis methods from the perspectives of user packages, user terminals, network performance, etc., and solidifies the model to form a pre-warning module. This improves processing efficiency, provides early warning of complaints, reduces the scale of user complaints, improves users' perception of complaint handling satisfaction, and enhances the company's reputation.

[0028] This example consists of the following steps: Step 1: Connect to the network real-time data system using the Kafka tool, including complaint order data, signaling data, MRO data, B-domain data, such as core network signaling data, user-level MRO data, and user basic attribute data; Mobile network service records

[0029] Mobile network service record data records the service history of each time a mobile user browses web pages or engages in instant messaging, including the base station and cell information where the service occurred.

[0030] Basic engineering parameters business

[0031] Among them, eNodeB and CELL ID represent the identifiers of the currently serving base station and cell, as well as basic information such as the latitude and longitude of the base station and cell.

[0032] B-domain order data

[0033] Domain B order data records order data related to user business transactions, including payments, changes to packages, new packages, account cancellation, etc.

[0034] Kafka data acquisition tools

[0035] Kafka is a high-throughput distributed publish-subscribe messaging system characterized by its speed, scalability, and persistence. Now an open-source system under the Apache Software Foundation, it is widely used by various commercial companies as part of the Hadoop ecosystem. Its greatest strength lies in its ability to process large amounts of data in real time to meet diverse needs, such as Hadoop-based batch processing systems, low-latency real-time systems, and the Spark streaming engine.

[0036] Step 2: Problem localization and analysis for individual complaining users: First, analyze the cell coverage performance of the user's service trajectory, and then analyze the success rate of various signaling in the core network, such as attachment success rate, handover success rate, PDN request success rate, etc. Point-of-care (POC) user complaints refer to service complaints arising from user dissatisfaction due to poor service experience. Currently, in actual network operations, these mainly include slow internet speeds, inability to access the internet, inability to make voice calls, and one-way voice calls. The causes of these problems are complex. Some are due to poor wireless channel quality caused by wireless coverage issues, resulting in slow internet speeds or even complete inability to access the internet. Other complaints are due to mismatches between the user's subscription information in the O domain and B domain, typically occurring when users renew their data plans after running out of data. Still other complaints are even caused by problems with the user's own terminal. The core data for locating point-of-care complaints is often hidden in O domain data, especially signaling data from the wireless and core network sides. By establishing data models, the true root cause of the problem can often be discovered.

[0037] Analysis and investigation dimensions of point-based complaint issues

[0038] Step 3: Problem localization and analysis for group user complaints: Focusing on three aspects of analysis models: user packages, user terminals, and network performance; Step 4: Conduct problem localization analysis of complaining users from various dimensions, including package clustering analysis, terminal brand and model clustering analysis, and service performance clustering analysis; complaining user package distribution is as follows: Figure 4 As shown.

[0039] Taking data from a certain province as an example, this study attempts to discover the correlation between user complaints across different service packages by aggregating and analyzing complaints from both users who file complaints and those who file repeated complaints. Using user service packages as the observation dimension, we separately statistically analyzed the exposure of each package among complaining and repeat complaining users. The results show that some packages rank higher among repeat complaining users than among complaining users. If the time window is large enough, the analysis results will be more statistically reliable, and we can prioritize investigating these packages for potential logical conflicts or issues with package template configuration.

[0040] Taking data from a certain province as an example, this study attempts to discover the correlation between terminal brands, models, and complaints by aggregating and analyzing user complaints across the terminal and brand dimensions. We focus on the brand and model of 4G and 5G terminals as the starting point for our analysis, examining the correlation between 4G and 5G terminal brands and complaints respectively. The results show that, compared with the brands and models of terminals used by all users across the network, the ranking of Xiaomi, Apple, Honor, and Redmi brands among complainants changes slightly, but the trends for 4G and 5G terminal brands are not significantly different. If there are obvious clustering characteristics for terminal brands and models, we can start from the perspective of terminal capabilities to investigate whether the user's service usage is affected by problems with the terminal itself. A diagram illustrating the aggregation analysis of complaints and repeat complaints across user terminal brands is shown below. Figure 5 , Figure 6 , Figure 7 and Figure 8 shown, specifically, Figure 5 This is a diagram showing the distribution of 4G terminal brands among all network users. Figure 6 This is a diagram showing the distribution of 5G terminal brands among all network users. Figure 7 This is a diagram showing the distribution of 4G terminal brands among users who have filed complaints. Figure 8 This is a diagram showing the distribution of 5G terminal brands among users who have filed complaints.

[0041] Simultaneously, the correlation between 4G terminal models, key 5G models, and complaints will be analyzed to attempt to discover the relationship between terminal models and complaints. For example... Figure 9 As shown in the results, when comparing and analyzing the terminal models used by all users across the network, we should focus on terminal models whose slope line (a straight line from 0 to 30) is below the band line (since only the top 30 terminal models are considered for the entire network, the maximum value of the band line is 30). We can prioritize investigating these terminal models.

[0042] Aggregate analysis of single and repeat complaint users by user terminal model.

[0043] Regarding the analysis of service performance for complaining users, this methodology initially suggests focusing on aspects such as frequent switching between different systems, end-to-end latency, and service speed to understand the convergence relationship between complaints and services. Taking frequent switching between different systems as an example, if users frequently initiate switching processes between 4G and 5G, or between 4G and 3G, it can lead to short-term service interruptions. Such interruptions can manifest as perceived issues like game lag or live streaming stuttering. For example, based on data analysis from a certain province, some complaining users frequently switched between 4G and 5G within one hour. For such issues, a detailed investigation of wireless parameters, especially the switching parameter settings for different frequencies, should be prioritized.

[0044] Correlation analysis between complaining users and the number of times they switch between different systems

[0045] End-to-end connection establishment latency is a direct indicator of slow internet speeds for users. Therefore, in the network-industry collaboration analysis, a correlation analysis was first conducted between user service flow (APP) and service latency. Taking data from a certain province as an example, by comparing the service latency of complaining users with the average latency of the entire network, several APPs showed downlink latency multiples exceeding 3, with iQiyi video reaching a downlink latency multiple of 5.02. Further investigation can be conducted into issues related to internet service DNS and end-to-end data parameter configurations.

[0046] Correlation analysis between user complaints and business performance

[0047] This embodiment enables problem localization based on big data, thereby improving the efficiency and accuracy of network problem localization.

[0048] This embodiment can construct an analysis method from three dimensions: point-based problem analysis, group user problem analysis, and pre-complaint analysis. It considers user packages, user terminals, and network performance, and solidifies the model into a pre-warning module. This improves processing efficiency, provides early warnings of complaints, reduces the scale of user complaints, and overall enhances user satisfaction with complaint handling, thereby improving the company's reputation. Details are as follows: (1) For point-based complaints, the time, location and user information of the user complaint can be combined to quickly trace back the user's historical signaling plane and user plane data. The cause of the problem can be located by the success or failure of the signaling process, the end-to-end establishment delay of the business, the uplink and downlink speed of the business, and whether the BO domain contract information is consistent. (2) To locate the problem of a certain type of user complaint, such as a set of complaints about slow Internet speed within a month, we can correlate the complaint timestamp with the time window of O domain data and perform cluster analysis from dimensions such as package profile, terminal profile, business profile, and geographical profile. If there are obvious clustering features, we can further conduct detailed analysis on a certain type of feature, thereby achieving the ability to locate the problem from surface to line to point. (3) For certain problems that can be solidified, machine learning models and monitoring tools can be formed to monitor the signaling status of users in real time, discover potential experience problems before users complain, and investigate network problems through SMS care, network optimization and other means to improve users' business satisfaction and service reputation.

[0049] Example 2: like Figure 1 and Figure 2As shown, this embodiment provides a network problem localization method based on big data. This method is applied to the network side and includes the following steps: Step S1: Obtain the user's network problem data; Specifically, step S1 includes: By using the Kafka tool to connect to the real-time network data system, we can obtain complaint order data, signaling data, MRO data, and B-domain data to obtain network problem data.

[0050] Step S2: Based on the user's network problem data, determine whether the user's network problem is an isolated issue or a problem affecting a group of users. If the network problem is an individual user's network issue, the problem will be located based on the expert database model; if the network problem is a group of users' network issues, the problem will be located based on the expert database model. The expert database model is a network problem localization model pre-built based on big data. The expert database model carries historical network problems and historical network problem localization results.

[0051] Specifically, the process of locating network problems for individual users based on the expert database model includes the following steps: A1: Obtain the cell coverage performance of the user's service trajectory and determine whether there are cell coverage problems based on the expert database model: If a cell coverage issue exists, the initial location of the network problem for an individual user is obtained; if no cell coverage issue exists, the initial location of the network problem for an individual user is not obtained. A2: Obtain the core network signaling success rate, which includes the attachment success rate, handover success rate, and PDN request success rate; A3: Based on the expert database model, one or more core network signaling success rates that are lower than expected or abnormal are identified, thus obtaining a secondary location of the network problem for individual users; A4: By combining the first and second localizations of the individual user's network problem, the localization of the individual user's network problem is obtained.

[0052] Based on the expert database model, the localization of network problems for group users is carried out, specifically including: Based on the expert database model and user package clustering, network problems for specific user groups can be identified. And / or, Based on the expert database model and clustering by user terminal brand and model, network problems among group users are located. And / or, Based on the expert database model and user service performance clustering, network problems of group users are located.

[0053] In one specific implementation, step S0 is included before step S1. Step S0: Construct an expert database model for network problem localization, which includes the following steps: Step S01: Based on historical data, obtain training and test samples for network problem localization; Step S02: Label the training samples, including network problems and their corresponding localization. Step S03: Train the initial expert database model using labeled training samples to obtain the trained model. The initial expert database model is a selected machine learning or deep learning model. Step S04: Use test samples to evaluate and optimize the trained model, thereby constructing an expert database model for network problem localization.

[0054] Example 3:

[0055] like Figure 1 and Figure 3 As shown, this embodiment provides a network problem localization device based on big data. This device is applied to the network side and includes: Acquisition unit 10 is used to acquire network problem data of users; As one specific implementation, the acquisition unit 10 includes: The interface module is used to connect to real-time network data systems via the Kafka tool. The acquisition module, connected to the docking module, is used to obtain complaint order data, signaling data, MRO data, and B-domain data from the real-time network data system to obtain network problem data.

[0056] The judgment unit 20, connected to the acquisition unit 10, is used to determine whether the user's network problem is an individual user's network problem or a group of users' network problems based on the user's network problem data. The first positioning unit 30 is connected to the judgment unit 20 and is used to locate the network problem of the individual user based on the expert database model after the judgment unit 20 determines that the user's network problem is an individual user's network problem. As one specific implementation, the first positioning unit 30 includes: The first acquisition module is used to acquire the cell coverage performance of the user's service trajectory; The first judgment module, connected to the first acquisition module, is used to determine whether a cell coverage problem exists based on the expert database model. If a cell coverage issue exists, the initial location of the network problem for an individual user is obtained; if no cell coverage issue exists, the initial location of the network problem for an individual user is not obtained. The second acquisition module is used to acquire the core network signaling success rate, which includes the attachment success rate, handover success rate, and PDN request success rate. The second judgment module, connected to the second acquisition module, is used to determine, based on the expert database model, one or more core network signaling success rates that are lower than expected or abnormal, and to obtain the second location of network problems for individual users. The module is connected to the first judgment module and the second judgment module respectively, and is used to combine the first location of the network problem of an individual user and the second location of the network problem of an individual user to obtain the location of the network problem of an individual user.

[0057] The second positioning unit 40 is connected to the judgment unit 20 and is used to locate the network problem of the group users according to the expert database model after the judgment unit 20 determines that the user's network problem is a group user network problem. The expert database model is a network problem localization model pre-built based on big data. The expert database model carries historical network problems and historical network problem localization results.

[0058] In one specific implementation, the second positioning unit 40 includes a first positioning module, a second positioning module, and a third positioning module connected in parallel. The first positioning module is used to locate network problems of a group of users based on the expert database model and user package clustering. The second positioning module is used to locate network problems of a group of users based on the expert database model and clustering based on user terminal brand and model. The third positioning module is used to locate network problems for a group of users based on the expert database model and user service performance clustering.

[0059] Example 4: This invention provides a network problem handling method based on big data, the method comprising the following steps: According to the network problem localization method based on big data described in Example 2, the user's network problem is located, and the localization analysis result is obtained; Determine whether a user's network problem constitutes a network problem complaint: If a user's network problem is a reported network problem, then the complaint will be processed based on the location analysis results; if a user's network problem is not a reported network problem, then an early warning will be issued based on the location analysis results.

[0060] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A network problem localization method based on big data, applied to the network side, characterized in that, The method includes the following steps: Obtain user network problem data; Based on user network problem data, determine whether the user's network problem is an isolated issue or a problem affecting a group of users: If the problem is an individual user's network issue, the network problem will be located based on the expert database model. If it is a network problem involving a group of users, then the problem will be located based on the expert database model. The method of locating network problems for individual users based on the expert database model specifically includes: utilizing radio and core network signaling data hidden in the O domain data, first analyzing the cell coverage performance of the user's service trajectory, then analyzing the success rate of various core network signaling, and combining this with the consistency of subscription information in the B and O domains to locate the cause of the problem; the method of locating network problems for groups of users based on the expert database model specifically includes: associating the complaint timestamp with the time window of the O domain data, performing cluster analysis from the dimensions of package profile, terminal profile, service profile, and geographical profile to investigate problems caused by mismatch of subscription information between the O and B domains or terminal problems; The expert database model is a network problem localization model pre-built based on big data, and the expert database model carries historical network problems and historical network problem localization results.

2. The network problem localization method based on big data according to claim 1, characterized in that, Obtain user network problem data, specifically including: By using the Kafka tool to connect to the real-time network data system, we can obtain complaint order data, signaling data, MRO data, and B-domain data to obtain network problem data.

3. The network problem localization method based on big data according to claim 1, characterized in that, The process of locating network problems for individual users based on the expert database model includes the following steps: A1: Obtain the cell coverage performance of the user's service trajectory, and determine whether there are cell coverage problems based on the expert database model: If a cell coverage issue exists, the initial location of the network problem for an individual user is obtained; if no cell coverage issue exists, the initial location of the network problem for an individual user is not obtained. A2: Obtain the core network signaling success rate, which includes the attachment success rate, handover success rate, and PDN request success rate; A3: Based on the expert database model, one or more core network signaling success rates that are lower than expected or abnormal are identified, thus obtaining a secondary location of the network problem for individual users; A4: By combining the first and second localizations of the individual user's network problem, the localization of the individual user's network problem is obtained.

4. The network problem localization method based on big data according to claim 1, characterized in that, The method of locating network problems for group users based on the expert database model specifically includes: Based on the expert database model and user package clustering, network problems for specific user groups can be identified. And / or, Based on the expert database model and clustering by user terminal brand and model, network problems among group users are located. And / or, Based on the expert database model and user service performance clustering, network problems of group users are located.

5. The network problem localization method based on big data according to any one of claims 1 to 4, characterized in that, Before obtaining the user's network problem data, the method further includes step S0. Step S0: Construct an expert database model for network problem localization, which includes the following steps: Step S01: Based on historical data, obtain training and test samples for network problem localization; Step S02: Label the training samples, including network problems and their corresponding localization. Step S03: Train the initial expert database model using labeled training samples to obtain a training model, wherein the initial expert database model is a selected machine learning or deep learning model; Step S04: Use test samples to evaluate and optimize the trained model, thereby constructing an expert database model for network problem localization.

6. A network problem localization device based on big data, applied to the network side, characterized in that, The device includes: The acquisition unit is used to acquire network problem data from users. The judgment unit, connected to the acquisition unit, is used to determine whether the user's network problem is an individual user's network problem or a group of users' network problems based on the user's network problem data. The first positioning unit, connected to the judgment unit, is used to locate the network problem of an individual user based on the judgment unit's determination that the user's network problem is an individual user's network problem, and then locate the network problem of an individual user based on the expert database model. The location of the network problem of an individual user based on the expert database model specifically includes: using the radio and core network side signaling data hidden in the O domain data, first analyzing the cell coverage performance of the user's service trajectory, then analyzing the success rate of various core network signaling, and locating the cause of the problem by combining whether the O domain subscription information is consistent. The second positioning unit, connected to the judgment unit, is used to locate the network problem of the group of users based on the expert database model after the judgment unit determines that the user's network problem is a group of users' network problems. The location of the network problem of the group of users based on the expert database model specifically includes: clustering analysis based on the time window association between the complaint timestamp and the O domain data, from the dimensions of package profile, terminal profile, service profile and geographical profile, in order to investigate the problem caused by the mismatch of the contract information between the O domain and the B domain or the terminal problem. The expert database model is a network problem localization model pre-built based on big data, and the expert database model carries historical network problems and historical network problem localization results.

7. The network problem location device based on big data according to claim 6, characterized in that, The acquisition unit includes: The interface module is used to interface with a real-time network data system via the Kafka tool; The acquisition module, connected to the docking module, is used to acquire complaint order data, signaling data, MRO data, and B-domain data from the real-time network data system to obtain network problem data.

8. The network problem location device based on big data according to claim 6, characterized in that, The first positioning unit includes: The first acquisition module is used to acquire the cell coverage performance of the user's service trajectory; The first judgment module, connected to the first acquisition module, is used to determine whether a cell coverage problem exists based on the expert database model. If a cell coverage issue exists, the initial location of the network problem for an individual user is obtained; if no cell coverage issue exists, the initial location of the network problem for an individual user is not obtained. The second acquisition module is used to acquire the core network signaling success rate, which includes the attachment success rate, the handover success rate, and the PDN request success rate. The second judgment module, connected to the second acquisition module, is used to determine, based on the expert database model, one or more core network signaling success rates that are lower than expected or abnormal, and to obtain the second location of network problems for individual users. The module is connected to the first judgment module and the second judgment module respectively, and is used to combine the first location of the network problem of an individual user with the second location of the network problem of an individual user to obtain the location of the network problem of an individual user.

9. The network problem location device based on big data according to any one of claims 6 to 8, characterized in that, The second positioning unit includes a first positioning module, a second positioning module, and a third positioning module connected in parallel. The first positioning module is used to locate network problems of a group of users based on the expert database model and user package clustering. The second positioning module is used to locate network problems of a group of users based on the expert database model and clustering based on user terminal brand and model. The third positioning module is used to locate network problems of a group of users based on the expert database model and user service performance clustering.

10. A network problem-solving method based on big data, characterized in that, The method includes the following steps: According to any one of claims 1 to 5, the network problem localization method based on big data is used to locate the user's network problem and obtain the localization analysis result; Determine whether a user's network problem constitutes a network problem complaint: If a user's network problem is a reported network problem, then the complaint will be processed based on the location analysis results; if a user's network problem is not a reported network problem, then an early warning will be issued based on the location analysis results.

Images