Method and device for early warning of radio group obstacle, electronic equipment and storage medium

Abstract

The application discloses a wireless group obstacle early warning method and device, electronic equipment and storage medium. The method comprises the following steps: in the case that it is monitored that there is a communication failure in a preset area, obtaining network element attribute data of a plurality of target wireless network elements in the preset area, wherein the plurality of target wireless network elements are used to represent wireless network elements that have failed in the preset area between a preset historical time and a current time; determining a target closeness degree of the plurality of target wireless network elements; predicting a wireless group obstacle of the preset area based on the target closeness degree to obtain a prediction result, wherein the prediction result is used to represent whether the preset area will have a wireless group obstacle at a preset future time; and in the case that the prediction result is that the preset area will have a wireless group obstacle at the preset future time, generating wireless group obstacle early warning information based on the network element attribute data. The application solves the technical problems that it is difficult to early warn the wireless group obstacle and the efficiency of judging the wireless group obstacle is low in the related art.

Description

Technical Field

[0001] This invention relates to the field of communications, and more specifically, to a method, apparatus, electronic device, and storage medium for early warning of wireless network faults. Background Technology

[0002] Wireless cluster failure refers to a situation where a large number of wireless network elements within a contiguous geographical area simultaneously or successively experience service outages within a short period of time, resulting in the inability to provide wireless service. Because wireless cluster failures involve numerous wireless network elements located in adjacent, contiguous geographical areas, when these elements simultaneously fail to provide service, users within the area cannot rely on wireless network elements at their current or adjacent locations, significantly impacting wireless communication for users within this geographical area.

[0003] Currently, related technologies typically only detect and generate alarm information after wireless network failures have occurred or spread over a large area, resulting in low timeliness. Alternatively, they rely on human judgment to determine if wireless network elements in a certain area are successively generating outage alarms. This subjective judgment process is heavily dependent on human experience, making it difficult to provide early warnings of wireless network failures and resulting in low efficiency in identifying wireless network failures.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method, apparatus, electronic device, and storage medium for early warning of wireless faults, in order to at least solve the technical problems in the related art of difficulty in providing early warning of wireless faults and low efficiency in judging wireless faults.

[0006] According to one aspect of the embodiments of this application, a method for early warning of wireless network failure is provided, comprising: when a communication failure is detected in a preset area, acquiring network element attribute data of multiple target wireless network elements in the preset area, wherein the multiple target wireless network elements are used to represent wireless network elements in the preset area that have failed between a preset historical time and the current time; determining the target similarity of the multiple target wireless network elements, wherein the target similarity is used to characterize the spatial clustering degree among the multiple target wireless network elements; performing wireless network failure prediction on the preset area based on the target similarity to obtain a prediction result, wherein the prediction result is used to indicate whether wireless network failure will occur in the preset area at a preset future time, the preset future time being after the current time; if the prediction result indicates that wireless network failure will occur in the preset area at a preset future time, generating wireless network failure early warning information based on the network element attribute data, wherein the wireless network failure early warning information includes: failure cause location results of multiple target wireless network elements failing.

[0007] According to another aspect of the present invention, a wireless cluster failure early warning device is also provided, comprising: an acquisition module, configured to acquire network element attribute data of multiple target wireless network elements in a preset area when a communication failure is detected in a preset area, wherein the multiple target wireless network elements represent wireless network elements in the preset area that have failed between a preset historical time and the current time; a determination module, configured to determine the target similarity of the multiple target wireless network elements, wherein the target similarity is used to characterize the spatial clustering degree among the multiple target wireless network elements; a prediction module, configured to perform wireless cluster failure prediction on the preset area based on the target similarity to obtain a prediction result, wherein the prediction result indicates whether wireless cluster failure will occur in the preset area at a preset future time, the preset future time being after the current time; and a generation module, configured to generate wireless cluster failure early warning information based on the network element attribute data when the prediction result indicates that wireless cluster failure will occur in the preset area at a preset future time, wherein the wireless cluster failure early warning information includes: failure cause location results of multiple target wireless network elements failing.

[0008] According to another aspect of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods of various embodiments of the present invention during runtime.

[0009] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0010] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0011] According to another aspect of the present invention, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0012] According to another aspect of the present invention, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of the present invention.

[0013] In this embodiment of the invention, in response to the detection of a communication fault in a preset area, the network element attribute data of multiple target wireless network elements that have already experienced faults in the preset area are first obtained. Then, the target similarity of the multiple target wireless network elements in the preset area can be determined. The target similarity is used to characterize the spatial clustering degree of the multiple target wireless network elements. Next, based on the target similarity, wireless group fault prediction is performed on the multiple target wireless network elements to obtain a prediction result. The prediction result is used to indicate whether wireless group faults will occur in the preset area at a preset future time. Finally, if the prediction result indicates that wireless group faults will occur in the preset area at a preset future time, wireless group fault warning information is generated, and fault analysis is performed on the multiple target wireless network elements to obtain the fault cause location result of the faults in the multiple target wireless network elements. It is noteworthy that, in response to the detection of a communication fault in a preset area, this application acquires the network element attribute data of multiple target wireless network elements that have already experienced faults within the preset area. By analyzing the spatial clustering degree of these multiple target wireless network elements that have experienced faults between a preset historical time and the current time, it determines whether a wireless group fault will occur in the preset area at a preset future time. If a wireless group fault is expected in the preset area at a preset future time, it can analyze and obtain wireless group fault warning information and fault cause location results. Based on the analysis of spatial clustering degree, it can more accurately determine the correlation between wireless network elements, thereby better predicting the occurrence trend of wireless group faults. It can detect wireless group fault trends earlier, generate wireless group fault warning information in a timely manner, and take corresponding measures before the fault occurs to prevent further escalation. Combined with fault analysis, it can quickly locate the cause of the fault, improve fault handling efficiency, and effectively improve network stability and reliability. This solves the technical problems of difficulty in providing early warning of wireless group faults and low efficiency in judging wireless group faults in related technologies. Attached Figure Description

[0014] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0015] Figure 1 This is a hardware structure block diagram of a computer terminal for implementing a method for early warning of wireless cluster obstacles according to an embodiment of this application.

[0016] Figure 2 This is a flowchart of a wireless fault warning method according to an embodiment of the present invention;

[0017] Figure 3 This is a schematic diagram of an optional wireless obstacle warning system according to an embodiment of the present invention;

[0018] Figure 4 This is a schematic diagram of an optional flexible periodic polling process according to an embodiment of the present invention;

[0019] Figure 5 This is a schematic diagram illustrating the process of determining the similarity of targets under an optional indoor scene type according to an embodiment of the present invention;

[0020] Figure 6 This is a schematic diagram illustrating the process of determining the similarity of targets under an optional outdoor scene type according to an embodiment of the present invention;

[0021] Figure 7 This is a schematic diagram of the target warning coefficient determination process under an optional indoor scene type according to an embodiment of the present invention;

[0022] Figure 8 This is a schematic diagram of the target warning coefficient determination process under an optional outdoor scene type according to an embodiment of the present invention;

[0023] Figure 9 This is a schematic diagram of an optional fault location analysis according to an embodiment of the present invention;

[0024] Figure 10 This is a schematic diagram of a wireless obstacle warning device according to an embodiment of the present invention. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] To address the problems existing in related technologies, embodiments of this application provide a method for predicting wireless obstacle clusters, which can be implemented in... Figure 1 The computer terminal shown is explained below.

[0028] The wireless obstacle prediction method provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal for implementing a method for predicting wireless obstacle clustering is shown. Figure 1 As shown, the computer terminal 10 may include one or more processors (shown as 102a, 102b, ..., 102n in the figure) (the processor may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission module 106 for communication functions connected via wired and / or wireless networks. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0029] It should be noted that the aforementioned one or more processors and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be implemented wholly or partially as software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be wholly or partially integrated into any other element in the computer terminal 10. As involved in the embodiments of this application, the data processing circuits serve as processor control (e.g., selection of a variable resistor termination path connected to an interface).

[0030] The memory 104 can be used to store software programs and modules for application software, such as the program instructions / data storage device corresponding to the wireless obstacle prediction method in this embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned wireless obstacle prediction method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0031] The transmission module 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission module 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission module 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0032] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the account to interact with the account interface of the computer terminal 10.

[0033] It should be noted here that, in some optional embodiments, the above... Figure 1 The computer terminal shown may include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that... Figure 1 This is only one instance of a specific particular instance, and is intended to illustrate the types of components that may exist in the aforementioned computer terminal.

[0034] In the above operating environment, this application provides an embodiment of a method for predicting wireless group faults. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0035] Figure 2 This is a flowchart of a wireless fault warning method according to an embodiment of the present invention, such as... Figure 2As shown, the method includes the following steps:

[0036] Step S202: If a communication fault is detected in a preset area, obtain the network element attribute data of multiple target wireless network elements in the preset area.

[0037] Among them, multiple target wireless network elements are used to represent wireless network elements in a preset area that have experienced failures between a preset historical time and the current time.

[0038] The aforementioned wireless network elements refer to various devices and components in a wireless communication system. They are the basic devices and components that constitute a wireless communication network, and may include, but are not limited to, base stations, wireless access points, wireless sensors, and antenna systems. These devices are the fundamental components of a wireless communication network, used to transmit and receive wireless signals. The functions of wireless network elements can include signal processing, spectrum management, access control, resource allocation, mobile handover, and interference management. Through these functions, wireless network elements can meet the requirements of wireless communication networks in terms of coverage, capacity, quality, and reliability, ensuring uninterrupted communication for users.

[0039] The aforementioned preset area can be a pre-defined geographical region, such as a district, county, community street, or multiple nearby residential areas. The preset area can be divided according to actual needs, and there are no restrictions here.

[0040] The aforementioned preset historical time can refer to a historical time before the current time. For example, it could be 10 minutes or 30 minutes before the current time, or it could be 10 PM yesterday. In other words, the time period from the preset historical time to the current time can be dynamically changing. The preset historical time can be determined according to actual needs, and there is no limitation here.

[0041] The aforementioned wireless cluster failure can refer to a large number of wireless network elements in a contiguous geographical area, such as macro base stations or indoor distribution radio remote units (RRUs), experiencing simultaneous or successive outages within a short period of time. This application can provide early warning of impending wireless cluster failures, enabling timely fault handling and preventing the escalation of wireless cluster failures.

[0042] The aforementioned network element attribute data refers to various parameters and status information of wireless network elements, including but not limited to device information, fault information, and historical data. Device information may include device model, serial number, manufacturer, and version number; fault information may include fault type, fault occurrence time, and fault duration; historical data may include historical operating data and historical fault records. By acquiring network element attribute data, fault cause analysis can be performed on multiple target wireless network elements, facilitating timely fault troubleshooting and improving the stability and reliability of the communication network.

[0043] In one optional embodiment, in response to the detection of a communication fault in a preset area (which could be one or more communication faults), and when it is not yet certain whether the preset area will enter a state of wireless network failure, the network element attribute data of multiple target wireless network elements within the preset area can be obtained first. Specifically, a network element monitoring system deployed within the preset area can monitor the status and attribute data of wireless network elements in real time. The network element monitoring system can collect various attribute data of multiple target wireless network elements such as wireless base stations and wireless transmission equipment, including but not limited to information such as power, signal strength, connection status, and transmission rate. Operators can also monitor and manage the communication network by deploying a network management system, which can obtain information on each wireless network element within the preset area, including historical data and real-time status. The network management system can easily obtain the network element attribute data of multiple target wireless network elements. Sensor technology can also be utilized. By deploying sensor devices within the preset area, the operating status of wireless network elements can be monitored in real time. The sensors can collect various data from the wireless network elements and transmit the data to the monitoring center through the communication network for further analysis and processing. Data mining and analysis techniques can be used to process and analyze the collected network element attribute data of multiple target wireless network elements. By establishing models and algorithms, faulty wireless network elements can be identified, and relevant personnel can be notified promptly. In practical applications, when acquiring network element attribute data for multiple target wireless network elements within a preset area, artificial intelligence and machine learning technologies can be used to improve the accuracy and efficiency of monitoring and early warning. By establishing monitoring models and predictive algorithms, automatic identification and early warning of communication faults can be achieved, reducing manual intervention and improving response speed. Furthermore, big data technology can be combined to deeply mine historical data, analyze the patterns and trends of communication faults, and provide a reference for future prediction and prevention. Through the analysis of large amounts of data, information and patterns hidden behind the data can be discovered, providing operators with more accurate early warning and fault handling solutions. Through automated monitoring and early warning systems, fault points can be quickly identified and addressed, reducing manual intervention and maintenance costs, and improving fault handling efficiency.

[0044] Step S204: Determine the target similarity of multiple target wireless network elements.

[0045] Among them, the target similarity is used to characterize the spatial clustering degree among multiple target wireless network elements.

[0046] The aforementioned target similarity can refer to a parameter used to characterize the spatial clustering characteristics of multiple target wireless network elements. Since the multiple wireless network elements that have failed between the preset historical time and the current time can be dynamically changed, the specific number of wireless network elements and which network element they are can change. That is, the multiple target wireless network elements can be dynamically changed, and the corresponding target similarity will also change. Therefore, the spatial clustering characteristics can specifically characterize at least the quantity characteristics and spatial distance characteristics of the multiple target wireless network elements.

[0047] In one optional embodiment, wireless cluster failure early warning can detect and locate a large number of wireless network element outages at an earlier stage, enabling timely repair measures and minimizing the impact on user experience and network stability. The early warning system can monitor and analyze multiple target wireless network elements within a preset area using a series of technical means to promptly identify potential cluster failures. Specifically, the similarity between multiple target wireless network elements within the preset area can be determined. First, monitoring and data collection of the wireless network elements within the preset area can be performed. This can be achieved through a network management system or specialized monitoring tools. The monitoring data can include information such as the operating status, signal strength, and load of the wireless network elements. Then, based on data from multiple wireless network elements that have failed between a preset historical time and the current time, the similarity between these target wireless network elements is calculated. The calculation of the similarity can be implemented in various ways, such as by calculating the distance between the wireless network elements to assess the similarity between multiple target wireless network elements. This can be evaluated using indicators such as the physical distance between the wireless network elements, signal strength, and load. By assessing the proximity of target wireless network elements, it's possible to detect spatial clustering characteristics, thereby determining the likelihood of wireless network clustering issues. In this process, evaluating the proximity of target wireless network elements allows for a more accurate assessment of wireless network clustering problems, helping to reduce false alarm rates and improve the accuracy of the early warning system. By acquiring the proximity of multiple target wireless network elements within a preset area, the early warning system can more accurately detect wireless network clustering issues, enabling proactive remediation measures and ultimately improving network reliability and user experience.

[0048] Step S206: Based on the similarity of targets, perform wireless obstacle prediction on the preset area to obtain the prediction result.

[0049] The prediction result is used to indicate whether a wireless cluster failure will occur in the preset area at a preset future time, which is after the current time.

[0050] The aforementioned preset future time can refer to a historical time after the current time. For example, it could be 1 minute or 30 minutes after the current time, or it could be a future time such as 10:00 AM tomorrow. The preset future time can be determined according to actual needs, and there is no limitation here.

[0051] In one optional embodiment, early warning of wireless cluster failures predicts the likelihood of a large-scale wireless network failure by analyzing the similarity between multiple target wireless network elements within a preset area. This early warning system helps operators promptly identify potential problems and take corresponding measures to avoid or reduce failures, improving network stability and reliability. Specifically, this step involves first collecting performance data of multiple target wireless network elements within the preset area, including signal strength, data transmission rate, and connection status. Then, the similarity between the multiple target wireless network elements is calculated based on this data, using distance metrics such as Euclidean distance or correlation analysis to assess their similarity. Analyzing this data yields the target similarity level. Next, wireless cluster failure prediction can be performed based on the target similarity level. Machine learning algorithms such as support vector machines, neural networks, or decision trees can be used to build a prediction model, learning from and analyzing historical data to predict future wireless cluster failure scenarios. The prediction result can be a binary classification (presence or absence of wireless cluster failure) or a continuous value (probability of wireless cluster failure), with the appropriate prediction method selected based on the actual situation. Through the above process, signs of wireless network failures can be detected early, allowing for timely measures to prevent the escalation and spread of the fault. Secondly, it can reduce the impact of failures on user experience and service quality, improving network reliability and stability. Furthermore, analyzing and predicting large amounts of wireless network data can help operators better plan network resources and optimize network structure, thereby improving network performance and efficiency.

[0052] Step S208: If the prediction result indicates that a wireless fault will occur in the preset area at a preset future time, generate a wireless fault warning message based on the network element attribute data.

[0053] Among them, the wireless cluster fault early warning information includes: the fault location results of multiple target wireless network elements experiencing faults.

[0054] In one optional embodiment, generating wireless cluster failure early warning information and performing fault analysis on multiple target wireless network elements can enable timely early warning and in-depth analysis of potentially faulty wireless network elements when prediction results indicate that a wireless cluster failure will occur in a preset area at a preset future time. This allows for the location of the fault cause and the implementation of corresponding maintenance measures, minimizing the impact of the fault on network performance. Specifically, when generating the wireless cluster failure early warning information, the system can first determine whether a wireless cluster failure will occur in a preset area at a preset future time based on the prediction results. If so, the system can automatically generate early warning information and send it to relevant maintenance personnel or the system. The early warning information may include the fault cause location results of multiple target wireless network elements, the early warning level, the early warning time, and the early warning objects, so that maintenance personnel can take timely and appropriate countermeasures. Next, when performing fault analysis on multiple target wireless network elements, the system can first monitor and analyze these wireless network elements. By monitoring the performance indicators, log information, alarm information, etc., of the wireless network elements, the system can gain a comprehensive understanding of the working status of the wireless network elements. Then, the system can infer and locate the possible causes of the wireless network element failures based on existing fault location algorithms. These fault location algorithms can analyze historical data from wireless network elements, surrounding environmental factors, device status, and other information to accurately identify the cause of the fault. Finally, the system can take corresponding maintenance measures based on the fault location results. These measures may include remote rebooting, parameter adjustment, and device replacement. Simultaneously, the system can monitor and provide feedback on the execution of these maintenance measures to ensure effective problem resolution. Through these steps, the system can issue timely warnings before a network-wide fault occurs, helping maintenance personnel quickly locate the cause of the fault and take effective measures, minimizing the impact of the fault on network performance. Furthermore, the system can improve the accuracy and efficiency of fault location, reduce maintenance costs, and enhance network stability and reliability.

[0055] In this embodiment of the invention, in response to the detection of a communication fault in a preset area, the network element attribute data of multiple target wireless network elements that have already experienced faults in the preset area are first obtained. Then, the target similarity of the multiple target wireless network elements in the preset area can be determined. The target similarity is used to characterize the spatial clustering degree of the multiple target wireless network elements. Next, based on the target similarity, wireless group fault prediction is performed on the multiple target wireless network elements to obtain a prediction result. The prediction result is used to indicate whether wireless group faults will occur in the preset area at a preset future time. Finally, if the prediction result indicates that wireless group faults will occur in the preset area at a preset future time, wireless group fault warning information is generated, and fault analysis is performed on the multiple target wireless network elements to obtain the fault cause location result of the faults in the multiple target wireless network elements. It is noteworthy that, in response to the detection of a communication fault in a preset area, this application acquires the network element attribute data of multiple target wireless network elements that have already experienced faults within the preset area. By analyzing the spatial clustering degree of these multiple target wireless network elements that have experienced faults between a preset historical time and the current time, it determines whether a wireless group fault will occur in the preset area at a preset future time. If a wireless group fault is expected in the preset area at a preset future time, it can analyze and obtain wireless group fault warning information and fault cause location results. Based on the analysis of spatial clustering degree, it can more accurately determine the correlation between wireless network elements, thereby better predicting the occurrence trend of wireless group faults. It can detect wireless group fault trends earlier, generate wireless group fault warning information in a timely manner, and take corresponding measures before the fault occurs to prevent further escalation. Combined with fault analysis, it can quickly locate the cause of the fault, improve fault handling efficiency, and effectively improve network stability and reliability. This solves the technical problems of difficulty in providing early warning of wireless group faults and low efficiency in judging wireless group faults in related technologies.

[0056] In the above embodiments of this application, the prediction of wireless cluster failure in a preset area based on the target similarity is performed to obtain the prediction result, including: when the target similarity is less than a first similarity threshold, determining that the prediction result is that the preset area will not experience wireless cluster failure at a preset future time, and repeating the step of determining the target similarity of multiple target wireless network elements; when the target similarity is greater than or equal to the first similarity threshold, determining that the prediction result is that the preset area will experience wireless cluster failure at a preset future time.

[0057] The aforementioned first similarity threshold can refer to a pre-set threshold used to determine whether a wireless cluster fault will occur in a preset area at a preset future time. It can be determined according to actual needs and is not limited here.

[0058] In one optional embodiment, during the early warning process of wireless cluster failure, the similarity of targets can be monitored to determine in real time whether the current similarity is less than a first similarity threshold. When the similarity is less than the first similarity threshold, the step of determining the similarity of multiple target wireless network elements can be repeated to continuously monitor the change in the similarity. When the similarity is greater than or equal to the first similarity threshold, it can be determined that a wireless cluster failure will occur in a preset area at a preset future time. When the similarity is greater than or equal to the first similarity threshold, it can be determined that the preset area has a tendency to enter a wireless cluster failure. This can help network operators take measures in advance to avoid the occurrence of wireless cluster failure, thereby improving the stability and reliability of the network. The early determination of the prediction results can help network operators react in a timely manner, ensure the normal operation of the network, effectively reduce the impact of wireless cluster failure on the network, and improve the overall performance and service quality of the network.

[0059] In the above embodiments of this application, the step of repeatedly performing the determination of the target similarity of multiple target wireless network elements includes: when the target similarity is less than a second similarity threshold, the step of repeatedly performing the determination of the target similarity of multiple target wireless network elements is performed based on a first polling period, wherein the second similarity threshold is less than the first similarity threshold; when the target similarity is greater than or equal to the second similarity threshold, the step of repeatedly performing the determination of the target similarity of multiple target wireless network elements is performed based on a second polling period, wherein the second polling period is less than the first polling period.

[0060] The first polling period mentioned above can be a pre-set larger polling period, that is, a smaller polling frequency can be achieved, such as 5 minutes. The first polling period can be determined according to actual needs, and there is no limitation here.

[0061] The second polling period mentioned above can be a pre-set smaller polling period, that is, a larger polling frequency can be achieved, such as 30 seconds, etc. The second polling period can be determined according to actual needs, and there is no limitation here.

[0062] In one optional embodiment, the wireless cluster failure early warning system predicts potential wireless cluster failures by monitoring the proximity of target wireless network elements, enabling timely intervention to reduce the failure rate and impact range. During the prediction process, a proximity threshold can be set to determine whether the distance between target wireless network elements is close, thus deciding whether to proceed with the next monitoring step. Specifically, when the target proximity is less than a second proximity threshold, the system can monitor the proximity between target wireless network elements through a first polling cycle. At this point, the target proximity is low and relatively safe, allowing for a longer polling cycle to balance monitoring changes in target proximity while conserving polling resources. In other words, when the target proximity is less than the set first proximity threshold, it is preliminarily determined that the distance between these wireless network elements is relatively large, and they are unlikely to enter a wireless cluster failure phase; therefore, the target proximity can be monitored using the first polling cycle. If the target similarity is greater than or equal to the first similarity threshold, the change in target similarity can be monitored based on the second polling cycle. This allows for more frequent monitoring of target similarity, meaning the system can monitor the similarity between target wireless network elements according to the set second polling cycle. Based on the target similarity, the system obtains a prediction result to determine whether a wireless group failure may occur. Through these steps, the system can dynamically predict and monitor based on the similarity between wireless network elements, promptly identifying potential group failure risks and taking corresponding intervention and handling measures. During this process, the monitoring cycle can be dynamically adjusted. Different monitoring cycles can be dynamically set according to actual conditions, enabling more accurate capture of wireless group failure trends and improving the accuracy and timeliness of predictions. Based on different similarity thresholds and monitoring cycles, the system can flexibly adjust the sensitivity and accuracy of early warnings to better adapt to the wireless group failure prediction needs in different environments. The target similarity-based wireless group failure prediction system, through setting dynamic and flexible periodic polling, can more effectively help operators and network maintenance personnel predict and handle potential wireless group failure problems, improving network operating efficiency and stability.

[0063] In the above embodiments of this application, determining that a wireless cluster failure will occur in a preset area at a preset future time includes: acquiring target warning parameters of multiple target wireless network elements, wherein the target warning parameters are used to represent parameters determined based on the number of users during the busy period of the preset area, the average resource utilization rate during the busy period, user feedback information on communication quality, and whether the current period belongs to the busy period; if the target warning parameters are greater than or equal to a preset warning parameter threshold, determining that a wireless cluster failure will occur in the preset area at a preset future time.

[0064] The aforementioned number of users during a busy period refers to the number of users within a specific time frame within a preset area. This time frame can be a period of high network activity, such as peak daytime hours or during specific events. By monitoring changes in the number of users during this time frame, the network load can be assessed.

[0065] The aforementioned average resource utilization during busy periods refers to the average network resource utilization rate within a specific time period in a preset area, including bandwidth utilization, channel utilization, etc. By monitoring these indicators, it is possible to understand whether network resources are sufficient to support current user demand and whether there is a resource shortage.

[0066] The aforementioned user feedback on communication quality refers to information provided by users through various means, including evaluations of call quality, data transmission speed, network coverage, and other aspects. Analyzing user feedback can help understand user satisfaction with network quality and identify potential problems in a timely manner.

[0067] Whether the current time period is considered a "busy period" can refer to whether the current time period is during a peak network activity period, such as daytime peak hours or specific event periods. By determining whether the current time period is a "busy period," the network load can be assessed, and timely measures can be taken to address potential problems.

[0068] In one optional embodiment, during the early warning process for wireless network clustering, the prediction result is determined to indicate that a preset area is trending towards wireless network clustering. To accurately predict the occurrence of wireless network clustering, target early warning parameters for multiple target wireless network elements can be monitored and analyzed to promptly detect anomalies and issue early warnings. Specifically, firstly, target early warning parameters for multiple target wireless network elements can be acquired. These parameters may include the number of users during busy periods, the average resource utilization rate during busy periods, user feedback on communication quality, and whether the current period is a busy period. These parameters reflect the load, resource utilization, and user experience of the wireless network and can serve as data for judging the trend of wireless network clustering. Next, the target early warning parameters can be monitored to determine whether they are greater than or equal to a preset early warning parameter threshold. When the similarity of the targets is greater than or equal to the set threshold, it indicates that there is a certain degree of similarity between multiple target wireless network elements, which may be simultaneously affected by some factor, leading to the occurrence of wireless network clustering. Furthermore, if the target early warning parameters are greater than or equal to the preset early warning parameter threshold, the prediction result is determined to be that a wireless network cluster will occur in the preset area at a preset future time, i.e., the system issues a wireless network clustering early warning notification. In the above process, by monitoring the early warning parameters of multiple target wireless network elements, we can gain a more comprehensive understanding of the operating status of the wireless network, promptly detect the risk of wireless cluster failures, and avoid false alarms by setting target similarity coefficients and target early warning parameters. This will improve the accuracy and reliability of early warnings, facilitate the timely issuance of wireless cluster failure early warning notifications, and enable maintenance personnel to respond quickly and take measures to reduce the impact of wireless network failures on users, thereby improving user experience and network quality.

[0069] In the above embodiments of this application, determining the target similarity of multiple target wireless network elements includes: obtaining the region type of a preset area, wherein the region type includes an indoor scene type and an outdoor scene type, the indoor scene type is used to represent an indoor scene without a backup battery to power the wireless network elements in the preset area, and the outdoor scene type is used to represent an outdoor scene with a backup battery to power the wireless network elements in the preset area; when the region type is an indoor scene type, performing station spacing normalization processing on multiple target wireless network elements based on a preset polygonal bounding box to obtain the target similarity; when the region type is an outdoor scene type, performing station spacing normalization processing on multiple target wireless network elements based on the station address hierarchy relationship between multiple target wireless network elements to obtain the target similarity.

[0070] In one optional embodiment, wireless cluster failure early warning can promptly detect and handle a large number of outages in a wireless network, thereby improving network stability and reliability. Obtaining the proximity of multiple target wireless network elements within a preset area helps operators more accurately determine areas where cluster failures may occur and take corresponding preventative measures. Specifically, for indoor scenarios, the distance between multiple target wireless network elements can be normalized using a preset polygonal bounding box. This process helps determine the distance between target wireless network elements and calculate their proximity. This method reveals that in the same indoor scenario, wireless network elements that are closer together are more likely to interfere with each other, thus helping to detect potential cluster failures in advance. For outdoor scenarios, distance normalization can be performed based on the hierarchical relationship between multiple target wireless network elements. In outdoor environments, the hierarchical relationship of wireless network elements is usually affected by factors such as terrain and buildings, and the distance and degree of mutual influence between different sites will also vary. By analyzing the hierarchical relationship, the proximity between different sites can be more accurately assessed, thus enabling targeted cluster failure early warning. In the above process, by assessing the proximity between target wireless network elements, areas prone to cluster failures can be more accurately identified, avoiding false alarms or missed alarms. For indoor scenarios, automated methods for normalizing station spacing can reduce the cost of manual intervention and improve the efficiency and real-time nature of early warnings. For outdoor scenarios, targeted early warning strategies can be developed based on the characteristics of different area types and station hierarchical relationships, improving the relevance and effectiveness of early warnings.

[0071] In the above embodiments of this application, generating wireless group fault early warning information based on network element attribute data includes: performing feature localization on multiple target wireless network elements based on network element attribute data to obtain a multi-level feature matrix, wherein the multi-level feature matrix is ​​used to represent the hierarchical attribute matrix of the multi-level structure to which the multiple target wireless network elements belong, and the multi-level structure includes at least radio frequency units, baseband units, and equipment rooms; performing fault analysis on multiple target wireless network elements based on the multi-level feature matrix to obtain fault cause localization results; and generating wireless group fault early warning information based on the fault cause localization results.

[0072] In one optional embodiment, early warning of wireless network cluster failures and fault analysis can be performed. When a large number of wireless network elements experience outages within the same geographical area, the cause of the failure can be quickly and accurately identified and repaired to ensure network stability and reliability. Specifically, firstly, feature localization can be performed on multiple target wireless network elements. Feature localization refers to extracting the characteristic information of each wireless network element by analyzing its performance parameters, alarm information, log records, etc., to facilitate subsequent fault location. During feature localization, the attributes of each level in the multi-level structure can be considered, including radio frequency units, baseband units, and equipment rooms. Secondly, based on the multi-level feature matrix obtained from feature localization, fault analysis can be performed on multiple target wireless network elements. In this process, the commonalities and differences between different wireless network elements can be identified by comparing their characteristic information. Through the analysis of this characteristic information, the possible location and scope of the fault cause can be further determined. Finally, based on the results of the fault analysis using the multi-level feature matrix, the fault cause of multiple target wireless network elements can be located. Through this process, the root cause of wireless network cluster failures can be accurately identified, and corresponding measures can be taken for repair and prevention. Meanwhile, this multi-level feature matrix-based fault analysis method can also help network operators better understand the entire network structure and improve the efficiency of fault location and handling. Through the above implementation process, the fault analysis method based on multi-level feature matrices enables network operators to better understand and manage the network structure, promptly identify and repair potential faults, thereby improving network stability and reliability. It can also improve the accuracy of fault location, accelerate fault handling, and reduce the impact of faults on network operation, thus improving the overall efficiency and reliability of network operation.

[0073] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0074] The Baseband Unit (BBU) is one of the core devices of a base station, responsible for baseband signal processing and base station control. The BBU can be located in a central equipment room or macro base station equipment room, and is used to process core network and user signaling data.

[0075] The Remote Radio Unit (RRU) is responsible for radio frequency (RF) processing, converting intermediate frequency (IF) signals into RF signals, and transmitting them through an antenna. Separating the RRU from the Base Unit (BBU) reduces signal attenuation and improves network efficiency.

[0076] An active antenna unit (AAU) integrates an RRU and an antenna, which improves network performance, achieves multi-channel primary transmission, reduces signal attenuation, and improves signal transmission efficiency.

[0077] An Application Programming Interface (API) is a predefined function used by developers to define communication methods between various functional components, enabling data interaction between users and developers.

[0078] A Hypertext Transfer Protocol Request (HTTP) is a request message from a client to a server. HTTP requests run at the application layer and include commonly used GET and POST requests.

[0079] Co-location relationship can refer to 4G and 5G RRUs and cells under the same wireless base station, where the latitude and longitude coordinates of these RRUs and cells are the same.

[0080] A polygonal frame can refer to viewing the wireless network coverage area as a polygonal region. The boundaries of this region are marked on a map using latitude and longitude coordinates. It is often used to mark enclosed areas such as university campuses, residential areas, industrial parks, large shopping malls, and scenic spots.

[0081] Wireless cluster failure refers to a situation where a large number of macro base stations or indoor distribution systems' RRUs simultaneously or successively fail to provide wireless service within a contiguous geographical area. Because wireless cluster failure involves numerous wireless network elements located in adjacent, contiguous geographical areas, when these elements fail simultaneously, users within the area cannot rely on their current or adjacent wireless network elements for wireless network service, significantly impacting wireless communication for users in this area. Within contiguous, adjacent geographical areas, regardless of whether it's a macro base station or an indoor distribution system, when a BBU fails, it causes its connected RRUs to fail, and RRUs to fail, leading to the failure of one or more logical cells configured to fail. When the number reaches a specified threshold, provincial and municipal telecommunications operators, as needed, determine that a wireless cluster failure has occurred.

[0082] The average resource utilization rate during busy hours, also known as the average PRB utilization rate, is derived from the fact that a Physical Resource Block (PRB) can be a basic channel resource unit in 4G and 5G networks, consisting of continuous subcarriers in the frequency domain and continuous orthogonal frequency division multiplexing (OFDM) symbols in the time domain. PRB utilization rate refers to the overall utilization rate calculated in both the time and frequency domains, typically referenced at the service access point between the Media Access Control (MAC) layer and the physical layer. The average PRB utilization rate during busy hours refers to the average PRB utilization rate of each 4G or 5G cell during its busiest period, and is closely related to the number of users and traffic throughput.

[0083] The technical solution proposed in this application will be described below with reference to an optional embodiment. This application proposes a system and method for automatic early warning of wireless group failures and automatic analysis and location of problematic links. This application belongs to the field of wireless communication, and specifically relates to a system and method for early warning of wireless group failure risks, and can automatically analyze and locate problematic links and propose solution suggestions.

[0084] Currently, the scale of 4G and 5G wireless networks is constantly expanding, and the number of users and types of services served are continuously increasing, which increases the complexity of operators' maintenance work, making maintenance extremely challenging. Once a wireless cluster failure occurs in a certain geographical area, a large number of users and services in that area will be unable to use the network normally, requiring wireless network maintenance personnel to restore the network as quickly as possible. At present, the network management platforms of various wireless equipment manufacturers only push alarms to each device independently. Some third-party platforms, after connecting to the network management interface, have the function of nearly real-time statistics of the total number of alarms across the network, but it is difficult to automatically determine whether a cluster failure has occurred. Multiple wireless failures occurring in different geographical areas within a similar time period do not constitute a wireless cluster failure. Currently, it is necessary to rely on the manual experience of wireless network maintenance personnel to determine whether multiple wireless network elements that have gone out of service cover a contiguous geographical area based on the geographical adjacency of the coverage area, and to combine this with whether the total number of related network elements that have gone out of service has reached a threshold, in order to determine whether a wireless cluster failure has occurred. Furthermore, it is difficult for the network management platforms of various wireless equipment manufacturers and various auxiliary platforms, as well as wireless network maintenance personnel, to detect potential wireless cluster failures in time before they occur, making timely warnings difficult. Meanwhile, the network management platforms and auxiliary platforms of various wireless equipment manufacturers push alarm information to wireless network maintenance personnel, but only provide limited information such as the number of alarms, the name of the alarm network element, and the alarm type. This makes it difficult to accurately pinpoint which network element(s) at different network levels are the main cause of the wireless cluster failure alarm. Moreover, once a wireless cluster failure occurs, various alarms, including main alarms, derivative alarms, and alarms outside the affected area, appear in the alarm system within a short period of time. Different alarm handling procedures apply to each type of alarm, making the location of a large number of alarms cumbersome, time-consuming, and labor-intensive. This hinders wireless network maintenance personnel from quickly and effectively identifying the main problem causing the wireless cluster failure alarm. Furthermore, many provincial and municipal wireless operators simultaneously run multiple network management systems from different manufacturers, requiring an administrator to switch between multiple systems concurrently. This further increases the complexity of alarm handling and poses a significant challenge to wireless network maintenance personnel in promptly addressing wireless cluster failures.

[0085] This application proposes a novel system for early warning and problem localization of wireless network element outages. Based on this new system, a novel method is used to determine the geographical proximity of multiple wireless alarms related to wireless network element outages generated within a similar time period to ascertain whether adjacent physical areas are covered. Then, when the proximity is high, the new system automatically increases the monitoring frequency of wireless network elements in the relevant area, promptly identifying new alarms in the area. The new method then determines whether the wireless network elements corresponding to these alarms are at risk of wireless network element outages, providing timely early warnings to wireless network maintenance personnel. Next, the new method locates the main problematic links leading to the wireless network element outage risk, and the results are promptly communicated to wireless network maintenance personnel to assist in quickly resolving the problem. Through timely early warning and alerts for wireless network element outages, and by assisting in quickly locating the main problematic links related to these alarms, wireless network maintenance personnel can take proactive intervention measures to prevent the occurrence of wireless network element outages or avoid further expansion of their impact, thereby effectively reducing the impact of wireless network element outages on users' wireless service experience.

[0086] This application proposes a system and method for automatic early warning and automatic analysis and location of wireless cluster faults. It can automatically warn of wireless cluster faults and quickly analyze and locate the problem links in cases of existing cluster faults. In wireless network operation and maintenance, the root causes of wireless cluster faults are mainly power supply issues, such as large-scale mains power outages or abnormal power supply to the main power lines of densely concentrated wireless network elements, or fiber optic transmission issues, such as faults in the main optical cables of macro base stations or indoor distributed systems, faults in related optical cables within centralized BBU deployment rooms, or faults in data transmission equipment connected to centralized BBUs. This application, in scenarios with high wireless network element density such as university campuses, residential areas, industrial parks, and large shopping malls, promptly acquires outage-related alarm information and assesses the similarity of wireless network elements to provide timely early warning or alarms for wireless cluster faults. For wireless macro base stations used for large-area outdoor wireless coverage, this application, by promptly acquiring low-voltage alarm information from the wireless macro base station's battery and combining this with the topological relationship information of adjacent locations of multiple alarm sites, predicts the risk of wireless cluster faults in advance. Meanwhile, this application can automatically locate alarm problems based on data such as the topology of network elements at all levels, alarm types, alarm network elements, and quantities, and promptly recommend handling solutions. By timely reminding wireless network maintenance administrators to intervene in the risk of wireless cluster failures, the scale of alarms can be prevented from further escalating.

[0087] The topology of this application system can be as follows. According to the technical solution of this application, four new modules are deployed: a wireless network management platform (hereinafter referred to as "network management"); a wireless network element equipment basic data platform (hereinafter referred to as "basic data platform"); and a base station equipment power and environmental data monitoring platform (hereinafter referred to as "environmental monitoring platform"), forming the system of this technical solution. The four newly deployed modules are: a system communication bus module, an alarm data processing module, an alarm data analysis module, and a notification management module. The main functions of the system communication bus module are: to obtain wireless network element alarm information from the network management platform; to obtain power equipment and line alarm information from the environmental monitoring platform; and to obtain network data at all levels related to various alarm information. The network element level includes RRU, BBU communication cards at all levels, BBU, RRU uplink aggregation equipment, BBU rack, A device (i.e., BBU uplink wireless data transmission equipment), and B device (i.e., A device uplink aggregation equipment). It obtains data from the wireless network management platform from RRU, BBU communication cards at all levels, and BBU, as well as data from the basic data platform's BBU rack, A device, B device, and wireless site parameter data. The acquired alarm data and related data of network elements at all levels are cached in the "data cache submodule" of the "alarm data processing module" to realize data interaction between modules of this system.

[0088] The alarm data processing module includes: an alarm polling submodule, a data caching submodule, an alarm data parsing submodule, and an alarm data processing submodule. Its main function is to obtain key alarm information of wireless network elements through a polling mechanism, i.e., key alarm information related to wireless group failures, and then parse and process it. The main functions of each submodule are as follows: The alarm polling submodule performs flexible periodic alarm polling on wireless network elements to determine whether the group failure warning threshold has been reached. The data caching submodule temporarily caches alarm data and related basic data of wireless network elements at all levels, facilitating quick retrieval by other submodules for processing and analysis. The alarm data parsing submodule parses the format of the acquired alarm data, enhancing the readability of the alarm data in this application system. The alarm data processing submodule performs format conversion processing on the parsed alarm data according to the system requirements of this application, facilitating subsequent data analysis. The alarm data analysis module includes: a group failure warning analysis submodule, a transmission layer equipment fault analysis submodule, a BBU fault analysis submodule, and an RRU fault analysis submodule. Among them: the group failure early warning and analysis submodule analyzes the group failure risk of the received alarm data information; after determining that there is a group failure risk, the BBU failure analysis submodule, RRU failure analysis submodule and other submodules are started to analyze the alarm cause of the corresponding network element at the network level.

[0089] The notification management module serves as an external notification tool, sending group failure early warning information or analysis results of group failure alarms to the wireless network maintenance administrator via group robots, SMS, and email, notifying relevant personnel to follow up and handle the issues. Regarding the alarm data interface principle, there are two main ways to obtain alarm data from the wireless network management platform: the first is to obtain it on demand from the northbound interface provided by the network management elements to the wireless communication operator; the second is to access the network management system and export alarm data from the network management alarm monitoring page. The second method is more complex, requiring manual access to the network management system through an account, and access is limited to designated IP addresses and trusted users, making it unsuitable for the system described in this patent. Furthermore, the industry currently advocates decoupling between components to reduce the complexity of software system implementation, while achieving better security and platform maintainability. The system in this application preferably adopts the first solution. The specific implementation details of the network management northbound interface will be embedded within the platform, with no external data exposure, and users will not be able to access the data generated during the interaction between the system and the network management system. Specifically, the network management system will provide an application interface for this system. Interaction between the network management system and the system will be achieved through this interface. Communication via this interface is bidirectional; only the system can initiate a request first, sending authentication information to the interface. Once authentication is successful, a "bridge" of mutual trust is established between the network management system and the system. Subsequently, the system sends a request body with a specific structure to the interface, containing the data the system wants to obtain from the network management system. For alarm data, this mainly includes "region," "time," "alarm code," and "alarm type." After receiving the request from the system, the network management interface first parses the request to ensure its validity and correct format. If the request body meets the requirements, the network management system will provide the corresponding data through the interface. If the requested data is invalid or incorrectly formatted, the network management system will also provide an error message. For security reasons, the interface has a single function; requests and responses between the network management system and this system can only be implemented through interfaces with specific functions, and no single interface will perform multiple functions. The alarm data acquisition and processing principle involves periodically polling the network management and environmental monitoring platforms for outage and power supply alarms through the "alarm polling submodule" in the "alarm data processing module".

[0090] The system employs a flexible periodic polling mechanism, for example, a default polling interval of 5 minutes, based on the first polling cycle, periodically requesting relevant alarm data from the network management system. When no relevant alarm data is available, the alarm data processing module, alarm data analysis module, and notification management module remain silent, without consuming additional system resources. Upon receiving alarm data, and when the alarm network element obtained by the "Group Fault Early Warning Analysis" submodule of the "Alarm Data Analysis Module" satisfies either the "Comprehensive Similarity Coefficient α" for Type 1 or the "Comprehensive Similarity Coefficient β" for Type 2, the polling submodule is notified to modify the polling cycle, changing the polling interval to 5 seconds, based on the second polling cycle. This can be configured as needed, continuously sending alarm information polling requests to the network management system at the new frequency. The polling target is all wireless network elements within a 3km radius of the aforementioned alarm network element, which can be configured as needed. The duration is set to 30 minutes, which can be configured as needed, or "α < α0" and "β < β0".

[0091] The alarm data parsing principle involves obtaining alarm data through a polling mechanism. Since the alarm data formats of the network management platform and the environmental monitoring platform differ and are inconsistent with the format required for subsequent analysis by this system, the "Alarm Data Parsing" submodule decodes the acquired data, verifies its integrity, and reassembles it according to the format required for subsequent system analysis. During the reassembly process, alarms are categorized into three types based on the physical structure of the wireless network elements and the logical architecture of the wireless network: transmission equipment level alarms, BBU level alarms, and RRU level alarms. The alarm data processing principle involves the "Alarm Data Processing" submodule extracting important readable information from the parsed alarm data, including alarm information related to equipment outages, such as cell unavailability alarms, low battery voltage alarms, AC power failure alarms, the administrative region, and the equipment involved in the alarm. The alarm data analysis module first statistically analyzes the number of major alarm codes (representing alarm types) related to wireless device outages to determine whether a wireless group failure has occurred and to predict the risk of such a failure. Once the conditions are met, the module promptly notifies the wireless network maintenance administrator. Then, based on the alarm codes and the topology of the wireless network devices, the module automatically analyzes and locates the main factors causing the wireless group failure, or the main factors leading to the wireless group failure warning, and promptly notifies the wireless network maintenance administrator of these factors.

[0092] This application employs a proximity judgment algorithm to determine the proximity of wireless network sites. Wireless cluster failure refers to a severe fault where a large number of wireless network elements serving a continuous geographical area go out of service. Therefore, a reasonable and efficient algorithm is needed to quickly determine whether the home sites of multiple wireless RRUs associated with outage alarms are located within a continuous geographical area. This application uses a "proximity judgment algorithm" to achieve this purpose. Currently, wireless cluster failures with a relatively high probability of occurrence and a relatively large service impact range in the existing network mainly fall into two types: The first type is a typical scenario where high density of wireless devices is deployed due to the construction of wireless indoor distribution systems in university campuses, industrial parks, large shopping malls and supermarkets, and residential areas, hereinafter referred to as "Cluster Failure Type 1". In this scenario, wireless RRUs generally lack battery backup. The second type is where nearby macro stations within a region go out of service successively within a short period of time, hereinafter referred to as "Cluster Failure Type 2". The AAU / RRUs of macro stations generally have battery backup. All RRUs are classified according to their cluster failure type, belonging to either Type 1 or Type 2. Two different types of network clusters are assessed using different "similarity judgment algorithms": for "Cluster Fault Type 1," a "polygon bounding box-based inter-station spacing normalization algorithm" is used; for "Cluster Fault Type 2," an "inter-station spacing normalization algorithm based on station address hierarchy" is used. Specifically, in indoor scenarios, the inter-station spacing of multiple target wireless network elements is normalized based on a preset polygon bounding box to determine their similarity; in outdoor scenarios, the inter-station spacing of multiple target wireless network elements is normalized based on their station address hierarchy to determine their similarity.

[0093] This algorithm, based on the similarity normalization of polygon bounding boxes, is suitable for "Type 1 obstacle clusters." A polygon bounding box uses latitude and longitude coordinates to mark the boundary of a closed region, such as a university campus, residential area, industrial park, or large shopping mall. Let the set of latitude and longitude coordinates of a polygon bounding box for a certain region, denoted as pol, be:

[0094] pol=((lon0,lat0),(lon1,lat1),…,(lon n ,lat n ));

[0095] lon1 and lat1 represent the longitude and latitude information of the first coordinate point, respectively; lon2 and lat2 represent the longitude and latitude information of the second coordinate point, respectively. n and lat nThese represent the longitude and latitude information of the nth coordinate point, respectively. In actual drawing, only the vertices of the polygonal region need to be marked. The polygonal frame can be drawn by connecting the longitude and latitude coordinate points on the visualization map. There are several methods to calculate the center point of the polygonal frame. It can be calculated using the geospatial center point or the shortest distance center point. The geospatial center point is calculated by converting the latitude and longitude spatial coordinate system to a planar coordinate system, and then converting the result back to latitude and longitude coordinates; the center point is located at the center of the polygon. The shortest distance center point represents the point where the sum of distances from the center point to multiple surrounding device coordinate points within the polygonal frame is the shortest; the center point is not necessarily located at the center of the polygon. The calculated coordinates of the polygonal frame center point are pol. center It can be represented as:

[0096] pol center =(lon) center ,lat center );

[0097] After calculating the center point of the polygon frame, the coordinates of all devices within that polygon frame are replaced with the coordinates of that center point. Each device within each polygon frame of the entire wireless network is processed in the same way, resulting in a table of coordinate information for all RRUs in the entire wireless network.

[0098] (RRU1,RRU2,…,RRU n );

[0099] The latitude and longitude of the first RRU is RRU1 = (lon1, lat1), the latitude and longitude of the second RRU is RRU2 = (lon2, lat2), and the latitude and longitude of the nth RRU is RRU1 = (lon1, lat1). n =(lon) n ,lat n ).

[0100] Next, the shortest distance RRU is calculated based on latitude and longitude coordinates. For all RRUs within all polygons, a search algorithm is used to match the distance between two RRU devices one by one. To avoid devices with the same site or latitude and longitude participating in the calculation, which could lead to different sites and polygons not being associated, the calculation process not only saves the device with a shortest distance of 0, but also the first RRU with a distance greater than 0. For example: RRU1 and (RRU2, ..., RRU... n The calculation yields the results D(RRU1,RRU2),...,D(RRU1,RRU2). n D(·) represents the latitude and longitude distance function. Assume RRU1 and RRU i If they are at the same station address or have the same polygonal bounding box, then D(RRU1,RRU) i ) = 0, RRU iThe shortest distance relationship is between RRU1 and RRU2; additionally, among all distance pairs greater than 0, D(RRU1, RRU2) is the shortest distance relationship. j If the minimum value is found, then the RRU is minimized. j It also has the shortest distance relationship with RRU1. After the first search, (RRU1, RRU1) will be... i ) and (RRU1,RRU j All RRUs are added to the table. Note the positional relationship: the first RRU is the original RRU, and the second is the target RRU, indicating that the RRU with the shortest distance to the original RRU is the target RRU. Continue the search process: RRU2 and (RRU1, RRU3, ..., RRU...) n (Comparison, RRU) n and (RRU1,…,RRU) n-1 Compare and find the RRUs whose shortest distance is equal to 0 (if it exists) and whose shortest distance is greater than 0.

[0101] After all RRUs have been searched, a table of neighboring RRUs with the shortest distance is obtained: [RRU] i RRU j ] represents matrix RRU i and RRU j This is the adjacent RRU matrix, where RRUs with the shortest distance are considered adjacent pairs. RRUs with a shortest distance of 0 are considered adjacent pairs, as are RRUs with a shortest distance greater than 0. The adjacent RRU table is then updated to include the RRUs in the matrix. i RRU with non-zero shortest distance j Using the distance between them as the reference denominator, the RRU is... i The remaining RRUs in the adjacent RRU matrix and RRU i The distances between them are normalized. For example, the distances between RRU1 and RRU2, RRU3, and RRU4 are 260 meters, 200 meters, and 300 meters, respectively. Among them, RRU3 has the shortest distance between RRU1 and other adjacent RRUs, so 200 meters is used as the normalization denominator. After normalization, the similarity between RRU1 and RRU2, RRU3, and RRU4 are 1.3, 1, and 1.5, respectively. The upper limit of the similarity can be set as needed, for example, to 3. Then, for each RRU within the polygon frame, a subset of adjacent RRUs is established, and further, subsets N1 and N2 of similarity are established. Where: N1 is the set of adjacent RRUs with a similarity of 1 for each RRU within the polygon frame, N1 = {n 11 ,n 21 ,…,n m1}, n m1Let N1 represent the subset of neighboring RRUs with a similarity of 1 to RRUm, such as {RRU1, RRU3} in the example above; N2 is the set of neighboring RRU subsets within the polygon frame that have a similarity in the interval (1, 3] for each RRU, N2 = {n 12 ,n 22 ,…,n m2}, n m2 This represents a subset of adjacent RRUs whose similarity to RRUm is in the interval (1, 3], such as {RRU1, RRU2} and {RRU1, RRU4} in the example above. For all RRUs in all polygon boxes, following the above method, first establish a subset of adjacent RRUs, and then establish subsets of similarity N1 and N2. Summarize N1 and N2 of all polygon boxes to form a "normalized set of adjacent RRUs" N, N = {N1, N2}. When two RRUs belonging to the "Group Fault Type 1" scenario belong to a subset of adjacent RRUs of N1 or N2, it means that these two RRUs have a "similarity of 1" or "similarity in the interval (1, 3]", respectively.

[0102] Based on the similarity normalization algorithm of site location hierarchy, this algorithm is applicable to "Type II of Group Fault". Unlike the relatively closed environment of Type I of Group Fault, the macro base station environment of Type II is relatively open. When a base station RRU in a certain location goes out of service, the original coverage area of ​​that base station may still receive wireless signals from adjacent first or second layer stations. For example: there are 3 macro base stations within 500 meters of a certain macro base station, and there are also 3 macro base stations within 500 to 1000 meters. The ones within 500 meters are considered first layer stations, and the macro base stations within 500 to 1000 meters are considered second layer stations. When this macro base station goes out of service, wireless users in the original coverage area of ​​this macro base station will most likely be able to receive wireless signals from the surrounding first layer stations. Even if all the first layer stations go out of service, users may receive wireless signals from the second layer stations, and wireless communication can still be barely maintained. This algorithm is used to determine the similarity of RRUs among multiple macro stations that have been deactivated or are at risk of deactivation. The specific algorithm is as follows: Each macro station in the entire network is uniquely numbered, and the set of all macro stations is represented as: (Base1, Base2, ..., Base...). n ), where the latitude and longitude of macro station number n is Base n =(lon) n ,lat n Using a search algorithm, the latitude and longitude distances between two macro stations within a specified range (e.g., all macro stations within a 1500-meter radius of macro station Base1) are matched. The distances from Base1 to all surrounding base stations within a 1500-meter radius are then obtained as: D(Base1,Base2), ..., D(Base1,Base2). m D(·) represents the latitude and longitude distance function. Similarly, we obtain Base2, Base3, ..., Base...n The distances to all surrounding macrostations within a 1500-meter radius of each macrostation are considered. To distinguish whether a given macrostation is a first-level or second-level station compared to the other macrostations within its 1500-meter radius, the distances need to be normalized according to the first- and second-level station distances. That is, the empirical average distance between first- and second-level stations is used as the reference denominator, and the actual distance between the two macrostations is used as the reference numerator to normalize the distances between them. The distance normalization function for a first-level station is:

[0103]

[0104] The above formula represents the normalized function of the inter-station spacing between macro stations i and j, where d1 is the inter-station spacing.

[0105] The normalized function of the two-level station is expressed as:

[0106]

[0107] The above formula represents the normalized function of the distance between the two-level stations i and j, where d2 is the distance between the two-level stations.

[0108] For macro site Base i and Base j :if and If all are within the interval (0,1], then Base j It is Base i The first floor station; if and If it is located in the interval (0,1], then Base j It is Base i The second-level stations. For example: assuming macro stations within a 500-meter radius are first-level stations, and those between 500 and 1000 meters are second-level stations. The distance from Base1 to Base2 is 400 meters, and the distance from Base1 to Base3 is 800 meters. The normalized function for the spacing between first-level stations can be calculated. The normalization function for the distance between the two stations is: Therefore, Base2 is a first-level station of Base1, and Base3 is a second-level station of Base1. For each macro station and surrounding macro stations within a radius of 1500 meters (set as needed), the above algorithm is used to obtain the first-level station subset and the second-level station subset for each macro station. All first-level station subsets are aggregated to form the first-level station set S1, S1 = {s 11 ,s 21 ,…,s n1}, s n1 This refers to the first-level station subset of the macro station numbered n. All second-level station subsets are aggregated to form the second-level station set S2, where S2 = {s...}12 ,s 22 ,…,s n2}, s n2 This refers to the second-level subset of macro stations numbered n. S1 and S2 are combined to form the macro station address hierarchy set S, where S = {S1, S2}. When two macro stations that trigger outage-related alarms belong to subset S1 or S2, it indicates that the hierarchical relationship between these two macro stations is either a first-level station or a second-level station. By introducing the "normalized adjacent RRU set" N and the "macro station address hierarchy set" S, the system avoids having to perform calculations every time when determining the proximity of radio stations involved in outage-related alarms, thus speeding up the judgment process. After the "normalized adjacent RRU set" N and the "macro station address hierarchy set" S are initially established, they need to be updated in a timely manner as new radio devices at new sites are added to the network, or as existing radio devices at new sites are gradually dismantled.

[0109] This wireless group failure early warning algorithm, based on the "comprehensive similarity coefficient" and "early warning coefficient" (i.e., target similarity and target early warning coefficient), addresses the issue that a large number of users' wireless communications will be significantly affected once a group failure occurs, leading to numerous wireless network quality complaints at the time or afterward. This application uses the scale of network elements involved in the wireless group failure, factors related to their similarity, and factors related to the severity of wireless user complaints as references to set a threshold for wireless group failure early warning. The system automatically determines whether the real-time values ​​of the "comprehensive similarity coefficient" and "early warning coefficient" exceed the early warning trigger threshold. For Type 1 wireless cluster faults, the potential severity of wireless complaints is related to the following parameters: the number of out-of-service RRUs or logical wireless cells (hereinafter referred to as "out-of-service network elements"); the proximity factor between the out-of-service RRU and the nearest neighboring RRU with the alarm (hereinafter referred to as "proximity factor"); the number of users under the out-of-service RRU or logical wireless cell during service busy hours (hereinafter referred to as "busy hour user number factor"); whether it is within the service busy hour period (hereinafter referred to as "time period factor"); the average resource utilization rate of the logical wireless cell during busy hours (hereinafter referred to as "load factor"); and the complaint tendency of the wireless site scenario (hereinafter referred to as "scenario factor"). For Type 2 wireless cluster faults, the potential severity of wireless complaints is related to the following parameters: the number of out-of-service macro base stations (hereinafter referred to as "out-of-service macro base stations"); the hierarchical relationship between the multiple out-of-service macro base stations (hereinafter referred to as "hierarchical factor"); the number of users under the out-of-service macro base station during service busy hours (hereinafter referred to as "busy hour user number factor"); whether it is within the service busy hour period; the average resource utilization rate of the logical wireless cell during busy hours; and the complaint tendency of the main scenarios covered by the wireless site (hereinafter referred to as "scenario factor").

[0110] The descriptions and values ​​of the above factors are as follows: Number of deactivated network elements: n net Similarity factor: ne rru={1,0.7,0.5}, similarity = 1, value 1; similarity in (1,1.5], value 0.7; similarity > 1.5, value 0.5, can be adjusted as needed; User count factor during busy time: n user ={1,0.5,0.4}, which represents the number of users during the self-busy period. If the number of users during the self-busy period is ≥100, the value is 1; if the number of users during the self-busy period is ≤50 and <100, the value is 0.8; if the number of users during the self-busy period is <50, the value is 0.4. This value can be adjusted as needed. Time period factor: t busy ={1,0.5}, which indicates whether the current time period is a self-busy period. A value of 1 is used when the time is self-busy, and a value of 0.5 is used when it is not self-busy. This value can be adjusted as needed. Load factor: p load ={1,0.6,0.3}, which represents the average resource utilization rate during peak hours. If the average PRB during peak hours is ≥70%, the value is 1; if 30% ≤ average PRB < 70%, the value is 0.6; if the average PRB < 30%, the value is 0.3. This value can be adjusted as needed. Scenario factor, td scene ={1, 0.7, 0.4}, which represents user feedback on communication quality. This is based on statistics from the entire year's complaint data. The top three scenarios with the most wireless complaints are assigned a value of 1; scenarios with the 4th to 6th highest number of complaints are assigned a value of 0.7; and the rest are assigned a value of 0.4. This value can be adjusted as needed. Number of macro base stations out of service: n base Hierarchical factor: r gap ={1,0.6}, where two macro stations are in a single-level station relationship and the value is 1; and two macro stations are in a two-level station relationship and the value is 0.6. Using the above factors, this application proposes two different types of early warning coefficient algorithms for wireless group faults. When the early warning coefficient exceeds the set threshold, the system issues a wireless group fault early warning notification.

[0111] In this application, based on the different regional types of the preset region, the similarity of the target can be represented as α or β, the target warning parameter can be represented as A or B, the first similarity threshold (α1 or β1), the second similarity threshold (α0 or β0), and the preset warning parameter threshold (A1 or B1).

[0112] The warning coefficient algorithm for Type 1 wireless obstacle clustering, corresponding to Type 1 wireless obstacle clustering, i.e., indoor scene type, can be expressed as:

[0113]

[0114] For wireless group faults of type one, the "comprehensive similarity coefficient" α, which comprehensively characterizes the number and similarity of out-of-service alarm RRUs, can be expressed as:

[0115]

[0116] The "Comprehensive Similarity Coefficient" α determines whether to increase the polling frequency of the alarm polling submodule and whether to issue a wireless group failure warning, based on the number of outage alarms and the similarity of the involved sites. The "Warning Coefficient" A, in addition to considering the number of outage alarms and the similarity of the involved sites, also considers other weighted factors related to the severity of wireless user complaints, namely the number of users during busy hours, time period, load, and scenario factors, to determine whether to issue a wireless group failure warning from another dimension. When multiple RRUs of type 1 fail out of service, the "Group Fault Early Warning Analysis" submodule calculates the "Comprehensive Similarity Coefficient" α and the "Early Warning Coefficient" A. Specifically: when "α < α0", the polling period of the "Alarm Polling" submodule remains unchanged, and the fault analysis submodules of various network elements, such as the "BBU Fault Analysis" submodule and the "RRU Fault Analysis" submodule, do not need to be started for fault analysis; when "α ≥ α0", the "Group Fault Early Warning Analysis" submodule feeds back information to the "Alarm Polling" submodule, notifying it to modify the polling period and increase the polling frequency, and the fault analysis submodule starts fault analysis; when "α ≥ α1", note that α1 > α0, or the early warning coefficient A satisfies "A ≥ A1", the "Group Fault Early Warning Analysis" submodule feeds back information to the "Notification Management" module, which then sends wireless group fault early warning information to the wireless network maintenance administrator. The early warning information includes key information such as the site name, RRU name, uplink BBU name, and the main possible causes of the wireless group fault early warning.

[0117] The warning coefficient algorithm for Type 2 wireless obstacle clustering, corresponding to Type 2 wireless obstacle clustering, i.e., outdoor scenario type, has the following warning coefficient B:

[0118]

[0119] For the corresponding type 2 wireless group fault, the "comprehensive similarity coefficient" β, which comprehensively characterizes the number of macrocell low voltage alarms and the similarity of the macrocells involved, is:

[0120]

[0121] The "Comprehensive Similarity Coefficient" β determines whether to increase the polling frequency of the alarm polling submodule and whether to issue a wireless group failure warning, based on the number of macro base station low voltage alarms and the inter-base station levels involved. The "Warning Coefficient" B, in addition to considering the number of macro base station low voltage alarms and the inter-base station levels involved, also considers other weighted factors related to the severity of wireless user complaints, namely, the number of users during busy hours, time period, load, and scenario factors, thus determining whether to issue a wireless group failure warning from another dimension. When "β < β0", the polling period of the "Alarm Polling" submodule remains unchanged. At the same time, the fault analysis submodules of various network elements, such as the "BBU Fault Analysis" submodule and the "RRU Fault Analysis" submodule, do not need to be started for fault analysis. When "β ≥ β0", the "Group Fault Early Warning Analysis" submodule feeds back information to the "Alarm Polling" submodule, notifying it to modify the polling period and increase the polling frequency. At the same time, the fault analysis submodule starts to perform fault analysis. When "β ≥ β1", note that β1 > β0, or the early warning coefficient B satisfies "B ≥ B1", the "Group Fault Early Warning Analysis" submodule feeds back information to the "Notification Management" module, which then sends wireless group fault early warning information to the wireless network maintenance administrator. The early warning information includes key information such as the macro base station name, the uplink BBU name, and the main possible causes of the wireless group fault early warning.

[0122] This section introduces multi-level early warning thresholds: a first similarity threshold (α1 and β1), a second similarity threshold (α0 and β0), and preset early warning parameter thresholds (A1 and B1). This tiered approach helps reduce false alarms and decreases the operational load on fault analysis submodules at various network element levels, such as the "BBU Fault Analysis" and "RRU Fault Analysis" submodules, as well as the "Alarm Polling" submodule. As the number of out-of-service RRUs or macrocell low-voltage alarms increases, and given the differences in the weights of various factors for each wireless network element, the "Comprehensive Similarity Coefficient" or "Early Warning Coefficient" increases at different rates. When either reaches a preset value, the system immediately issues a wireless group fault early warning notification.

[0123] This application presents an algorithm for locating and analyzing alarms related to wireless network element outages. Among all levels of wireless network elements, the Remote Root Unit (RRU) is the network element closest to the wireless user. This application analyzes and locates the relevant links affecting the wireless service capability of the RRU. The fault links that can cause RRU outages extend from the RRU network element to its upstream network element links, mainly including: RRU, BBU, and transport layer equipment. Specifically: the RRU link includes optical path faults, power supply faults, etc.; the BBU link includes communication board faults, optical path faults, power supply faults, etc.; and the transport layer equipment includes equipment port faults, optical path faults, power supply faults, etc. The degree of impact on wireless group failures, from high to low, is: transport layer equipment → BBU → RRU. Assuming an RRU failure, only RRU alarms will be triggered, but not BBU alarms. Except for wireless group failures caused by mains power outages or trunk optical cable interruptions, the possibility of a large number of RRU devices failing simultaneously is very low. In this case, analyzing only this level of RRU is insufficient to find the true cause of the fault; it is also necessary to combine analysis with upstream equipment, namely BBU and transport layer equipment. However, a failure in the upstream equipment of the RRU can cause a large number of RRUs to fail simultaneously. Therefore, after determining the type of group failure, the cause of the group failure should be located based on the characteristics of that type of group failure.

[0124] This section introduces the RRU feature localization method. Features refer to RRU hierarchical attributes; for example, by specifying the data center, rack, BBU, or board, the target RRU can be located. When the data center, rack, BBU, or board malfunctions, it will trigger an alarm in the RRU. A one-dimensional feature matrix is ​​established for each RRU: RRU i = [board number, BBU number, rack number, data center number], then all RRUs are: RRU = [RRU1, RRU2, ..., RRU] N ] T Finding the maximum number of identical elements in each column of the RRUs involved in the wireless group failure warning system yields the following result: RRU num =[Num 板卡 Num BBU Num 机架 Num 机房 Then find RRU. num The category corresponding to the most elements in the list, which is cat = argmax(RRU) num This category is for locating the cause of RRU group failures. For example, a wireless group failure warning involves multiple target RRU alarms: RRU1 = [1,0,1,1], RRU2 = [1,0,1,1], RRU3 = [1,0,2,1], RRU = [RRU1,RRU2,RRU3]. T RRU num =[2,1,2,3],argmax(RRU) numThe result is categorized as "Data Center," therefore these RRU alarms share a common characteristic: they all belong to the same data center. A fault in a certain component within this data center causes multiple target RRU alarms. Further on-site fault analysis and localization are then performed on the data center. Similarly, a BBU feature matrix can be established: BBU i =[Chain Link Number, B Device Number, A Device Number, Office Number, Chassis Number]. By analyzing these attributes, the problematic link can be quickly located. After completing data analysis and integration, the data analysis module packages the analysis results, including: analysis files, statistical files, and alarm notification forms. The analysis and statistical files contain alarm network element location information, cause analysis, and operation guidelines. The alarm notification form contains information such as alarm occurrence time, alarm type, number of alarms, assessment threshold, and repair time.

[0125] The notification principle for early warning information involves the alarm data analysis module integrating, statistically analyzing, and processing the alarm information provided by the network management system. The resulting data is then transmitted to the notification management module via the communication bus. The data analysis results include: analysis files, statistical files, and alarm notification forms. The analysis and statistical files contain alarm element location information, cause analysis, and operation guidelines. The alarm notification forms include information such as alarm occurrence time, alarm type, number of alarms, and assessment thresholds. These notifications are sent to designated wireless network maintenance personnel via the group robot interface and SMS interface to track and handle the issue, ensuring timely understanding of the scale of the fault and its progress. Furthermore, analysis and statistical reports are sent to a designated email address. Upon receiving the alarm notification form, wireless network maintenance personnel can promptly log in to their email to view detailed alarm information, analysis reports, and fault handling suggestions.

[0126] This application proposes a triple security and reliability guarantee mechanism. The system comprises several modules: a network management platform, an environmental monitoring platform, an alarm data processing module, an alarm data analysis module, and a notification management module. Alarm data is transmitted along a data bus, and each module has a clearly defined function and does not overlap with others. Guarantee Mechanism One: The system adopts an interface development approach. This application applies the concept of low coupling in current software development to wireless alarm monitoring development. Through interfaces, the functional components are decoupled, improving software reusability and avoiding the problems of unclear functions and redundant, chaotic code in traditional development. Simultaneously, it significantly reduces the difficulty of later code maintenance and version iteration. Some local networks run multiple network management systems from different manufacturers simultaneously, requiring an administrator to manage multiple systems, which is quite difficult. This application's new interface development model also reduces the difficulty of handling alarms from multiple network management systems. Security Mechanism Two: An independent authentication mechanism using interface development allows specific information to be sent only to specific modules. Each module has its own independent authentication mechanism, requiring verification before sending a request. All links and requests are encrypted twice. The system runs in the cloud, and maintenance personnel can only see pushed alarm-related information, but cannot access core data or understand the relationships between modules. This ensures the confidentiality of critical information. Security Mechanism Three: Simple configuration and easy to learn. Before use, the system requires only a small amount of input parameter data such as city, email, and group robot ID. After the software starts running, maintenance personnel do not need to be on duty. The system has a fault tolerance and error correction mechanism that automatically checks the results of each run. If a module experiences an error, the error correction mechanism is automatically activated, and the data stream bus sends a request to that module again. If the module still cannot provide a correct response after a period of time, the system will directly intervene, stop running, and notify the system maintenance personnel. The results of each run are stored in the cloud for easy retrospective analysis.

[0127] This application proposes a similarity normalization calculation method based on polygon bounding boxes. For each wireless network element within the polygon bounding box, the straight-line distance between each element and its nearest neighbor is used as the denominator for normalization, while the straight-line distance between that element and all other network elements within the polygon bounding box is used as the numerator. The resulting value is used to measure the similarity between that element and all other network elements within the box. The key point of this technology is to express the subjective judgment ability of humans in determining "a certain area of ​​wireless network elements successively generating outage alarms" through digital quantification. This technology can assist in quickly and automatically determining the risk of wireless cluster failure due to a series of outage-related alarms generated by wireless network elements within a polygon bounding box within a short period of time. When using this technology to determine the similarity of wireless sites involved in outage-related alarms within a polygon bounding box, it avoids the need for calculations every time, thus speeding up the judgment process.

[0128] This application proposes a similarity normalization calculation method based on site hierarchy. This technique uses empirical data on the distance between first-level and second-level macrocells as the denominator and the actual straight-line distance between macrocells as the numerator. The resulting value is used to measure the similarity and site hierarchy relationship between macrocells involved in low battery voltage alarms. This technique expresses the subjective judgment ability of humans in determining the "site hierarchy topology relationship between several macrocells corresponding to low battery voltage alarms" through digital quantification. It can quickly and automatically determine the approximate distance and site hierarchy relationship between macrocells corresponding to multiple randomly generated low battery voltage alarms. Using this technique can assist in quickly and automatically determining the risk of wireless cluster faults in the area where the macrocells corresponding to low battery voltage alarms are located. Furthermore, when using this technique to determine the similarity of wireless macrocells involved in low battery voltage alarms, it avoids performing calculations every time, thus speeding up the judgment process.

[0129] This application proposes an automatic decision-making method for sending wireless cluster failure early warning notifications. This technology uses factors such as the number of wireless network elements involved in out-of-service alarms, their similarity, the scale of affected users, user complaint tendencies, and wireless network load—factors that influence the provision of basic wireless service capabilities within the target area and the severity of potential wireless complaints in the event of a wireless cluster failure—as the basis for automatically issuing wireless cluster failure early warning information. As the number of out-of-service RRUs or macro base station low-voltage alarms increases, and with the differences in the weights of various factors for each wireless network element, the "comprehensive similarity coefficient" or "early warning coefficient" increases at different rates. When either reaches a preset value, the system immediately issues a wireless cluster failure early warning notification, avoiding both false alarms and untimely warnings.

[0130] This application proposes a flexible periodic polling mechanism. It acquires alarm information related to service outages from the wireless network management system and macro base station power monitoring platform. The number and similarity of the wireless sites involved in the acquired alarms serve as the basis for automatically adjusting the polling period flexibly from long to short. Simultaneously, it specifically sets the range of network elements to be polled. This improves the timeliness of alarm information acquisition and helps to rationally allocate system computing resources, avoiding rapid increases and waste of computing resource demands.

[0131] This application proposes an automatic analysis and location method for outage-related alarms. Starting from several dimensions, including RRUs, BBUs, and transport layer devices, it uses a list of outage-related alarms obtained during the polling cycle, combined with alarm code information, and automatically analyzes and locates the problem using a feature matrix of wireless network elements according to the network topology hierarchy. This automatic analysis using a feature matrix of wireless network elements based on the network topology hierarchy is logically clear, accurately locates the alarm problem, and has a fast analysis speed, helping wireless network maintenance personnel to handle problems efficiently and accurately.

[0132] Figure 3 This is a schematic diagram of an optional wireless obstacle warning system according to an embodiment of the present invention, such as... Figure 3 As shown in the diagram, the system includes an interactive network management platform, a basic data platform, an environmental monitoring platform, a system communication bus module, an alarm data processing module, an alarm analysis module, and a notification management module. The network management platform includes a northbound interface, the basic data platform includes an open interface, the environmental monitoring platform includes an open interface, the alarm data processing module includes a data caching submodule, an alarm data parsing submodule, an alarm data processing submodule, and an alarm polling submodule, the alarm analysis module includes a group fault early warning analysis submodule, an RRU fault analysis submodule, a BBU fault analysis submodule, and a transport layer device fault analysis submodule, and the notification management module includes a group robot interface, an SMS interface, and an email interface.

[0133] Figure 4 This is a schematic diagram of an optional flexible periodic polling process according to an embodiment of the present invention, such as... Figure 4 As shown, alarm polling is performed according to a preset polling period to determine if there are any alarms. If there are no alarms, the alarm data processing module, alarm data analysis module, and notification management module remain silent, and the process jumps to execute alarm polling according to the preset polling period. If there are alarms, the "group failure early warning analysis submodule" analyzes the alarm data and obtains a comprehensive similarity coefficient α or β to determine whether it belongs to case 1: α≥α0 or β≥β0, or case 2: α<α0 or β<β0. If it belongs to case 1, the polling period is set to 5 seconds, the polling period is changed, and the process jumps to execute alarm polling according to the preset polling period. If it belongs to case 2, the polling period is set to 5 minutes, the polling period is changed, and the process jumps to execute alarm polling according to the preset polling period.

[0134] Figure 5 This is a schematic diagram illustrating the process of determining target similarity under an optional indoor scene type according to an embodiment of the present invention, such as... Figure 5As shown, a set of latitude and longitude coordinates pol for all RRUs within a certain polygonal frame is established. The latitude and longitude of the center point of the RRUs within the polygonal frame are calculated. All RRU coordinates within the polygonal frame are replaced by the latitude and longitude of the center point. Using the above steps, a coordinate information table of all RRUs within all polygonal frames of the entire network is obtained. The shortest distance between any two RRUs in the RRU coordinate information table is calculated to obtain the "shortest distance close RRU table". Normalization is performed using the shortest non-zero distance between adjacent RRUs as the denominator. An upper limit value for the degree of proximity is set to establish a "subset of adjacent RRUs". Further, "subsets of similarity" N1 and N2 are established. Following the above steps, "subsets of similarity" N1 and N2 for all polygonal frames are established. The N1 and N2 of all polygonal frames are summarized to form a "normalized set of adjacent RRUs", N, N = {N1, N2}. When two RRUs that have failed belong to subsets N1 or N2, they represent different degrees of proximity.

[0135] Figure 6 This is a schematic diagram illustrating the process of determining target similarity under an optional outdoor scene type according to an embodiment of the present invention, such as... Figure 6 As shown, each macro station in the entire network is uniquely numbered. The straight-line distance between each macro station and all other macro stations within a specified distance range is calculated to obtain a distance function. Based on the preset average station spacing values ​​of the first and second tier stations, the distance function is normalized. The first-tier station subset and the second-tier station subset of each macro station are obtained. All first-tier station subsets are summarized to form the first-tier station set S1; similarly, the second-tier station set S2 is formed. S1 and S2 are summarized to form the macro station address hierarchy set S. When two macro stations that have experienced a failure belong to subsets S1 or S2, it represents different degrees of proximity.

[0136] Figure 7 This is a schematic diagram illustrating the target warning coefficient determination process under an optional indoor scene type according to an embodiment of the present invention, as shown below. Figure 7 As shown, the warning coefficient A and the "comprehensive similarity system" α of wireless group fault type 1 are calculated. When α≥α0, the polling period is modified and the polling frequency is increased. The "alarm analysis module" performs fault analysis. When α≥α1 or A≥A1, the "group fault warning analysis submodule" feeds back the analysis results to the "notification management" module. The "notification management" module sends wireless group fault warning information to the wireless network maintenance administrator.

[0137] Figure 8 This is a schematic diagram illustrating the target warning coefficient determination process under an optional outdoor scene type according to an embodiment of the present invention, as shown below. Figure 8As shown, the warning coefficient β and the "comprehensive similarity system" β of wireless group faults of type 2 are calculated. When β≥β0, the polling period is modified and the polling frequency is increased. The "alarm analysis module" performs fault analysis. When β≥β1 or B≥B1, the "group fault warning analysis submodule" feeds back the analysis results to the "notification management" module. The "notification management" module sends wireless group fault warning information to the wireless network maintenance administrator.

[0138] Figure 9 This is a schematic diagram illustrating an optional fault location analysis according to an embodiment of the present invention, such as... Figure 9 As shown, a one-dimensional feature table is established for each RRU based on the board number, room number, rack number, and BBU number; these are then merged to obtain a key information table for all RRUs; faulty RRUs are associated, and the number of each feature in the key information table of the faulty RRUs is counted to obtain a feature quantity table; the feature corresponding to the element with the largest value in the feature quantity table is then calculated; the cause of the fault is located in this feature.

[0139] According to another aspect of the embodiments of this application, a wireless fault warning device is also provided. This device can execute the wireless fault warning method of the above embodiments. The specific implementation method and preferred application scenarios are the same as those of the above embodiments, and will not be repeated here.

[0140] Figure 10 This is a schematic diagram of a wireless obstacle warning device according to an embodiment of this application, such as... Figure 10 As shown, the device includes the following: an acquisition module 1002, a determination module 1004, a prediction module 1006, and a generation module 1008.

[0141] The system comprises the following modules: an acquisition module, used to acquire network element attribute data of multiple target wireless network elements within a preset area when a communication fault is detected therein; a determination module, used to determine the target similarity of the multiple target wireless network elements, where the target similarity characterizes the spatial clustering degree among the multiple target wireless network elements; a prediction module, used to predict wireless group faults in the preset area based on the target similarity, and obtain a prediction result, where the prediction result indicates whether wireless group faults will occur in the preset area at a preset future time, which is after the current time; and a generation module, used to generate wireless group fault warning information based on the network element attribute data when the prediction result indicates that wireless group faults will occur in the preset area at a preset future time, where the wireless group fault warning information includes: fault cause location results of multiple target wireless network elements experiencing faults.

[0142] The prediction module is further configured to determine that the prediction result is that no wireless cluster failure will occur in the preset area at a preset future time when the target similarity is less than the first similarity threshold, and to repeatedly execute the step of determining the target similarity of multiple target wireless network elements; and to determine that the prediction result is that wireless cluster failure will occur in the preset area at a preset future time when the target similarity is greater than or equal to the first similarity threshold.

[0143] The prediction module is further configured to repeatedly execute the step of determining the target similarity of multiple target wireless network elements based on a first polling period when the target similarity is less than a second similarity threshold, wherein the second similarity threshold is less than the first similarity threshold; and to repeatedly execute the step of determining the target similarity of multiple target wireless network elements based on a second polling period when the target similarity is greater than or equal to the second similarity threshold, wherein the second polling period is less than the first polling period.

[0144] The prediction module is also used to acquire target warning parameters for multiple target wireless network elements. The target warning parameters are used to represent parameters determined based on the number of users during the busy period of the preset area, the average resource utilization rate during the busy period, user feedback information on communication quality, and whether the current period belongs to the busy period. If the target warning parameters are greater than or equal to the preset warning parameter threshold, the prediction result is determined to be that a wireless group failure will occur in the preset area at a preset future time.

[0145] The determination module is further used to obtain the region type of the preset area. The region type includes an indoor scene type and an outdoor scene type. The indoor scene type represents an indoor scene without a backup battery to power the wireless network elements in the preset area, and the outdoor scene type represents an outdoor scene with a backup battery to power the wireless network elements in the preset area. When the region type is an indoor scene type, the distance between multiple target wireless network elements is normalized based on a preset polygonal bounding box to obtain the target similarity. When the region type is an outdoor scene type, the distance between multiple target wireless network elements is normalized based on the site hierarchy relationship between multiple target wireless network elements to obtain the target similarity.

[0146] The generation module is also used to perform feature localization on multiple target wireless network elements based on network element attribute data to obtain a multi-level feature matrix. The multi-level feature matrix is ​​used to represent the hierarchical attribute matrix of the multi-level structure to which the multiple target wireless network elements belong. The multi-level structure includes at least radio frequency units, baseband units, and equipment rooms. Based on the multi-level feature matrix, the module performs fault analysis on multiple target wireless network elements to obtain fault cause localization results. Based on the fault cause localization results, the module generates wireless group fault early warning information.

[0147] Embodiments of this application also provide an electronic device, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of the present invention during runtime.

[0148] The aforementioned memory can refer to devices inside a computer used to store data and programs, including RAM, hard disks, etc. RAM can be used to temporarily store running programs and data, while hard disks can be used to store programs and data long-term. Memory enables the computer to read and write data and execute programs. The aforementioned processor is responsible for executing instructions in computer programs and performing data processing. It can also be responsible for controlling and executing various operations, including arithmetic operations, logical operations, and data transmission.

[0149] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0150] The aforementioned computer storage media can refer to the media used in computer memory to store certain discontinuous physical quantities. Computer storage media mainly include semiconductors, magnetic cores, magnetic drums, magnetic tapes, laser discs, etc. Computer-readable storage media include stored programs, which can be a set of instructions that a computer can recognize and execute, running on an electronic computer to meet certain information needs.

[0151] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0152] The aforementioned computer program products can refer to software programs that have been written, tested, and released, and can run on computers or other devices. Computer program products can include application programs, operating systems, utility software, etc., used to achieve specific functions or solve specific problems.

[0153] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of the present invention.

[0154] The aforementioned non-volatile computer-readable storage medium can refer to a medium for storing data. Non-volatile computer-readable storage media can retain data without loss when power is off and can be used to store long-term data, such as operating systems, applications, and user files. Non-volatile storage media can include hard disk drives, solid-state drives, optical disks, and flash memory storage devices, etc.

[0155] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of the present invention.

[0156] The aforementioned computer program can refer to a set of instructions used to tell the computer to perform specific tasks or operations. Computer programs can be written by programmers using specific programming languages ​​and can include algorithms, data structures, logic, and control flow. Computer programs can be used for a variety of purposes, including application software, operating systems, etc.

[0157] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0158] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0159] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0160] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0161] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0162] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for early warning of wireless cluster faults, characterized in that, include: When a communication failure is detected in a preset area, the network element attribute data of multiple target wireless network elements in the preset area are obtained, wherein the multiple target wireless network elements are used to represent wireless network elements in the preset area that have experienced failures between a preset historical time and the current time. Determine the target proximity of the plurality of target wireless network elements, wherein the target proximity is used to characterize the spatial clustering degree among the plurality of target wireless network elements; Based on the similarity of the targets, a wireless cluster fault prediction is performed on the preset area to obtain a prediction result, wherein the prediction result is used to indicate whether a wireless cluster fault will occur in the preset area at a preset future time, and the preset future time is after the current time; If the prediction result indicates that a wireless cluster failure will occur in the preset area at the preset future time, a wireless cluster failure early warning information is generated based on the network element attribute data. The wireless cluster failure early warning information includes: the fault location results of the multiple target wireless network elements experiencing failures.

2. The method for early warning of wireless network faults according to claim 1, characterized in that, Based on the similarity of the targets, wireless obstacle prediction is performed on the preset area to obtain prediction results, including: If the target similarity is less than a first similarity threshold, the prediction result is determined to be that the preset area will not experience wireless network failure at the preset future time, and the step of determining the target similarity of the multiple target wireless network elements is repeated. If the similarity of the target is greater than or equal to the first similarity threshold, the prediction result is determined to be that the preset area will experience wireless clustering at the preset future time.

3. The method for early warning of wireless cluster faults according to claim 2, characterized in that, Repeatedly performing the step of determining the target similarity of the plurality of target wireless network elements includes: If the target similarity is less than the second similarity threshold, the step of determining the target similarity of the plurality of target wireless network elements is repeated based on the first polling cycle, wherein the second similarity threshold is less than the first similarity threshold; If the target similarity is greater than or equal to the second similarity threshold, the step of determining the target similarity of the plurality of target wireless network elements is repeated based on the second polling period, wherein the second polling period is less than the first polling period.

4. The method for early warning of wireless network faults according to claim 2, characterized in that, Determining that the prediction result indicates a wireless cluster failure will occur in the preset area at the preset future time includes: Obtain target warning parameters for the plurality of target wireless network elements, wherein the target warning parameters are used to represent parameters determined based on the number of users during the busy period of the preset area, the average resource utilization rate during the busy period, user feedback information on communication quality, and whether the current period belongs to the busy period; If the target warning parameter is greater than or equal to a preset warning parameter threshold, the prediction result is determined to be that the preset area will experience wireless network failure at the preset future time.

5. The early warning method for wireless cluster faults according to claim 1, characterized in that, Determining the target similarity of the plurality of target wireless network elements includes: Obtain the region type of the preset region, wherein the region type includes an indoor scene type and an outdoor scene type. The indoor scene type is used to represent an indoor scene that does not have a backup battery to power the wireless network elements in the preset region, and the outdoor scene type is used to represent an outdoor scene that has a backup battery to power the wireless network elements in the preset region. When the area type is the indoor scene type, the distance between the stations of the multiple target wireless network elements is normalized based on a preset polygonal frame to obtain the similarity of the targets. When the area type is the outdoor scene type, based on the site hierarchy relationship between the multiple target wireless network elements, the inter-site spacing of the multiple target wireless network elements is normalized to obtain the target similarity.

6. The method for early warning of wireless cluster faults according to claim 1, characterized in that, Based on the network element attribute data, wireless cluster failure early warning information is generated, including: Based on the network element attribute data, feature localization is performed on the multiple target wireless network elements to obtain a multi-level feature matrix. The multi-level feature matrix is ​​used to represent the hierarchical attribute matrix of the multi-level structure to which the multiple target wireless network elements belong. The multi-level structure includes at least a radio frequency unit, a baseband unit, and an equipment room. Based on the multi-level feature matrix, fault analysis is performed on the multiple target wireless network elements to obtain the fault cause location result; The wireless cluster fault early warning information is generated based on the fault cause location results.

7. A wireless obstacle warning device, characterized in that, include: The acquisition module is used to acquire network element attribute data of multiple target wireless network elements in the preset area when a communication failure is detected in the preset area. The multiple target wireless network elements are used to represent wireless network elements in the preset area that have failed between a preset historical time and the current time. A determining module is used to determine the target similarity of the plurality of target wireless network elements, wherein the target similarity is used to characterize the spatial clustering degree among the plurality of target wireless network elements; The prediction module is used to predict wireless clustering in the preset area based on the similarity of the target, and obtain a prediction result, wherein the prediction result is used to indicate whether wireless clustering will occur in the preset area at a preset future time, and the preset future time is after the current time; The generation module is used to generate wireless cluster failure early warning information based on the network element attribute data when the prediction result indicates that a wireless cluster failure will occur in the preset area at the preset future time. The wireless cluster failure early warning information includes: the fault location results of the multiple target wireless network elements experiencing failures.

8. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, executes the wireless cluster fault warning method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the wireless obstacle warning method according to any one of claims 1 to 6.

10. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the wireless cluster fault warning method according to any one of claims 1 to 6.

Images