Wireless network fault processing method and device, computer equipment and storage medium

Abstract

The invention relates to a wireless network fault processing method and device, computer equipment and a computer readable storage medium. The method comprises the following steps: simulating a pre-constructed digital twinborn model for a target wireless network to obtain simulation operation data of the target wireless network; inputting the simulation operation data into a plurality of types of pre-constructed wireless network fault prediction models, and obtaining a predicted fault type of the target wireless network through the plurality of types of wireless network fault prediction models; matching the predicted fault type by using a knowledge graph for the wireless network fault and a text processing model to obtain fault analysis information of the target wireless network; the fault analysis information is used for performing fault processing on the target wireless network. By adopting the method, the accuracy degree of fault type prediction of the target wireless network can be improved through multiple types of wireless network fault prediction models, so that the reliability degree of wireless network fault prediction analysis can be improved.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for handling wireless network faults. Background Art

[0002] With the development of communication technology, wireless network monitoring and troubleshooting are becoming more and more important. As the scale of wireless networks continues to expand and application scenarios become more complex, traditional wireless network fault handling technologies, such as manual inspections and network equipment log analysis, usually have problems such as low efficiency and difficulty in finding faults, and are unable to meet the needs of wireless network fault detection.

[0003] In response to the above situation, a technology has emerged that combines digital twin technology, big model technology and knowledge graph technology to predict wireless network faults. This technology can build a virtual model of the wireless network, simulate the actual state of the wireless network and synchronously output the network status data. It can then combine big model technology and knowledge graph technology to perform fault prediction analysis on the network status data.

[0004] However, the technology of combining digital twin technology, big model technology and knowledge graph technology to perform wireless network fault prediction analysis still has the problem of low reliability of prediction analysis. Summary of the invention

[0005] Based on this, it is necessary to provide a wireless network fault handling method, apparatus, computer equipment, computer-readable storage medium and computer program product that can improve the reliability of predictive analysis in response to the above technical problems.

[0006] In a first aspect, the present application provides a wireless network fault handling method, comprising:

[0007] Simulating a digital twin model pre-built for a target wireless network to obtain simulated operation data of the target wireless network;

[0008] Inputting the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models;

[0009] The predicted fault type is matched by using a knowledge graph and a text processing model for wireless network faults to obtain fault analysis information of the target wireless network; the fault analysis information is used to perform fault processing on the target wireless network.

[0010] In one embodiment, the multiple types of wireless network fault prediction models include: at least two of: a support vector machine model, a neural network model, and a decision tree model; the inputting of the simulation operation data into the pre-constructed multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models, includes: inputting the simulation operation data into the support vector machine model, and obtaining fault characterization classification data corresponding to the simulation operation data through the support vector machine model; inputting the fault characterization classification data into the neural network model, and obtaining the probability of the target wireless network corresponding to different fault types through the neural network model; inputting the probability into the decision tree model, and obtaining the predicted fault type of the target wireless network from each fault type through the decision tree model.

[0011] In one of the embodiments, the wireless network fault handling method also includes: obtaining actual operation data of the target wireless network; performing data preprocessing on the actual operation data to obtain processed actual operation data; mapping the processed actual operation data to the digital twin model, adjusting the model parameters of the digital twin model according to the mapping results, and obtaining an updated digital twin model of the target wireless network.

[0012] In one of the embodiments, the digital twin model includes: a network configuration parameter model; before simulating the digital twin model pre-built for the target wireless network, it also includes: collecting network parameters of the target wireless network according to actual network configuration information of the target wireless network, and obtaining digital twin usage information of the target wireless network; obtaining target network parameters matching the digital twin usage information from the network parameters; and constructing the network configuration parameter model using the target network parameters.

[0013] In one of the embodiments, the digital twin model includes: a user behavior model; before simulating the digital twin model pre-built for the target wireless network, it also includes: obtaining user behavior characteristics of the scenario where the target wireless network is located, and the life cycle corresponding to the target wireless network; obtaining target user behavior characteristics matching the life cycle from the user behavior characteristics; and using the target user behavior characteristics to construct the user behavior model.

[0014] In one of the embodiments, the knowledge graph stores fault information of multiple historical faults; the fault information includes fault type information, fault cause information and fault handling method information of the historical faults; the method also includes: displaying the fault information of the historical faults; using the knowledge graph and text processing model for wireless network faults to match the predicted fault type to obtain fault analysis information of the target wireless network, including: matching the predicted fault type with each fault information stored in the knowledge graph to obtain target fault information whose fault type information is the predicted fault type; based on the fault cause information and fault handling method information corresponding to the target fault information, generating the fault analysis information of the target wireless network through the text processing model.

[0015] In one of the embodiments, after matching the predicted fault type using a pre-built knowledge graph and text processing model of wireless network faults to obtain fault analysis information of the target wireless network, the method further includes: generating fault warning information of the target wireless network based on the fault analysis information; the fault warning information includes the predicted fault type, fault cause information and fault handling method information of the target wireless network; sending the fault warning information to a fault handling end for display; and the fault handling end is used to perform fault handling on the target wireless network based on the predicted fault type, fault cause information and fault handling method information contained in the displayed fault warning information.

[0016] In one of the embodiments, the wireless network fault handling method also includes: based on the simulation operation data, obtaining the spatial distribution information and time change trend information of the digital twin model corresponding to multiple network performance indicators; displaying the spatial distribution information through a three-dimensional heat map, and displaying the time change trend information through a time series graph.

[0017] In a second aspect, the present application further provides a wireless network fault processing device, including:

[0018] A simulation data acquisition module, used to simulate a digital twin model pre-built for a target wireless network to obtain simulated operation data of the target wireless network;

[0019] A fault type prediction module, used for inputting the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models;

[0020] A fault information analysis module is used to match the predicted fault type using a knowledge graph and a text processing model for wireless network faults to obtain fault analysis information of the target wireless network; the fault analysis information is used to perform fault processing on the target wireless network.

[0021] In a third aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described in any one of the embodiments of the first aspect when executing the computer program.

[0022] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in any one of the embodiments of the first aspect.

[0023] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the method described in any one of the embodiments of the first aspect.

[0024] The above-mentioned wireless network fault handling method, device, computer equipment, computer-readable storage medium and computer program product obtain simulated operation data of the target wireless network by simulating the digital twin model pre-built for the target wireless network; input the simulated operation data into the pre-built multiple types of wireless network fault prediction models, and obtain the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models; use the knowledge graph and text processing model for wireless network faults to match the predicted fault type to obtain the fault analysis information of the target wireless network; the fault analysis information is used to perform fault handling on the target wireless network. In this embodiment, the simulated operation data can be obtained through the digital twin model of the target wireless network, and then the simulated operation data can be input into the pre-built multiple types of wireless network fault prediction models. After obtaining the predicted fault type, the knowledge graph and text processing model for wireless network faults can be used to obtain the fault analysis information for fault handling. The multiple types of wireless network fault prediction models can improve the accuracy of the predicted fault type of the target wireless network, thereby improving the reliability of the wireless network fault prediction analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0026] FIG1 is a schematic diagram of a flow chart of a method for handling wireless network failures in one embodiment;

[0027] FIG2 is a schematic diagram of a process of obtaining a predicted fault type of a target wireless network in one embodiment;

[0028] FIG3 is a schematic diagram of a process for updating a digital twin model in one embodiment;

[0029] FIG4 is a schematic diagram of a module interface of a wireless network real-time monitoring and fault prediction system in one embodiment;

[0030] FIG5 is a schematic flow chart of a method for real-time monitoring and fault prediction of a wireless network in one embodiment;

[0031] FIG6 is a structural block diagram of a wireless network fault processing device in one embodiment;

[0032] FIG. 7 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0034] In one embodiment, as shown in FIG1 , a method for handling wireless network faults is provided. This embodiment uses the method applied to a terminal as an example for illustration. It is understandable that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0035] Step S101, simulating a digital twin model pre-built for a target wireless network to obtain simulated operation data of the target wireless network.

[0036] The digital twin model refers to a pre-built twin model used to simulate the target wireless network, which may include: a three-dimensional scene model, which may include information such as building structure, geographic location, and room layout; a device model, which may include information such as device model, parameter configuration, and status information; a network parameter model, which includes information such as network standards, frequency bands, channel allocation, and security policies; and a user behavior model, such as the user's network usage intensity, common application types, and device types in different time periods. The target wireless network refers to the wireless network that requires fault prediction analysis, and the simulated operation data is the simulated operation data for the target wireless network obtained by simulating the digital twin model.

[0037] Specifically, the user can pre-lay out a digital twin model for the target wireless network in the terminal, and then simulate the digital twin model to obtain simulated operation data for the target wireless network through the digital twin model.

[0038] Step S102: inputting the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models.

[0039] A wireless network fault prediction model is a model used to predict the type of wireless network faults, and multiple types of wireless network fault prediction models represent that there are multiple types of the above models, for example, it can include support vector machine models, neural network models, decision tree models and other models that can be used to identify fault types, and the predicted fault type is the fault category of the target wireless network predicted by the above multiple models.

[0040] Specifically, after obtaining the simulated operation data of the target wireless network, the terminal can input the simulated operation data into various types of wireless network fault prediction models, and then output the predicted fault type of the target wireless network through the various types of wireless network fault prediction models.

[0041] Step S103, using the knowledge graph and text processing model for wireless network failures, matching the predicted failure type to obtain failure analysis information of the target wireless network; the failure analysis information is used to perform failure processing on the target wireless network.

[0042] The knowledge graph for wireless network failures refers to a pre-built knowledge graph for handling wireless network failures, and the text processing model is a model for generating and understanding natural language, such as a large language model, and the fault analysis information is information obtained after analyzing the faults of the target wireless network, which can be used to complete the fault handling of the wireless network.

[0043] Specifically, after the terminal identifies the fault type of the wireless network, it can use the text processing model and the knowledge graph of the wireless network fault to match the fault analysis information corresponding to the predicted fault type as the fault analysis information of the target wireless network. The fault analysis information can then be used to complete the fault handling of the target wireless network.

[0044] In the above wireless network fault handling method, the simulated operation data of the target wireless network is obtained by simulating the digital twin model pre-built for the target wireless network; the simulated operation data is input into the pre-built multiple types of wireless network fault prediction models, and the predicted fault type of the target wireless network is obtained through the multiple types of wireless network fault prediction models; the predicted fault type is matched by using the knowledge graph and text processing model for wireless network faults to obtain the fault analysis information of the target wireless network; the fault analysis information is used to perform fault handling on the target wireless network. In this embodiment, the simulated operation data can be obtained by the digital twin model of the target wireless network, and then the simulated operation data can be input into the pre-built multiple types of wireless network fault prediction models. After obtaining the predicted fault type, the knowledge graph and text processing model for wireless network faults can be used to obtain the fault analysis information for fault handling. The multiple types of wireless network fault prediction models can improve the accuracy of the predicted fault type of the target wireless network, thereby improving the reliability of the wireless network fault prediction analysis.

[0045] In one embodiment, the multiple types of wireless network fault prediction models include: at least two of a support vector machine model, a neural network model, and a decision tree model; as shown in FIG2 , step S102 may further include:

[0046] Step S201 : inputting simulation operation data into a support vector machine model, and obtaining fault characterization classification data corresponding to the simulation operation data through the support vector machine model.

[0047] The support vector machine model can strengthen the data used to determine fault characterization, that is, it can be used to strengthen the classification accuracy of fault characterization data, such as alarms, number of access users, average uplink / downlink transmission rate, number of reconnections, wireless resource occupancy rate, used transmission bandwidth, etc. Therefore, the terminal can input the simulation operation data into the support vector machine model, and implement data classification through the support vector machine model to obtain fault characterization classification data. Compared with the simulation operation data, the fault characterization classification data can enhance the accuracy of data classification.

[0048] Step S202: inputting the fault characterization classification data into a neural network model, and obtaining the probability of the target wireless network corresponding to different fault types through the neural network model.

[0049] The neural network model can perform small-grained accurate predictions based on the fault characterization classification refined by the support vector machine. Specifically, after the terminal obtains the fault characterization classification data through the support vector machine model, it can also input the fault characterization classification data into the neural network model. The neural network model can output the probability of the target wireless network corresponding to different fault types.

[0050] Step S203: input the probability into a decision tree model, and obtain the predicted fault type of the target wireless network from each fault type through the decision tree model.

[0051] The decision tree model focuses on different scenarios in which wireless networks are located, such as shopping malls, universities, office buildings, etc. The terminal obtains the prediction results of each score output by the neural network and combines them with the scenario requirements to generate the decision output with the highest feasibility probability, thereby obtaining the final predicted fault type of the target wireless network from a variety of fault types.

[0052] In this embodiment, the predicted fault type can be obtained by integrating multiple wireless network fault prediction models, such as combining support vector machines, neural networks, decision trees and other models. Since the support vector machine can enhance the accuracy of data classification used to determine fault characterization, the neural network performs small-grained accurate prediction based on the fault characterization classification refined by the support vector machine, and the decision tree can process the prediction results of each score output by the neural network and generate the decision output with the highest feasibility probability in combination with the scenario requirements. Through the combination of the above-mentioned multiple models, the limitation of insufficient capacity of a single model is broken, and the ability of high-precision intelligent output is achieved, thereby improving the accuracy and reliability of fault prediction and diagnosis.

[0053] In one embodiment, as shown in FIG3 , the wireless network fault handling method may further include:

[0054] Step S301: Acquire actual operation data of the target wireless network.

[0055] The actual operation data refers to the operation data generated during the actual operation of the target wireless network, which can be collected by sensors and other means. For example, the actual operation data can include environmental information such as temperature, humidity, light intensity, etc., and network equipment operation data such as signal strength, delay, packet loss rate, load, equipment status, etc., and user behavior data such as user location, network usage time, application type, device type, traffic, etc.

[0056] Specifically, after the target wireless network is put into use, the actual operation information of the target wireless network can be collected through sensors and other means. For example, environmental information can be collected through sensors, network protocol analysis can be used to obtain operation data of network devices, and network analysis can be used to obtain user behavior data. Through the above methods, the actual operation data of the target wireless network can be collected.

[0057] Step S302, performing data preprocessing on the actual operation data to obtain processed actual operation data;

[0058] Step S303: Map the processed actual operation data to the digital twin model, adjust the model parameters of the digital twin model according to the mapping results, and obtain the updated digital twin model of the target wireless network.

[0059] After obtaining the actual operation data, in order to ensure data quality, the collected actual operation data can be cleaned and preprocessed first, for example, the collected actual operation data can be removed from abnormal data and data conversion can be performed to ensure data quality. After obtaining the processed actual operation data, the actual operation data can be mapped to the digital twin model to adjust the model parameters of the digital twin model to achieve the update of the digital twin model. Among them, the process of mapping the preprocessed data to the digital twin model and updating the state of the model can be to map the collected signal strength data to the signal strength information of the corresponding area, and to map the device status data to the status information of the corresponding device. The process of updating the model can be to dynamically update the digital twin model according to the results of data mapping, for example, to add new devices, modify device parameters, update user behavior data, etc., to ensure the real-time and accuracy of the model.

[0060] In this embodiment, the actual operation data of the target wireless network can be obtained, and after preprocessing the actual operation data, the digital twin model can be updated using the preprocessed actual operation data. In this way, the real-time and accuracy of the digital twin model's simulation operation of the target wireless network can be further improved.

[0061] In one embodiment, the digital twin model includes: a network configuration parameter model; before step S101, it may also include: collecting network parameters of the target wireless network according to the actual network configuration information of the target wireless network, and obtaining digital twin usage information of the target wireless network; obtaining target network parameters matching the digital twin usage information from the network parameters; and constructing a network configuration parameter model using the target network parameters.

[0062] The actual network configuration information refers to the network configuration information of the target wireless network, while the network parameters refer to the parameters related to the target wireless network, such as wireless network standards, frequency bands, channel allocations, security policies, etc., which can be collected based on the actual network configuration information. The digital twin usage information characterizes the network usage that the digital twin model needs to simulate, and the target network parameters refer to the network parameters used to build the network configuration parameter model, which can be determined based on the network usage that the digital twin model needs to simulate.

[0063] For example, network parameters may include parameters directly related to network coverage, such as signal strength, transmit power, and antenna azimuth, and may also include parameters such as the number of connected users and user service-related attributes. If the target wireless network simulated by the current digital twin network is a network usage in the planning stage, since the network planning stage focuses on the network's coverage capability and the network capacity only needs to be estimated theoretically, parameters directly related to network coverage, such as signal strength, transmit power, and antenna azimuth, can be used as target network parameters, and the above parameters can be used to construct a network configuration parameter model. However, network parameters such as the number of connected users and user service-related attributes are not used to construct the network configuration parameter model.

[0064] Specifically, when constructing a network configuration parameter model, you can first collect the network parameters of the target wireless network, and then establish a corresponding digital twin model switch for each network parameter. Then, you can determine the target network parameters based on the digital twin usage information of the target wireless network, and set the digital twin model switch corresponding to the target network parameter to on, and set the digital twin model switches of other network parameters except the target network parameter to off, so that the network configuration parameter model can be constructed through the target network parameters.

[0065] In this embodiment, the network configuration parameter model can be constructed based on the target network parameters, and the target network parameters are determined according to the digital twin usage information of the target wireless network. In this way, unnecessary digital twin network parameters can be reduced to construct the network configuration parameter model while ensuring that the network configuration parameter model can meet the requirements of the system application, thereby improving the construction efficiency of the network configuration parameter model.

[0066] In one embodiment, the digital twin model includes: a user behavior model; before step S101, it may also include: obtaining user behavior characteristics of the scene where the target wireless network is located, and the life cycle corresponding to the target wireless network; obtaining target user behavior characteristics that match the life cycle from the user behavior characteristics; and using the target user behavior characteristics to construct a user behavior model.

[0067] The user behavior model is a model used to simulate the user behavior of the target wireless network. The model can be constructed based on the historical data of the scene where the target wireless network is located or the preset user behavior characteristics, such as the network usage intensity, common application types, and device types of users in different time periods. The life cycle corresponding to the target wireless network is the life cycle of the target wireless network that needs to be simulated by the digital twin model. The target user behavior characteristics refer to the user behavior characteristics used to construct the user behavior model. The user behavior characteristics can be determined based on the life cycle of the target wireless network to be simulated.

[0068] Similar to network parameters, for the network maintenance and operation required for different life cycles of the network, the full set of user behaviors is not an option that must be used for every problem to be investigated. Therefore, when constructing a user behavior model, you can also select appropriate user behavior features, that is, select target user behavior features from user behavior features to realize the construction of the user behavior model.

[0069] Specifically, when constructing a user behavior model, you can first collect user behavior characteristics of the scenario where the target wireless network is located. For example, if the scenario where the target wireless network is located is a school, you can collect user behavior characteristics of the school scenario. Then you can establish a corresponding digital twin model switch for each user behavior characteristic. You can also determine the target user behavior characteristics based on the life cycle of the target wireless network, and set the digital twin model switch corresponding to the target user behavior characteristics to on, and set the digital twin model switches of other user behavior characteristics except the target user behavior characteristics to off, so as to construct a user behavior model through the target user behavior characteristics.

[0070] In this embodiment, the user behavior model can be constructed based on the target user behavior characteristics, and the target user behavior characteristics are determined according to the life cycle of the target wireless network. In this way, when constructing the user behavior model, it is possible to avoid always generating a full set model of all attributes, thereby reducing the model complexity and improving the performance of the digital twin network.

[0071] In one embodiment, the knowledge graph stores fault information of multiple historical faults; the fault information includes fault type information, fault cause information and fault handling method information of the historical faults; the wireless network fault handling method may also include: displaying the fault information of the historical faults; step S103 may further include: matching the predicted fault type with each fault information stored in the knowledge graph to obtain target fault information whose fault type information is the predicted fault type; based on the fault cause information and fault handling method information corresponding to the target fault information, generating fault analysis information of the target wireless network through a text processing model.

[0072] Historical faults refer to wireless network faults that have occurred in the past, and the fault information of historical faults refers to the fault-related information of historical faults, which may include the type of fault, i.e., fault type information, the cause of the fault, i.e., fault cause information, and the method of handling the fault, i.e., fault handling method information. The above fault information may be stored in the knowledge graph for wireless network faults. In addition, the above knowledge graph may also be visualized to facilitate users to understand information such as fault type, symptoms, and solutions. The target fault information is the fault information stored in the knowledge graph, and the fault type information is the fault information of the predicted fault type.

[0073] Specifically, the fault data of historical faults can be used to construct a knowledge graph of wireless network faults. The knowledge graph can include the fault type, fault cause, and fault handling method of historical faults, etc., and can be displayed in a visual way. After that, if the wireless network fault prediction model identifies the predicted fault type of the target wireless network, the predicted fault type can be matched with the fault information stored in the knowledge graph to obtain the target fault information whose fault type is the predicted fault type. After that, a text processing model, such as a large language model, can be used to combine the fault cause information and the fault handling method information to generate the fault analysis information of the target wireless network, such as generating a text containing the fault cause of the target wireless network and the fault handling method for the target wireless network as the fault analysis information of the target wireless network.

[0074] In this embodiment, the knowledge graph of wireless network faults can store fault information of various historical faults. After obtaining the predicted fault type through the wireless network fault prediction model, the predicted fault type can be matched with the fault information of the historical faults in the knowledge graph. After obtaining the target fault information, a text processing model, i.e., a large language model, can be used to generate fault analysis information in combination with the fault cause information and fault handling method information corresponding to the target fault information. In this way, the target wireless network fault can be handled more accurately and effectively.

[0075] In one embodiment, after step S103, the following may also be included: generating fault warning information of the target wireless network based on the fault analysis information; the fault warning information includes the predicted fault type, fault cause information and fault handling method information of the target wireless network; sending the fault warning information to the fault handling end for display; the fault handling end is used to perform fault handling on the target wireless network according to the predicted fault type, fault cause information and fault handling method information contained in the displayed fault warning information.

[0076] Fault warning information is used to send to the fault processing end to remind the fault processing end to handle the network fault of the target wireless network. The warning information includes the fault type of the target wireless network, that is, the predicted fault type, the cause of the fault, that is, the fault cause information, and the handling method for the fault, that is, the fault handling method information, etc. Specifically, after the terminal obtains the fault analysis information, it can generate fault warning information based on the fault analysis information. The fault warning information may include the fault type, possible causes, recommended solutions, etc., and send it to the fault processing end used to handle the wireless network fault. The fault processing end can be the terminal itself or the operation and maintenance service end that remotely handles the fault.

[0077] Specifically, after the terminal obtains the fault analysis information, it can also generate fault warning information for the target wireless network. The fault warning information includes the predicted fault type, fault cause information, and fault handling method information of the target wireless network. The fault warning information can then be sent to the fault handling end, which handles the fault of the target wireless network based on the above warning information. In this way, the efficiency of fault handling can be improved.

[0078] In this embodiment, fault analysis information can also be used to generate fault warning information of the target wireless network, and the fault warning information can be sent to the fault processing end, so that the fault processing end can perform fault processing according to the fault warning information. In this way, the efficiency of fault processing can be improved.

[0079] In one embodiment, the wireless network fault handling method may also include: based on simulation operation data, obtaining the spatial distribution information and time change trend information of the digital twin model corresponding to multiple network performance indicators; displaying the spatial distribution information through a three-dimensional heat map, and displaying the time change trend information through a time series graph.

[0080] A variety of network performance indicators refer to a variety of indicators used to characterize network performance, including signal strength, delay, jitter and other indicators. In this embodiment, after the terminal completes the collection of simulation operation data, it can also obtain the spatial distribution information and time change trend information of various network performance indicators and display them. The spatial distribution information can be displayed in the form of a three-dimensional heat map, that is, the spatial distribution of indicators such as signal strength, delay, and jitter in the digital twin model is displayed using a three-dimensional heat map to help users intuitively understand network performance, while the time change trend information can be displayed through a time series graph, that is, the time series graph is used to display the change trend of network operation data over time, such as signal strength, load conditions, number of users, etc., to facilitate users to analyze network performance changes and fault trends.

[0081] In this embodiment, the spatial distribution information and time change trend information of various network performance indicators can also be displayed through three-dimensional heat maps and time series graphs. This method can help users intuitively understand network performance and facilitate users to analyze network performance changes and fault trends, thereby further improving the intelligence of wireless network fault handling.

[0082] In one embodiment, a method for real-time monitoring and fault prediction of wireless networks based on digital twins is also provided. Through multi-source data fusion, deep learning, knowledge graphs and other technologies, a more comprehensive and intelligent digital twin model is constructed, which can more accurately predict wireless network faults and provide more effective solutions to improve user experience, improve network efficiency, and reduce operation and maintenance costs. This embodiment organically combines the steps of digital twin model construction, real-time data collection, multi-model training, knowledge graph construction, fault diagnosis, fault prediction, warning information generation, data visualization, etc., to provide users with more intelligent wireless network monitoring and fault prediction services, wherein the modules and interfaces of the wireless network real-time monitoring and fault prediction system can be shown in Figure 4.

[0083] As shown in FIG5 , the implementation of the solution specifically includes the following steps:

[0084] 1. Digital twin model construction:

[0085] Model and required information, including but not limited to:

[0086] Build a 3D scene model, including information such as building structure, geographic location, and room layout.

[0087] Build a device model based on the actual network equipment, including device model, parameter configuration, status information, etc.

[0088] Build a network parameter model, including network standards, frequency bands, channel allocation, security policies, etc.

[0089] Build a user behavior model based on historical data or preset user behavior characteristics.

[0090] Further steps are:

[0091] 1.1 Scene model construction:

[0092] Users upload interior floor plans, supporting common formats such as JPG, PNG, and PDF.

[0093] Using image recognition technology, we can automatically identify walls, doors, windows, furniture and other elements in the floor plan, and extract their size, material and other information.

[0094] A fast 3D scene modeling method based on a plan view uses a 3D modeling engine to automatically generate a 3D scene model. Walls, doors, windows, furniture and other elements are presented in 3D form and given corresponding material properties for subsequent signal propagation simulation.

[0095] 1.2 Equipment model construction:

[0096] Collect the type and configuration information of actual network equipment, such as router model, transmission power, antenna pattern, channel information, etc.; base station equipment model, coverage range, load conditions, equipment status, etc.

[0097] Build a network device model based on the device information. For example, create a router model that includes parameters such as signal transmission power, antenna pattern, channel information, etc.; create a base station device model that includes information such as coverage range, load conditions, and device status, and establish a corresponding digital twin model switch for each model parameter.

[0098] 1.3 Network parameter construction:

[0099] According to the actual network configuration, collect network parameter information, such as wireless network standards, frequency bands, channel allocation, security policies, etc.

[0100] Based on the collected network parameter information, a network parameter model is constructed, and a corresponding digital twin model switch is established for each model parameter. The switch of the model parameter is used to support the digital twin model to dynamically determine the network capabilities required for the instantiated digital twin network according to the purpose of the network digital twin. For example, in the network planning stage, the focus is on the network coverage capability, while the network capacity only needs to be estimated theoretically. Therefore, when the system recognizes that the current digital twin network is about to simulate the network purpose of a planning stage, the digital twin switch corresponding to the parameters directly related to network coverage, such as the received signal strength, transmission power, and antenna azimuth, is set to on, and the digital twin switch corresponding to the number of connected users, user service-related attributes, and other parameters should be set to off. This achieves the instantiation of the digital twin network required to meet the system application and saves unnecessary instantiation of digital twin network parameters.

[0101] 1.4 User behavior construction:

[0102] Build a user behavior model based on historical data or preset user behavior characteristics, such as the user's network usage intensity, common application types, device types, etc. in different time periods.

[0103] The construction of the user behavior model involves the collection and analysis of user behavior data, which can be designed and implemented according to the specific situation. Add new parameters to the digital twin model corresponding to the user behavior, and establish a corresponding digital twin model switch for each model parameter. The digital twin model switch of the user behavior will determine whether the user behavior reflects this user behavior feature in the instantiated digital twin network instance. For the network operation and maintenance and operation operations required for different life cycles of the network, the full set of user behaviors is not an option that must be used for every problem to be investigated. By configuring a control switch for each user behavior in the model, the full set model of all attributes can be avoided when the digital twin network is instantiated, thereby reducing the model complexity and improving the performance of the digital twin network.

[0104] 2. Real-time data collection:

[0105] The analysis objects include but are not limited to:

[0106] Environmental information such as temperature, humidity, light intensity, etc. is collected through sensors.

[0107] Through network protocol analysis, the operating data of network devices can be obtained, such as signal strength, delay, packet loss rate, load conditions, device status, etc.

[0108] Through network analysis, user behavior data is obtained, such as user location, network usage time, application type, device type, traffic, etc.

[0109] The expansion steps are:

[0110] 2.1 Sensor data collection:

[0111] Deploy sensors, such as temperature sensors, humidity sensors, and light sensors, to collect environmental information.

[0112] Sensor data collection involves sensor deployment and the selection of data transmission protocols, which need to be designed and implemented according to specific circumstances.

[0113] 2.2 Network equipment data collection:

[0114] Through network protocol analysis, the operating data of network devices can be obtained, such as signal strength, delay, packet loss rate, load conditions, device status, etc.

[0115] Use SNMP, NetFlow and other protocols for data collection.

[0116] 2.3 User behavior data collection:

[0117] Through network analysis, user behavior data is obtained, such as user location, network usage time, application type, device type, traffic, etc.

[0118] Data collection can be performed using technologies such as WiFi / Bluetooth / AOA positioning and data packet analysis.

[0119] 3. Data mapping and synchronization:

[0120] Clean and preprocess the collected data to ensure data quality.

[0121] Map the preprocessed data to the digital twin model and update the state of the model. For example, map the collected signal strength data to the signal strength information of the corresponding area, and map the device status data to the status information of the corresponding device.

[0122] Using deep learning models, we analyze real-time data in digital twin models and dynamically adjust model parameters. For example, we can dynamically adjust the coverage of base station equipment based on changes in user locations, or dynamically adjust channel allocation based on equipment load conditions.

[0123] The specific steps are as follows:

[0124] 3.1 Data cleaning and preprocessing:

[0125] Clean and preprocess the collected data, for example, remove abnormal data, perform data conversion, etc., to ensure data quality.

[0126] 3.2 Data Mapping:

[0127] Map the preprocessed data to the digital twin model and update the model status. For example, map the collected signal strength data to the signal strength information of the corresponding area, and map the device status data to the status information of the corresponding device.

[0128] 3.3 Model update:

[0129] Based on the results of data mapping, the digital twin model is dynamically updated, for example, adding new devices, modifying device parameters, updating user behavior data, etc., to ensure the real-time and accuracy of the model.

[0130] This step reflects the dynamic update capability of the digital twin model, which is also one of the important innovations of the present invention.

[0131] 4. Multi-model fusion training and reasoning:

[0132] Combining the advantages of multiple machine learning models, for example, combining support vector machines, neural networks, decision trees and other models, support vector machines are used to enhance the classification accuracy of data used to determine fault characterization (including but not limited to alarms, number of access users, average uplink / downlink transmission rates, number of reconnections, wireless resource occupancy rate, and used transmission bandwidth). The neural network performs small-grained accurate predictions based on the fault characterization classification refined by the support vector machine (based on the mathematical space classification basis obtained by the support vector machine, each classification is predicted separately; the classification results of the support vector machine include the degree of time correlation and the influence of spatial topological connection relationships from the perspective of physical meaning analysis). The decision tree processes the prediction results of each score output by the neural network and generates the decision output with the highest feasibility probability in combination with scenario requirements (including but not limited to specific scenarios for wireless network deployment, such as shopping malls, universities, office buildings, etc.). Finally, the big language model (big model for short) is good at processing semantic understanding and translation, but lacks professional knowledge and processing capabilities. The big model is used for demand analysis and intent recognition translation, and is passed to the support vector machine. After the support vector machine, neural network, and decision tree are executed in series, the decision output is fed back to the big model. The big model integrates the results with the requirements in text to form a complete natural language style solution and feeds back to the network operation and maintenance personnel.

[0133] It breaks through the limitation of insufficient capabilities of a single model and realizes the capability of high-precision intelligent output, thereby improving the accuracy and reliability of fault prediction and diagnosis.

[0134] The specific steps are as follows:

[0135] 4.1 Data preprocessing:

[0136] Preprocess historical network operation data and fault data, such as time series conversion and feature engineering, to make the data more suitable for model training.

[0137] 4.2 Model training:

[0138] Based on the data set, support vector machine, neural network, decision tree and other models are trained in turn.

[0139] The support vector machine is used to enhance the accuracy of data classification. The neural network performs small-grained accurate predictions based on the refined classification of the support vector machine. The decision tree processes the prediction results of each score output by the neural network and generates the decision output with the highest feasibility probability based on scenario requirements.

[0140] The data classification objects implemented by the support vector machine include, but are not limited to, alarms, number of access users, average uplink / downlink transmission rate, number of reconnections, wireless resource occupancy rate, and used transmission bandwidth.

[0141] The basis of neural network prediction includes, but is not limited to, mathematical space classification obtained by support vector machine. The classification results of support vector machine include the influence of time correlation and spatial topological connection relationship from the perspective of physical meaning.

[0142] The decision tree focuses on different scenarios where wireless networks are located, including but not limited to specific scenarios where wireless networks are deployed, such as shopping malls, universities, and office buildings.

[0143] This solution calls on the general natural language understanding, intent escape, and text integration capabilities of the big model. It can be directly applied without retraining the big model, thereby reducing the computing power cost of using the big model in the solution.

[0144] 4.3 Model application:

[0145] Used for network fault diagnosis and analysis based on the trained model.

[0146] 5. Fault diagnosis based on knowledge graph:

[0147] The big model technology is used to build a knowledge graph of wireless network failures, which includes information such as common failure types, symptoms, solutions, device associations, and failure causes.

[0148] When a failure occurs, the big model can analyze the cause of the failure based on real-time data and knowledge graphs, and provide more accurate and effective solutions.

[0149] The specific steps are as follows:

[0150] 5.1 Knowledge Graph Construction:

[0151] By using big model technology, combined with expert experience and historical fault data, a knowledge graph of wireless network faults is constructed, which contains information such as common fault types, symptoms, solutions, device associations, and fault causes.

[0152] 5.2 Fault diagnosis:

[0153] When a fault occurs, the big model can analyze the cause of the fault based on real-time data and knowledge graphs, and provide more accurate and effective solutions. For example, based on information such as device logs, signal strength changes, and user feedback, it can be determined that the device may have a hardware fault, software fault, or configuration problem.

[0154] 6. Fault prediction and early warning:

[0155] Analyze the prediction results of the large model to determine the possibility and severity of the failure.

[0156] Based on the prediction results, early warning information is generated, including fault type, possible causes, and recommended solutions.

[0157] Present the warning information to users in the form of charts, text, voice, etc., and select the appropriate push method, such as SMS, email, APP notification, etc.

[0158] The specific steps are as follows:

[0159] 6.1 Analysis of prediction results:

[0160] Analyze the prediction results of the large model to determine the possibility and severity of the failure, and set different warning thresholds based on user needs.

[0161] 6.2 Generation of early warning information:

[0162] Based on the prediction results, early warning information is generated, including fault type, possible causes, and recommended solutions.

[0163] 6.3 Warning information push:

[0164] Present the warning information to users in the form of charts, text, voice, etc., and select the appropriate push method, such as SMS, email, APP notification, etc.

[0165] 7. Visualization of multi-dimensional data display:

[0166] Use three-dimensional heat maps to display the spatial distribution of indicators such as signal strength, delay, and jitter in the digital twin model, helping users to intuitively understand network performance.

[0167] Use time series graphs to show the changing trends of network operation data over time, such as signal strength, load conditions, number of users, etc., to facilitate users to analyze network performance changes and fault trends.

[0168] The wireless network fault knowledge graph is visualized to facilitate users to understand fault types, symptoms, solutions and other information.

[0169] Specific expansion may include:

[0170] 7.1 Three-dimensional heat map:

[0171] Use three-dimensional heat maps to display the spatial distribution of indicators such as signal strength, delay, and jitter in the digital twin model, helping users to intuitively understand network performance.

[0172] 7.2 Time Series Plot:

[0173] Use time series graphs to show the changing trends of network operation data over time, such as signal strength, load conditions, number of users, etc., to facilitate users to analyze network performance changes and fault trends.

[0174] 7.3 Knowledge Graph Visualization:

[0175] The wireless network fault knowledge graph is visualized to facilitate users to understand fault types, symptoms, solutions and other information.

[0176] 8. Troubleshooting:

[0177] After receiving the warning information, users can take measures on their own according to the warning content, such as adjusting network parameters, restarting devices, replacing devices, etc.

[0178] For complex faults, users can choose remote operation and maintenance services, where professionals will perform remote diagnosis and processing.

[0179] The specific steps are as follows:

[0180] 8.1 User autonomous processing:

[0181] After receiving the warning information, users can take measures on their own according to the warning content, such as adjusting network parameters, restarting devices, replacing devices, etc.

[0182] 8.2 Remote Operation and Maintenance:

[0183] For complex faults, users can choose remote operation and maintenance services, where professionals will perform remote diagnosis and processing.

[0184] The user can also evaluate the service effect provided by the device of the present invention and give adjustment instructions for the device, which will be applied to the real-time data collection of the next cycle.

[0185] Furthermore, this embodiment can be applied to the following scenarios:

[0186] Scenario 1: Smart home network fault warning based on multi-source data fusion.

[0187] Build a digital twin model of the home network, including information such as house structure, router location, network parameters, user devices, and environmental information such as temperature and humidity.

[0188] Collect real-time data such as signal strength, traffic, number of users, user device status, temperature, etc. through the router.

[0189] Use multiple models to train a fault prediction model and analyze user feedback text.

[0190] When the system detects that the temperature in the living room is too high and the signal strength decreases, the model can predict that the router may overheat and issue a warning message, suggesting that the user lower the room temperature or restart the router.

[0191] Scenario 2: Enterprise wireless network fault diagnosis based on knowledge graph.

[0192] Build a digital twin model of the enterprise wireless network, including information such as office areas, wireless devices, network parameters, and user devices.

[0193] Collect real-time data such as signal strength, latency, number of users, traffic, device status, etc. through wireless devices.

[0194] Build a knowledge graph of wireless network failures, including common failure types, symptoms, solutions, device associations, failure causes, and other information.

[0195] When a wireless device fails, the model can analyze the cause of the failure based on the knowledge graph and real-time data. For example, if the device log shows "insufficient memory", the knowledge graph contains information that the "insufficient memory" failure may be related to "device configuration error", "malware infection" and other reasons.

[0196] Based on the diagnosis results, the model provides solutions to the operation and maintenance personnel, for example, suggesting checking the device configuration or performing a virus scan.

[0197] Scenario 3: Campus wireless network optimization based on dynamic parameter adjustment.

[0198] Build a digital twin model of the campus wireless network, including information such as teaching buildings, dormitories, public areas, wireless devices, network parameters, and user devices.

[0199] Collect real-time data such as signal strength, latency, number of users, traffic, etc. through wireless devices.

[0200] Use deep learning models to analyze real-time data in digital twin models, for example, identify user-dense areas and analyze users’ network usage patterns in different time periods.

[0201] Based on the analysis results, the parameters in the digital twin model are dynamically adjusted. For example, during breaks, more signal strength is concentrated in areas where students are concentrated, and channel allocation is optimized to avoid network congestion.

[0202] The real-time monitoring and fault prediction method of wireless network based on digital twin proposed in this embodiment can realize the instantiation of digital twin network required by system application by configuring control switches for each collected data parameter in the model, and save unnecessary instantiation of digital twin network parameters, thereby reducing model complexity and improving digital twin network performance. And it can adaptively adjust dynamic parameters based on deep learning, adapt to changes in network environment and user behavior characteristics in real time, improve network performance, and reduce maintenance costs. In addition, the advantages of multiple machine learning models can be combined, for example, support vector machines, neural networks, decision trees and other models can be combined, and the advantages of different models can be comprehensively utilized to improve the accuracy of fault prediction and diagnosis and improve the reliability of early warning information.

[0203] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0204] Based on the same inventive concept, the embodiment of the present application also provides a wireless network fault handling device for implementing the wireless network fault handling method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more wireless network fault handling device embodiments provided below can refer to the limitations of the wireless network fault handling method above, and will not be repeated here.

[0205] In one embodiment, as shown in FIG6 , a wireless network fault processing device is provided, including: a simulation data acquisition module 601, a fault type prediction module 602 and a fault information analysis module 603, wherein:

[0206] A simulation data acquisition module 601 is used to simulate a digital twin model pre-built for a target wireless network to obtain simulated operation data of the target wireless network;

[0207] The fault type prediction module 602 is used to input the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtain the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models;

[0208] The fault information analysis module 603 is used to match the predicted fault type using the knowledge graph and text processing model for wireless network faults to obtain fault analysis information of the target wireless network; the fault analysis information is used to perform fault processing on the target wireless network.

[0209] In one embodiment, multiple types of wireless network fault prediction models include: at least two of a support vector machine model, a neural network model, and a decision tree model; a fault type prediction module 602 is further used to input simulation operation data into the support vector machine model, and obtain fault characterization classification data corresponding to the simulation operation data through the support vector machine model; input the fault characterization classification data into the neural network model, and obtain the probability of the target wireless network corresponding to different fault types through the neural network model; input the probability into the decision tree model, and obtain the predicted fault type of the target wireless network from each fault type through the decision tree model.

[0210] In one embodiment, the wireless network fault handling device also includes: a twin model update module, which is used to obtain actual operation data of the target wireless network; perform data preprocessing on the actual operation data to obtain processed actual operation data; map the processed actual operation data to the digital twin model, adjust the model parameters of the digital twin model according to the mapping results, and obtain the updated digital twin model of the target wireless network.

[0211] In one embodiment, the digital twin model includes: a network configuration parameter model; a wireless network fault handling device, further including: a twin model construction module, which is used to collect network parameters of the target wireless network according to the actual network configuration information of the target wireless network, and obtain digital twin usage information of the target wireless network; obtain target network parameters that match the digital twin usage information from the network parameters; and use the target network parameters to construct a network configuration parameter model.

[0212] In one embodiment, the digital twin model includes: a user behavior model; a wireless network fault handling device, and also includes: a twin model construction module, which is used to obtain user behavior characteristics of the scene where the target wireless network is located, and the life cycle corresponding to the target wireless network; from the user behavior characteristics, obtain the target user behavior characteristics that match the life cycle; and use the target user behavior characteristics to construct a user behavior model.

[0213] In one embodiment, the knowledge graph stores fault information of various historical faults; the fault information includes fault type information, fault cause information and fault handling method information of the historical faults; the fault information analysis module 603 is also used to display the fault information of the historical faults; the predicted fault type is matched with each fault information stored in the knowledge graph to obtain target fault information whose fault type information is the predicted fault type; based on the fault cause information and fault handling method information corresponding to the target fault information, the fault analysis information of the target wireless network is generated through a text processing model.

[0214] In one embodiment, the fault information analysis module 603 is also used to generate fault warning information of the target wireless network based on the fault analysis information; the fault warning information includes the predicted fault type, fault cause information and fault handling method information of the target wireless network; the fault warning information is sent to the fault handling end for display; the fault handling end is used to perform fault handling on the target wireless network according to the predicted fault type, fault cause information and fault handling method information contained in the displayed fault warning information.

[0215] In one embodiment, the wireless network fault handling device also includes: a simulation data display module, which is used to obtain the spatial distribution information and time change trend information corresponding to multiple network performance indicators of the digital twin model based on the simulation operation data; display the spatial distribution information through a three-dimensional heat map, and display the time change trend information through a time series graph.

[0216] Each module in the above wireless network fault handling device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module above.

[0217] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG7. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a wireless network fault handling method is implemented. The display unit of the computer device is used to form a visually visible picture, which may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0218] Those skilled in the art will understand that the structure shown in FIG. 7 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0219] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0220] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0221] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0222] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0223] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0224] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0225] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A wireless network fault handling method, characterized in that: The method comprises: Simulating a digital twin model pre-built for a target wireless network to obtain simulated operation data of the target wireless network; Inputting the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models; The predicted fault type is matched by using a knowledge graph and a text processing model for wireless network faults to obtain fault analysis information of the target wireless network; the fault analysis information is used to perform fault processing on the target wireless network.

2. The method according to claim 1, characterized in that The multiple types of wireless network fault prediction models include: at least two of a support vector machine model, a neural network model, and a decision tree model; The step of inputting the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models, includes: Inputting the simulated operation data into the support vector machine model, and obtaining fault characterization classification data corresponding to the simulated operation data through the support vector machine model; Inputting the fault characterization classification data into the neural network model, and obtaining the probability of the target wireless network corresponding to different fault types through the neural network model; The probability is input into the decision tree model, and the predicted fault type of the target wireless network is obtained from each fault type through the decision tree model.

3. The method according to claim 1, characterized in that The method further comprises: Acquiring actual operation data of the target wireless network; Performing data preprocessing on the actual operation data to obtain processed actual operation data; The processed actual operation data is mapped to the digital twin model, and the model parameters of the digital twin model are adjusted according to the mapping result to obtain the updated digital twin model of the target wireless network.

4. The method according to claim 1, characterized in that The digital twin model includes: a network configuration parameter model; before simulating the digital twin model pre-built for the target wireless network, it also includes: According to the actual network configuration information of the target wireless network, the network parameters of the target wireless network are collected, and the digital twin usage information of the target wireless network is obtained; From the network parameters, obtain target network parameters matching the digital twin usage information; The network configuration parameter model is constructed using the target network parameters.

5. The method according to claim 1, characterized in that The digital twin model includes: a user behavior model; before simulating the digital twin model pre-built for the target wireless network, it also includes: Acquire user behavior characteristics of the scene where the target wireless network is located, and the life cycle corresponding to the target wireless network; Acquire target user behavior characteristics matching the life cycle from the user behavior characteristics; The user behavior model is constructed using the target user behavior characteristics.

6. The method according to claim 1, characterized in that The knowledge graph stores fault information of various historical faults; The fault information includes fault type information, fault cause information and fault handling method information of the historical fault; the method further includes: Displaying the fault information of the historical fault; The method of matching the predicted fault type by using the knowledge graph and text processing model for wireless network faults to obtain fault analysis information of the target wireless network includes: Matching the predicted fault type with each fault information stored in the knowledge graph to obtain target fault information whose fault type information is the predicted fault type; Based on the fault cause information and fault handling method information corresponding to the target fault information, fault analysis information of the target wireless network is generated through the text processing model.

7. The method according to any one of claims 1 to 6, characterized in that: After matching the predicted fault type using the pre-built knowledge graph and text processing model of wireless network faults to obtain the fault analysis information of the target wireless network, the method further includes: Generate fault warning information of the target wireless network according to the fault analysis information; the fault warning information includes the predicted fault type, fault cause information and fault handling method information of the target wireless network; The fault warning information is sent to a fault processing terminal for display; the fault processing terminal is used to perform fault processing on the target wireless network according to the predicted fault type, fault cause information and fault processing method information contained in the displayed fault warning information.

8. The method according to claim 1, characterized in that The method further comprises: Based on the simulation operation data, obtaining spatial distribution information and time change trend information of the digital twin model corresponding to multiple network performance indicators; The spatial distribution information is displayed through a three-dimensional heat map, and the time change trend information is displayed through a time series map.

9. A wireless network fault handling device, characterized in that: The device comprises: A simulation data acquisition module, used to simulate a digital twin model pre-built for a target wireless network to obtain simulated operation data of the target wireless network; A fault type prediction module, used for inputting the simulation operation data into pre-built multiple types of wireless network fault prediction models, and obtaining the predicted fault type of the target wireless network through the multiple types of wireless network fault prediction models; A fault information analysis module is used to match the predicted fault type using a knowledge graph and a text processing model for wireless network faults to obtain fault analysis information of the target wireless network; the fault analysis information is used to perform fault processing on the target wireless network.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

12. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images