A method, device and electronic equipment for locating an implicit fault of a mobile communication network

Abstract

This invention discloses a method, apparatus, and electronic device for locating latent faults in mobile communication networks. The method includes: first, acquiring component data of all network nodes, where these components can be basic elements or signaling elements obtained through target model decomposition. The target model is a service model following the consistency principle. Next, the component data of the network nodes are synthesized to obtain the target model value for each network node. Then, the existence of a latent network fault is determined based on the difference in target model values ​​among network nodes at the same time and level. Specifically, if the difference between the target model value of a certain network node and the target model values ​​of other network nodes at the same time and level exceeds a preset threshold, the network node is determined to have a latent network fault. This method can quickly locate latent faults in mobile communication networks, thereby effectively ensuring network stability and reliability.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, and electronic device for locating latent faults in mobile communication networks. Background Technology

[0002] Mobile communication networks operated by telecom operators are vast and complex, typically employing a multi-layered, multi-node network architecture to ensure stability and reliability. Such an architecture usually includes multiple layers such as a core network, transmission network, and access network, with each layer containing multiple network nodes. These nodes are interconnected through various protocols and interfaces, forming a complex network system. By using user behavior analytics technology and conducting in-depth analysis of user service-related data, it is possible to understand user habits and behavioral patterns, thereby providing better services.

[0003] However, faults are inevitable in mobile communication networks. Current technologies primarily detect network faults by monitoring network device alarms, performance metrics, and logs. For example, monitoring network device alarm information, CPU utilization, memory usage, and network traffic can help determine if a device is faulty. Furthermore, analyzing network device alarms, logs, and user signaling can also help identify abnormal behavior and fault information.

[0004] Nevertheless, existing technologies still have some problems and limitations in fault detection. First, these technologies mainly rely on monitoring and analyzing alarm information, performance indicators, and logs of network devices, which typically can only detect obvious faults. For some latent faults, such as discrepancies between service model data of other network nodes and other node data caused by hidden problems of a certain node device, existing technologies struggle to effectively identify them. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the above-mentioned shortcomings of the prior art by proposing a method, device and electronic device for locating latent faults in mobile communication networks. This method can quickly and accurately identify latent faults in mobile communication networks, thereby ensuring the stability and reliability of the network.

[0006] In a first aspect, the present invention provides a method for locating latent faults in mobile communication networks, the method comprising the following steps:

[0007] Step S1: Obtain the component element data of all network nodes;

[0008] The constituent element data is obtained by decomposing the target model; the target model is a service model that conforms to the consistency principle and is constructed in advance based on the operator's network topology and the service types carried by each network node.

[0009] Step S2: Synthesize the component element data of the network nodes to obtain the target model value of the network nodes;

[0010] Step S3: Based on the differences in target model values ​​between network nodes at the same time and level, determine whether there are any hidden network faults;

[0011] If the target model value of a certain network node differs from the target model value of other network nodes at the same time and level by more than a preset threshold, then the network node is determined to have a hidden network fault, thereby completing the location of the hidden network fault.

[0012] Furthermore, the constituent element data in step S1 specifically includes: basic element data and signaling element data;

[0013] The basic element data is obtained from a professional network management or integrated network system; the signaling element data is obtained from the service provider through the capability open API interface.

[0014] Furthermore, the target model is an average call duration model per user and / or an average internet speed model per user;

[0015] When the target model is the average internet traffic per user model, the constituent elements of the network node include total interface traffic and the number of internet users;

[0016] Step S2 specifically includes:

[0017] The total traffic of the interface and the number of internet users are combined to obtain the average internet traffic per user model value of the network node. The combination formula is as follows:

[0018] Average internet traffic per user = Total interface traffic / Number of internet users.

[0019] Furthermore, after step S3, the method further includes step S4:

[0020] Step S4: Report the hidden network faults to the terminal APP so that professionals can handle them.

[0021] Furthermore, in step S3, the differences between the target model values ​​specifically include: the difference in error, the difference in standard deviation, or the difference in variance between the target model values;

[0022] When the difference between the target model values ​​is the difference in standard deviation, step S3 specifically includes the following steps:

[0023] Step S31: Calculate the standard deviation between the target model values, using the following formula:

[0024]

[0025] Where Y is the standard deviation; μ is the mean of the target model values; N is the number of target model values, and N is a natural number greater than 1; X i Let i be the target model value of the i-th target model, where i ∈ N;

[0026] Step S32: Calculate the deviation coefficient based on the standard deviation difference;

[0027] The deviation coefficient C v The calculation formula is as follows:

[0028] C v = (Y / μ)*100%;

[0029] Where Y is the standard deviation; μ is the mean of the target model values;

[0030] Step S33: Determine whether the deviation coefficient exceeds a preset deviation threshold.

[0031] If the deviation coefficient exceeds the preset deviation threshold, the corresponding network node is determined to have a hidden fault.

[0032] Secondly, the present invention provides a device for locating latent faults in mobile communication networks, the device comprising:

[0033] The acquisition unit is used to acquire the component element data of all network nodes;

[0034] The constituent element data is obtained by decomposing the target model; the target model is a service model that conforms to the consistency principle and is constructed in advance based on the operator's network topology and the service types carried by each network node.

[0035] A synthesis unit, connected to the acquisition unit, is used to synthesize the component element data of the network node to obtain the target model value of the network node;

[0036] The determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the difference between the target model values ​​of network nodes at the same time and level.

[0037] If the difference between the target model value of a certain network node and the target model value of other network nodes at the same time and level exceeds a preset threshold, the determination unit determines that the network node has a hidden network fault, thereby completing the location of the hidden network fault.

[0038] Furthermore, the acquisition unit includes:

[0039] The first acquisition unit, connected to the synthesis unit, is used to acquire basic element data of all network nodes from a professional network management or integrated network system, so that the synthesis unit can synthesize the basic element data of the network nodes to obtain the target model value of the network nodes.

[0040] The second acquisition unit, connected to the synthesis unit, is used to acquire signaling element data of all network nodes from the service provider through the capability open API interface, so that the synthesis unit can synthesize the signaling element data of the network nodes to obtain the target model value of the network nodes.

[0041] Further, the synthesis unit includes:

[0042] The first synthesis unit, connected to the acquisition unit, is used to synthesize an average call duration model per user.

[0043] The second synthesis unit, connected to the acquisition unit, is used to synthesize an average internet speed model per user.

[0044] The first synthesis unit stores the following formula:

[0045] Average internet traffic per user = Total interface traffic / Number of internet users.

[0046] Furthermore, the determination unit includes:

[0047] The first determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the error difference between the target model values ​​of network nodes at the same time and level.

[0048] The second determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the difference in standard deviation between the target model values ​​of network nodes at the same time and level.

[0049] The third determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the variance difference between the target model values ​​of network nodes at the same time and level.

[0050] The first determination unit includes:

[0051] The first calculation module is used to calculate the standard deviation between the target model values;

[0052] The first calculation module stores the formula for calculating the standard deviation:

[0053]

[0054] Where Y is the standard deviation; μ is the mean of the target model values; N is the number of target model values, and N is a natural number greater than 1; Xi Let i be the target model value of the i-th target model, where i ∈ N;

[0055] The second calculation module, connected to the first calculation module, is used to calculate the deviation coefficient based on the standard deviation difference;

[0056] The second calculation module stores the deviation coefficient C. v The calculation formula is as follows:

[0057] C v = (Y / μ)*100%;

[0058] Where Y is the standard deviation; μ is the mean of the target model values;

[0059] The determination module, connected to the second calculation module, is used to determine whether the deviation coefficient exceeds a preset deviation threshold.

[0060] If the deviation coefficient exceeds the preset deviation threshold, the corresponding network node is determined to have a hidden fault.

[0061] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to implement the method for locating latent faults in a mobile communication network as described in the first aspect.

[0062] This invention is based on a consistency principle service model, which can quickly and accurately identify hidden faults in mobile communication networks, thereby ensuring network stability and reliability.

[0063] The specific beneficial effects are as follows:

[0064] 1. Rapid Fault Location: Based on the consistency principle, this invention utilizes a business model that comprehensively analyzes data from multiple network nodes to quickly detect and locate hidden faults. This method can instantly identify data inconsistencies between different network nodes, providing a strong basis for timely problem-solving.

[0065] 2. Accurate Data Analysis: This invention fully utilizes the correlation between different network nodes and applies the consistency principle to conduct comprehensive data analysis. This method can effectively capture potential hidden network problems, significantly improve the accuracy of fault location, enable network maintenance personnel to obtain accurate fault diagnosis information in real time, shorten fault handling time, and thus provide more stable network services.

[0066] 3. Improved user experience: By quickly and accurately detecting and locating network faults, this invention effectively reduces service interruptions and user complaints caused by faults, thereby improving user satisfaction and loyalty.

[0067] 4. Reduced operation and maintenance costs: This invention reduces the need for manual intervention through automated fault detection and location processes, alleviating the workload of operation and maintenance personnel and thus saving labor and time costs.

[0068] 5. Enhanced network security: This invention can quickly locate faults, thereby reducing potential security risks and improving the security of the entire network.

[0069] 6. Support for decision-making: Through precise fault analysis and location, this invention provides support for network operation and maintenance decisions, helps optimize network resource allocation, and achieves more efficient network management. Attached Figure Description

[0070] Figure 1 This is a schematic diagram of a method for locating latent faults in a mobile communication network according to an embodiment of the present invention.

[0071] Figure 2 This is a schematic diagram of the multi-level communication network service carrying of the operator in an embodiment of the present invention;

[0072] Figure 3 This is a schematic diagram of the multi-level communication network service bearer topology for 4G data services of a provincial operator in an embodiment of the present invention;

[0073] Figure 4 This is a schematic diagram illustrating the anomaly location of a multi-level communication network service bearer topology for a provincial operator's 4G data service in an embodiment of the present invention.

[0074] Figure 5 This is a schematic diagram of a device for locating latent faults in a mobile communication network according to an embodiment of the present invention.

[0075] Reference numerals: 10, acquisition unit; 20, synthesis unit; 30, determination unit. Detailed Implementation

[0076] 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.

[0077] 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.

[0078] It is understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0079] 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.

[0080] 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.

[0081] 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.

[0082] 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.

[0083] 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.

[0084] Example 1:

[0085] This embodiment provides a method for locating latent faults in mobile communication networks. Primarily used for network operation and maintenance, this method relies on real-time monitoring and data analysis to quickly identify potential latent faults, thereby improving troubleshooting efficiency and ensuring the stability of communication services. When handling user complaints, this method can quickly pinpoint the root cause of the problem, improving user experience. Simultaneously, it supports network optimization, enabling maintenance personnel to anticipate and resolve network performance bottlenecks. Furthermore, this method also has a risk assessment function to ensure the reliability of the network architecture, thereby improving the overall quality and security of the mobile communication network.

[0086] The technical solution of this embodiment is based on the consistency principle of user service models. By collecting relevant data of specified service models on different network nodes at each network level, and using a pre-built system to process this data, it analyzes whether there are significant differences between the service model values ​​of a certain network node and other network nodes at the same level (e.g., expressed by apparent error, standard deviation, or variance). If the difference is found to be too large, it can be determined that there is a hidden fault in the network node.

[0087] To ensure service security and sufficient network capacity in complex multi-level communication networks, specific services are typically randomly distributed to multiple network nodes at the same level for shared support. The "service model consistency principle" in this embodiment refers to the fact that during service scheduling, operators randomly distribute a large number of similar services with specific weights to network elements at the next lower network level based on factors such as the capacity of the next-level network. Due to the randomness of service distribution, theoretically, at the same point in time (within a certain period), the total service model values ​​(such as total call duration, total internet traffic, etc.) carried by different network elements at the next lower level will show significant differences due to the different distribution weights at the previous level. However, some average-valued service model values ​​(such as average call duration per user, average internet traffic per user, etc.) will not show significant deviations across different network levels and nodes.

[0088] In carrier networks, under normal circumstances (without network anomalies such as individual network node failures), the values ​​of some average-based service models will not change abnormally due to the location of the data source (different network layers or network nodes). Therefore, when calculating the same amount of time, the calculation results of average-based service models at different network layers or network element nodes often exhibit very similar characteristics.

[0089] like Figure 2 As shown, this embodiment is based on a multi-level distribution network structure, containing network nodes at different levels. Specifically, it includes a multi-level control and distribution network:

[0090] Level 1 control and distribution network:

[0091] This is the highest level in the entire network, responsible for overall control and management. Its main task is to coordinate the operation of lower-level nodes and data distribution.

[0092] Secondary distribution network nodes:

[0093] These nodes are located below the primary control network and are primarily responsible for receiving data or instructions from the primary network and distributing them to lower-level nodes.

[0094] Three-tier distribution network nodes:

[0095] These nodes, located below the secondary nodes, are responsible for further distributing data or instructions. N-level distribution network nodes:

[0096] N-level distribution network nodes:

[0097] These nodes represent lower-level distribution nodes in the network, typically at levels beyond the third level.

[0098] This multi-level distribution network architecture is typically used in large-scale data distribution systems, such as Content Delivery Networks (CDNs) or distributed computing systems. Each level of nodes is responsible for distributing data or tasks to nodes closer to the end user, improving data transmission efficiency and response speed. Through this hierarchical architecture, the system can effectively manage and distribute large amounts of data while ensuring network stability and reliability.

[0099] like Figure 1 As shown in the figure, this embodiment provides a method for locating latent faults in mobile communication networks, which specifically includes the following steps:

[0100] Step S1: Obtain the component element data of all network nodes; the component element data is obtained by decomposing the target model; the target model is a service model that conforms to the consistency principle and is pre-constructed based on the operator's network topology and the service types carried by each level of network nodes.

[0101] Specifically, the component element data includes: basic element data and signaling element data. The basic element data is obtained from professional network management or integrated network systems, while the signaling element data is obtained from service providers through capability open API interfaces. Component element data can be divided into basic element data and signaling element data, both of which provide crucial information for network management and fault location. Basic element data is the core information required for the network to maintain normal operation, including device status, performance indicators, device traffic, and user access status. This data helps operators monitor the health of the network in real time and provides necessary basis for network optimization. The acquisition of basic element data mainly relies on professional network management systems (network management systems) and integrated network systems. These systems possess powerful data collection and analysis capabilities, enabling real-time monitoring of various devices and nodes in the network, recording their operating status and performance. Through effective data collection, operators can promptly identify potential problems and take corresponding measures to ensure network stability and reliability. Signaling element data focuses on communication and signal exchange between devices in the network, mainly involving control signaling and user requests. This data is typically obtained from service providers through capability open API interfaces. This open interface allows for seamless data exchange between different systems, enabling operators to obtain real-time signaling information regarding call connection status, data transmission status, and other network functions. By analyzing signaling element data, operators can gain a deeper understanding of user behavior, respond promptly to customer needs, and thus improve user experience. Target models include various models such as the average call duration per user and the average internet speed per user.

[0102] Step S2: Synthesize the component data of the network nodes to obtain the target model value of the network nodes.

[0103] Specifically, when the target model is an average internet traffic per user model, the component data of the network node includes total interface traffic and the number of internet users. Synthesizing the component data of the network node involves combining the total interface traffic and the number of internet users to obtain the average internet traffic per user model value for the network node. The synthesis formula is as follows:

[0104] Average internet traffic per user = Total interface traffic / Number of internet users.

[0105] The synthesis of network node component data aims to generate target model values ​​for network nodes, helping operators evaluate network performance and user experience. In this process, the target model value is derived by combining different data points. Specifically, when the target model is an average traffic per user model, it relies on two key components: total interface traffic and the number of users accessing the network. This synthesis provides operators with a more intuitive understanding of network usage, enabling better management and optimization. Specifically, the total interface traffic represents the sum of all user traffic passing through the network node within a specific time period. This data is collected in real-time through network device monitoring functions to ensure an accurate reflection of overall network consumption. On the other hand, the number of users accessing the network indicates the total number of users simultaneously using that interface, reflecting network user activity and load. Combining these two data points allows operators to gain a clearer understanding of the current network load and efficiency. The average traffic per user model value is calculated using the synthesis formula: "Average traffic per user model value = Total interface traffic / Number of users accessing the network." This formula means that it divides the total traffic by the actual number of users to obtain the average traffic consumed by each user. This indicator can not only help operators identify abnormal traffic in the network, but also provide a basis for analyzing user behavior during peak and off-peak periods, thereby providing a scientific basis for network optimization and resource allocation.

[0106] Step S3: Based on the differences between the target model values ​​of network nodes at the same time and level, determine whether there is a hidden network fault; if the difference between the target model value of a certain network node and the target model value of other network nodes at the same time and level exceeds a set threshold (the threshold is preset), then it is determined that the network node has a hidden network fault, thereby completing the location of the hidden network fault.

[0107] Specifically, in step S3, the differences between the target model values ​​include: the difference in error, the difference in standard deviation, or the difference in variance between the target model values;

[0108] Error variance refers to the difference between the target model value and the actual observed value, reflecting the accuracy of the model's predictions. It can be measured by calculating the difference between each model value and the actual value (such as absolute error or relative error). A large error variance usually indicates that the model's predictive ability is insufficient or unstable, while a smaller error variance indicates that the model can reflect the actual situation well. Therefore, error variance is an important indicator for evaluating model performance and optimizing the model.

[0109] Variance variance refers to the degree of fluctuation among a set of target model values. It is measured by calculating the deviation of model values ​​from their mean. A larger variance indicates more significant differences and fluctuations among model values, reflecting instability and inconsistency in model predictions; conversely, a smaller variance indicates relatively concentrated model values ​​and more consistent prediction results. Analyzing variance variance helps assess the stability of a model under different conditions and optimize the model to improve its predictive reliability.

[0110] When the difference between the target model values ​​is the difference in standard deviation, step S3 specifically includes the following steps:

[0111] Step S31: Calculate the standard deviation between the target model values, using the following formula:

[0112]

[0113] Where Y is the standard deviation; μ is the mean of the target model values; N is the number of target model values, and N is a natural number greater than 1; X i Let i be the target model value of the i-th target model, where i ∈ N;

[0114] Step S32: Calculate the deviation coefficient based on the standard deviation difference;

[0115] The deviation coefficient C r The calculation formula is as follows:

[0116] C v = (Y / μ)*100%;

[0117] Where Y is the standard deviation; μ is the mean of the target model values;

[0118] Step S33: Determine whether the deviation coefficient exceeds a preset deviation threshold.

[0119] If the deviation coefficient exceeds the preset deviation threshold, the corresponding network node is determined to have a hidden fault.

[0120] Assume there are 5 network nodes (N=5), and the total interface traffic and number of users on each node are as follows: Node 1 has a total interface traffic of 1000GB and 100 users, so its average traffic per user model value = total interface traffic / number of users = 10GB / user; Node 2 has a total interface traffic of 1200GB and 100 users, so its average traffic per user model value is 12GB / user; Node 3 has a total interface traffic of 1100GB and 100 users, so its average traffic per user model value is 11GB / user; Node 4 has a total interface traffic of 1300GB and 100 users, so its average traffic per user model value is 13GB / user; and Node 5 has a total interface traffic of 1400GB and 100 users, so its average traffic per user model value is 14GB / user. The target model values ​​for each node are: X1 = 10, X2 = 12, X3 = 11, X4 = 13, X5 = 14.

[0121] Step S31: Calculate the standard deviation between the target model values:

[0122] First, calculate the average value μ of these target model values:

[0123] μ=(10+12+11+13+14) / 5=12;

[0124] Then, calculate the square of the difference between each value and the mean:

[0125] (10-12) 2 =4, (12-12) 2 =0,(11-12) 2 =1,

[0126] (13-12) 2 =1,(14-12) 2 =4,

[0127] Next, calculate the average of these squared differences:

[0128] (4+0+1+1+4) / 5=2;

[0129] Finally, calculate the standard deviation.

[0130] Step S32: Calculate the deviation coefficient:

[0131] C v =(Y / μ)*100%=(1.414 / 12)*100%≈11.78%.

[0132] Step S33: Determine whether the deviation coefficient exceeds the preset deviation threshold.

[0133] Assume the preset deviation threshold is 10%. Due to the calculated deviation coefficient C... v The percentage is approximately 11.78%, which is greater than 10%, therefore the corresponding network node is determined to have a hidden fault.

[0134] For multi-level, multi-node network architectures of operators, this embodiment is based on the consistency principle of user service models. It collects user service-related data from each network node at each network level. After processing and analyzing this collected data by a pre-installed system, if significant differences are found between the service model data of a network node or its associated superior and subordinate network nodes and other network nodes at the same level, it can be determined that the network node with the discrepancies may have a hidden fault. This method effectively helps operators quickly identify and locate potential problems in complex network environments, improving network stability and reliability. This embodiment, by collecting user service-related data from each network node at each network level and using a pre-installed system to process and analyze this data, can identify significant differences between service model data, expressed for example, through apparent error, standard deviation, or variance. By applying the consistency principle of user service models, inconsistencies between a network node or its associated superior and subordinate network nodes and other network nodes at the same level can be discovered and accurately located, thereby effectively identifying potential network faults.

[0135] Step S4: Report hidden network faults to the terminal app so that professionals can handle them. Reporting hidden network faults to the terminal app can be achieved in several ways to ensure users are promptly informed of network conditions. First, operators can push real-time fault information directly to the user's terminal app via network protocols, ensuring the timeliness and accuracy of the information. Second, when hidden network faults occur, operators can also use SMS notifications to send fault details and related processing progress to the user's mobile phone, which is particularly suitable for users who may be unable to connect to the network for a short period. In addition, the push notification function within the terminal app can also be used to inform users of the latest fault updates, and users can view historical notifications and fault handling status at any time. To enhance user interaction, the app can also set up a problem feedback channel to encourage users to proactively report problems encountered during use, allowing operators to analyze and handle faults based on the feedback. Sending fault notifications using the user's registered email address or social media account is also an effective communication method. Through these comprehensive methods, users can smoothly obtain information related to hidden network faults, improving user satisfaction. When the deviation coefficient Cv of the service model value calculated by the pre-configured system device exceeds a set value T, the system will determine that the corresponding network node has a hidden fault. For example, if the deviation coefficient Cv calculated by the system exceeds the set value T of 3%, it indicates that a network node with a discrepancy has been found to have a hidden fault. Once the system identifies a network node with a hidden fault, it will automatically notify professionals in the form of an alarm. Upon receiving the notification, the professionals will, based on the number of network node elements that triggered the alarm, combined with the network topology and service characteristics, further use manual assistance to more accurately locate the specific network node element that has experienced the fault.

[0136] This paper details the discovery and location process of a hidden Gi-FW fault by a telecommunications operator. In this scenario, the bearer topology of the operator's 4G data service multi-level communication network is as follows: Figure 3 As shown.

[0137] Following the Spring Festival one year, this telecom operator received numerous user complaints about lag and buffering issues with 4G internet access in multiple cities across the province; however, the exact locations of the users were uncertain. Initial investigations ruled out wireless network faults; no fault alarms were reported at any level of network nodes, including the bearer network and core network, and daily performance indicators were normal. To further investigate, the maintenance team conducted extensive dial-up tests on the reported 4G internet problems and thoroughly tracked signaling, but found no anomalies. Before developing the method described in this invention, two months of investigation and analysis failed to clearly define the problem, leaving the maintenance team in a difficult situation.

[0138] After a series of analyses and in-depth considerations, the inventor's R&D team discovered that 4G internet access services are distributed proportionally and randomly to different network nodes based on the carrying capacity of each network level. Under normal circumstances, the "average traffic per user model" value of each network node should remain consistent; that is, at the same time, the service model value of different network nodes should not differ significantly due to differences in network level and node. However, if an anomaly occurs in a network node at a certain level, it will usually affect the consistency of the "busy-hour average traffic per user model" value.

[0139] Based on this principle, the inventors developed the method in this embodiment. According to the method in this embodiment, the maintenance team collected, calculated, and analyzed data on the "busy-hour average internet traffic per user model" for network nodes at the SAEGW level. Through analysis, the team quickly discovered that the apparent error of the busy-hour average internet traffic per user for the four network nodes SAEGW13, SAEGW14, SAEGW15, and SAEGW16 was nearly 4 percentage points lower than that of the other 15 peer network nodes, while the deviations for the other nodes were all within 1%. This difference provided a clue that a hidden fault might exist.

[0140] Further analysis determines whether there are potential problems or hidden dangers in the upstream or downstream network nodes of the network node in question. Combining the service characteristics of the upstream and downstream network elements at the SAEGW level, and through comprehensive judgment, the network node affecting user complaints is accurately located in the Gi-FW device (e.g., in computer room 5, a downstream network node at the SAEGW level) of the computer room. Figure 4 (As shown).

[0141] After in-depth analysis of the Gi-FW equipment in data center 5, the root cause of the complaints was quickly identified: an anomaly occurred in the internal forwarding port of a certain service board on the equipment. Calculations revealed that although this latent fault affected only 0.875% of all 4G users in the province, its limited impact made it difficult to pinpoint the specific cause of the 4G internet access problems reported by users through conventional dial-up testing and signaling tracing.

[0142] The above process fully demonstrates the effective application of the technical solution of the present invention in fault location of complex networks.

[0143] This embodiment's method, based on a business model adhering to the consistency principle, can quickly locate hidden faults in mobile communication networks, thereby ensuring network stability and reliability. Through comprehensive analysis of data from multiple network nodes, this embodiment can rapidly discover and locate hidden faults, promptly identify data inconsistencies between different nodes, and provide strong evidence for problem-solving. Furthermore, this method fully utilizes the correlation between nodes to achieve comprehensive data analysis, improving the accuracy of fault location and enabling maintenance personnel to obtain accurate fault diagnosis information in real time, significantly shortening fault handling time and thus improving network service stability. Thanks to this rapid and accurate fault location capability, this embodiment effectively reduces service interruptions and user complaints, thereby improving user satisfaction and loyalty. Simultaneously, this automated fault detection and location process reduces the need for manual intervention, alleviating the workload of maintenance personnel and saving manpower and time costs. In addition, this embodiment can quickly locate faults, reduce potential security risks, and enhance network security. It also provides support for network operation and maintenance decisions, helping to optimize resource allocation and achieve more efficient network management.

[0144] Example 2:

[0145] like Figure 5 As shown, this embodiment provides a device for locating latent faults in mobile communication networks. The device includes:

[0146] Acquisition unit 10 is used to acquire the component element data of all network nodes;

[0147] The constituent element data are basic elements or signaling elements obtained by decomposing the target model; the target model is a service model that conforms to the consistency principle and is constructed in advance based on the operator's network topology and the service types carried by each network node.

[0148] The synthesis unit 20, connected to the acquisition unit 10, is used to synthesize the constituent element data of the network node to obtain the target model value of the network node.

[0149] The determination unit 30, connected to the synthesis unit 20, is used to determine whether there is a hidden network fault based on the difference between the target model values ​​of network nodes at the same time and level.

[0150] If the difference between the target model value of a certain network node and the target model value of other network nodes at the same time and level exceeds a preset threshold, the determination unit determines that the network node has a hidden network fault, thereby completing the location of the hidden network fault.

[0151] As one specific implementation, the acquisition unit 10 includes:

[0152] The first acquisition unit, connected to the synthesis unit, is used to acquire basic element data of all network nodes from a professional network management or integrated network system, so that the synthesis unit can synthesize the basic element data of the network nodes to obtain the target model value of the network nodes.

[0153] The second acquisition unit, connected to the synthesis unit, is used to acquire signaling element data of all network nodes from the service provider through the capability open API interface, so that the synthesis unit can synthesize the signaling element data of the network nodes to obtain the target model value of the network nodes.

[0154] As one specific implementation, the synthesis unit 20 includes:

[0155] The first synthesis unit, connected to the acquisition unit, is used to synthesize an average call duration model per user.

[0156] The second synthesis unit, connected to the acquisition unit, is used to synthesize an average internet speed model per user.

[0157] The first synthesis unit stores the following formula:

[0158] Average internet traffic per user = Total interface traffic / Number of internet users.

[0159] As one specific implementation, the determination unit 30 includes:

[0160] The first determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the error difference between the target model values ​​of network nodes at the same time and level.

[0161] The second determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the difference in standard deviation between the target model values ​​of network nodes at the same time and level.

[0162] The third determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the variance difference between the target model values ​​of network nodes at the same time and level.

[0163] The first determination unit includes:

[0164] The first calculation module is used to calculate the standard deviation between the target model values;

[0165] The first calculation module stores the formula for calculating the standard deviation:

[0166]

[0167] Where Y is the standard deviation; μ is the mean of the target model values; N is the number of target model values, and N is a natural number greater than 1; X i Let i be the target model value of the i-th target model, where i ∈ N;

[0168] The second calculation module, connected to the first calculation module, is used to calculate the deviation coefficient based on the standard deviation difference;

[0169] The second calculation module stores the deviation coefficient C. v The calculation formula is as follows:

[0170] C v = (Y / μ)*100%;

[0171] Where Y is the standard deviation; μ is the mean of the target model values;

[0172] The determination module, connected to the second calculation module, is used to determine whether the deviation coefficient exceeds a preset deviation threshold.

[0173] If the deviation coefficient exceeds the preset deviation threshold, the corresponding network node is determined to have a hidden fault.

[0174] The apparatus in this embodiment is capable of performing the method in Embodiment 1.

[0175] Example 3:

[0176] This embodiment provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to implement the method for locating latent faults in a mobile communication network as described in Embodiment 1.

[0177] 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 method for locating latent faults in a mobile communication network, characterized in that, The method includes the following steps: Step S1: Obtain the component element data of all network nodes; The constituent element data is obtained by decomposing the target model; the target model is a service model that conforms to the consistency principle and is constructed in advance based on the operator's network topology and the service types carried by each network node. Step S2: Synthesize the component element data of the network nodes to obtain the target model value of the network nodes; Step S3: Based on the differences in target model values ​​between network nodes at the same time and level, determine whether there are any hidden network faults; If the difference between the target model value of a network node and the target model value of other network nodes at the same time and level exceeds a preset threshold, the network node is determined to have a hidden network fault, thereby completing the location of the hidden network fault.

2. The method for locating latent faults in mobile communication networks according to claim 1, characterized in that, The constituent element data in step S1 specifically includes: basic element data and signaling element data; The basic element data is obtained from a professional network management or integrated network system; the signaling element data is obtained from the service provider through the capability open API interface.

3. The method for locating latent faults in mobile communication networks according to claim 1, characterized in that, The target model is an average call duration per user model and / or an average internet speed per user model. When the target model is the average internet traffic per user model, the constituent elements of the network node include total interface traffic and the number of internet users; Step S2 specifically includes: The total traffic of the interface and the number of internet users are combined to obtain the average internet traffic per user model value of the network node. The combination formula is as follows: Average internet traffic per user = Total interface traffic / Number of internet users.

4. The method for locating latent faults in mobile communication networks according to claim 1, characterized in that, Following step S3, the method further includes step S4: Step S4: Report the hidden network faults to the terminal APP so that professionals can handle them.

5. The method for locating latent faults in a mobile communication network according to any one of claims 1 to 4, characterized in that, In step S3, the differences between the target model values ​​specifically include: the error difference, standard deviation difference, or variance difference between the target model values. When the difference between the target model values ​​is the difference in standard deviation, step S3 specifically includes the following steps: Step S31: Calculate the standard deviation between the target model values, using the following formula: Where Y is the standard deviation; μ is the mean of the target model values; N is the number of target model values, and N is a natural number greater than 1; X i Let i be the target model value of the i-th target model, where i ∈ N; Step S32: Calculate the deviation coefficient based on the standard deviation difference; The deviation coefficient C v The calculation formula is as follows: C v =(Y / μ)*100%; Where Y is the standard deviation; μ is the mean of the target model values; Step S33: Determine whether the deviation coefficient exceeds a preset deviation threshold. If the deviation coefficient exceeds the preset deviation threshold, the corresponding network node is determined to have a hidden fault.

6. A device for locating latent faults in mobile communication networks, characterized in that, include: The acquisition unit is used to acquire the component element data of all network nodes; The constituent element data is obtained by decomposing the target model; the target model is a service model that conforms to the consistency principle and is constructed in advance based on the operator's network topology and the service types carried by each network node. A synthesis unit, connected to the acquisition unit, is used to synthesize the component element data of the network node to obtain the target model value of the network node; The determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the difference between the target model values ​​of network nodes at the same time and level. If the difference between the target model value of a certain network node and the target model value of other network nodes at the same time and level exceeds a preset threshold, the determination unit determines that the network node has a hidden network fault, thereby completing the location of the hidden network fault.

7. The device for locating latent faults in mobile communication networks according to claim 6, characterized in that, The acquisition unit includes: The first acquisition unit, connected to the synthesis unit, is used to acquire basic element data of all network nodes from a professional network management or integrated network system, so that the synthesis unit can synthesize the basic element data of the network nodes to obtain the target model value of the network nodes. The second acquisition unit, connected to the synthesis unit, is used to acquire signaling element data of all network nodes from the service provider through the capability open API interface, so that the synthesis unit can synthesize the signaling element data of the network nodes to obtain the target model value of the network nodes.

8. The device for locating latent faults in mobile communication networks according to claim 6, characterized in that, The synthesis unit includes: The first synthesis unit, connected to the acquisition unit, is used to synthesize an average call duration model per user. The second synthesis unit, connected to the acquisition unit, is used to synthesize an average internet speed model per user. The first synthesis unit stores the following formula: Average internet traffic per user = Total interface traffic / Number of internet users.

9. The device for locating latent faults in mobile communication networks according to any one of claims 6 to 8, characterized in that, The determination unit includes: The first determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the error difference between the target model values ​​of network nodes at the same time and level. The second determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the difference in standard deviation between the target model values ​​of network nodes at the same time and level. The third determination unit, connected to the synthesis unit, is used to determine whether there is a hidden network fault based on the variance difference between the target model values ​​of network nodes at the same time and level. The first determination unit includes: The first calculation module is used to calculate the standard deviation between the target model values; The first calculation module stores the formula for calculating the standard deviation: Where Y is the standard deviation; μ is the mean of the target model values; N is the number of target model values, and N is a natural number greater than 1; X i Let i be the target model value of the i-th target model, where i ∈ N; The second calculation module, connected to the first calculation module, is used to calculate the deviation coefficient based on the standard deviation difference; The second calculation module stores the deviation coefficient C. v The calculation formula is as follows: C v =(Y / μ)*100%; Where Y is the standard deviation; μ is the mean of the target model values; The determination module, connected to the second calculation module, is used to determine whether the deviation coefficient exceeds a preset deviation threshold. If the deviation coefficient exceeds the preset deviation threshold, the corresponding network node is determined to have a hidden fault.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to implement the method for locating latent faults in a mobile communication network as described in any one of claims 1-5.

Images