A 5G wireless base station operation and maintenance management system and method

Abstract

The present invention relates to the technical field of base station operation and maintenance management, and specifically, to a 5G wireless base station operation and maintenance management system and method. It includes: a real-time monitoring module monitors the 5G network performance indicators of the base station in real time through the stream processing framework Apache Kafka; a real-time analysis module calculates the average value of the signal strength through a window function, detects the abnormal value of the data transmission speed through the IQR statistical method, and evaluates the trend of the connection quality through the ARIMA event sequence analysis method. The fault diagnosis module diagnoses network faults based on the real-time analysis results through fault tree analysis and provides repair suggestions. The automated operation and maintenance module uses a random forest algorithm to achieve fault prediction, automatic repair and dynamic resource allocation. The present invention is designed to process a large amount of real-time data streams, including signal strength, data transmission speed, and connection quality through the stream processing framework Apache Kafka, and has low latency characteristics.

Description

Technical Field

[0001] The present invention relates to the technical field of base station operation and maintenance management, and specifically, to a 5G wireless base station operation and maintenance management system and method. Background Art

[0002] In the past, in the operation and maintenance management of base stations, the monitoring frequency and data volume were relatively low, and relatively simple monitoring tools and technologies were used. The monitoring scope and depth were limited by the technical characteristics of the 4G / LTE network. The analysis tools and technologies may be relatively basic and lack advanced data analysis functions. The fault response time is long and may take several minutes or even longer. The diagnostic tools may lack advanced fault prediction functions. The automation level is relatively low, and some operation and maintenance tasks may still need to be completed manually. Automated tools may lack advanced self-repair and optimization functions. Therefore, a 5G wireless base station operation and maintenance management system and method are provided. Summary of the invention

[0003] The purpose of the present invention is to provide a 5G wireless base station operation and maintenance management system and method to solve the problems of small monitoring range, low efficiency, poor data analysis function and low degree of automation proposed in the above background technology.

[0004] To achieve the above objectives, on the one hand, the present invention aims to provide a 5G wireless base station operation and maintenance management system, including:

[0005] A real-time monitoring module, which monitors the 5G network performance indicators of the base station in real time through the stream processing framework Apache Kafka;

[0006] A real-time analysis module, wherein the real-time analysis module calculates the average value of the signal strength through a window function, detects abnormal values ​​of the data transmission speed through an IQR statistical method, evaluates the trend of the connection quality through an ARIMA event sequence analysis method, reduces the overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal through weighted averaging and introducing environmental factors to optimize the window function;

[0007] The optimized window function is:

[0008]

[0009] in,

[0010]

[0011]

[0012] Indicates the smoothed time series signal at time The value of Indicates the dynamically adjusted window size; Indicates that the window function is at time The weight of Represents the time series signal at time The value of represents the linear combination function based on environmental factors; Indicates the change threshold; Indicates the current window size; Indicates the minimum window size; Indicates the maximum window size; Indicates time The rate of change of the signal; Represents the time series signal at time The value of Represents the time series signal at time The value of

[0013] A fault diagnosis module, which diagnoses network faults through fault tree analysis based on real-time analysis results and provides repair suggestions;

[0014] The automated operation and maintenance module uses a random forest algorithm to achieve fault prediction, automatic repair, and dynamic resource allocation.

[0015] As a further improvement of the technical solution, the real-time monitoring module monitors the 5G network performance indicators of the base station in real time through the stream processing framework Apache Kafka, including the following steps:

[0016] S2.1. Deploy Kafka server. 5G network performance indicators include signal strength, data transmission speed and connection quality. Create different Topics for each performance indicator.

[0017] S2.2. Deploy data collectors at base stations to regularly collect signal strength, data transmission speed, and connection quality data, and send them to corresponding Kafka Topics;

[0018] S2.3. Use the Kafka consumer API to subscribe to the previously created Topics and read the messages.

[0019] As a further improvement of the technical solution, the real-time analysis module includes a data unit, a transmission unit, and a connection unit;

[0020] The data unit uses a sliding window function to aggregate signal strength data and calculate the average value, reduces the overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal by weighted averaging and introducing environmental factors to optimize the window function;

[0021] The transmission unit uses the IQR statistical method to identify outliers in data transmission speed;

[0022] The connection unit uses the ARIMA time series analysis method to evaluate the changing trend of connection quality over time.

[0023] As a further improvement of the technical solution, the data unit uses a window function to aggregate the signal strength data and calculates the average value, reduces the overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal by weighted averaging and introducing environmental factors to optimize the window function, specifically:

[0024] The specific calculation of the average value before window function optimization is:

[0025]

[0026] in, Indicates signal strength data, Represents the average value of all signal strength values ​​in the window, Indicates the window size, Indicates the index of the signal strength data point currently being processed. Used to traverse the signal strength data points within the window;

[0027] The specific calculation of the average value after the window function is optimized is:

[0028] S4.1. Use the IQR method to detect outliers in the signal strength data. For the detected outliers, use the linear interpolation method to replace them with the average of the neighboring values.

[0029] S4.2, obtain GIS data such as base station location coordinates, topography, and building distribution; monitor the environmental conditions around the base station through sensors, including temperature, humidity, and wind speed; assign an environmental factor to each base station based on the degree of impact of the environmental conditions on the base station;

[0030] S4.3, calculate the rate of change of signal strength over time, evaluate the stability of signal strength according to the rate of change, and dynamically adjust the window size according to the stability;

[0031] S4.4. Dynamically adjust the weights according to the changes in signal strength and external environmental conditions, combine outlier processing, GIS data integration, dynamic adjustment of window size and adaptive weight function to form the final optimized window function.

[0032] As a further improvement of the technical solution, the transmission unit uses an IQR statistical method to identify abnormal values ​​of the data transmission speed, including the following steps:

[0033] S5.1. Collect the data transmission speed records over a period of time as a data set;

[0034] S5.2, sort the data in ascending order;

[0035] S5.3. Calculate the first quartile Q1, where Q1 is the value at the 25% position among all the data, and calculate the third quartile Q3, where Q3 is the value at the 75% position among all the data;

[0036] S5.4. Calculate the interquartile range (IQR). ;

[0037] S5.5. Define outlier limits; lower limit: , upper limit: ; Any data point below the lower limit is considered an outlier, and any data point above the lower limit is considered an outlier;

[0038] S5.6. Traverse each data point in the data set and mark those values ​​below the lower limit and above the upper limit as outliers.

[0039] As a further improvement of the technical solution, the connection unit uses the ARIMA time series analysis method to evaluate the change trend of the connection quality over time, including the following steps:

[0040] S6.1. Collect data about the quality of the connection, including packet loss, latency, and jitter;

[0041] S6.2. Draw a time series graph to intuitively understand the changing trend of connection quality and calculate basic statistics, including mean, standard deviation, minimum and maximum values;

[0042] S6.3. Use the ADF test to check whether the data is stationary. If not, perform differential processing until the data is stationary.

[0043] S6.4. Determine the number of autoregressive terms, difference order, and moving average terms of model parameters based on the ACF and PACF graphs;

[0044] S6.5. Check whether the residual sequence of the fitted ARIMA model is white noise;

[0045] S6.6. Use the determined ARIMA model to predict future connection quality, and the prediction results are used to evaluate the trend of connection quality over time.

[0046] As a further improvement of this technical solution, the

[0047] In S6.3, the specific steps to use the ADF test to check whether the data is stationary are:

[0048] S6.31, construct ADF regression model;

[0049] S6.32. Select the number of lag periods Perform ADF regression and calculate ADF statistics;

[0050] S6.33. Compare the calculated ADF statistic with the critical value T. If the ADF statistic is less than the critical value, the data is considered to be stationary; if the ADF statistic is greater than the critical value, the data is considered to be non-stationary.

[0051] As a further improvement of the technical solution, the fault diagnosis module diagnoses network faults through fault tree analysis based on real-time analysis results, including the following steps:

[0052] S8.1. Determine the specific network failure to be analyzed: 5G base station suffered DoS attack resulting in service interruption as the top event;

[0053] S8.2. List the direct causes of the top event, including low signal strength, abnormal data transmission, and decreased connection quality;

[0054] S8.3. Use OR gates to represent the logical relationship between basic events and construct a complete fault tree structure;

[0055] S8.4. Collect historical network security related information, including historical data on signal strength, historical data on data transmission speed, and historical data on connection quality;

[0056] S8.5. Based on historical data, calculate the probability of occurrence of each basic event, including the probability of using historical data to calculate the signal strength below the threshold R, the probability of using historical data to calculate the transmission speed abnormality, and the probability of using historical data to calculate the connection quality below the standard T. Evaluate which basic event has the greatest impact on the top event and determine the key failure point; find the minimum set of basic events that cause the top event to occur;

[0057] S8.6. Based on the results of the fault tree analysis, adjust the network architecture, including adjusting the signal coverage to improve signal strength, real-time traffic monitoring and anomaly detection mechanism, timely discovering data transmission speed anomalies, and optimizing the connection quality monitoring mechanism.

[0058] As a further improvement of the technical solution, the automated operation and maintenance module uses a random forest algorithm to achieve fault prediction, automatic repair and dynamic resource allocation, including the following steps:

[0059] S9.1. Collect historical data related to the fault, including signal strength, data transmission speed, and connection quality, extract features reflecting the status of the device and potential failure modes from the raw data, and divide the data set into a training set and a test set;

[0060] S9.2. Train the random forest model using the training set data, select hyperparameters, including the number of trees q and the maximum depth e, and use cross-validation to evaluate the performance of the model and adjust the hyperparameters.

[0061] S9.3. Use the trained model to predict new data to identify potential failure modes;

[0062] S9.4. Based on the results of fault prediction, define a set of automatic repair strategies. When the model predicts a potential fault, the corresponding repair strategy is automatically triggered;

[0063] S9.5. Dynamically adjust resource allocation based on prediction results.

[0064] On the other hand, the present invention provides a 5G wireless base station operation and maintenance management method, based on the above-mentioned 5G wireless base station operation and maintenance management system, comprising the following steps:

[0065] S10.1. Real-time monitoring of 5G network performance indicators of base stations through the stream processing framework Apache Kafka, including signal strength, data transmission speed, and connection quality;

[0066] S10.2, calculating the average value of signal strength by window function, detecting abnormal values ​​of data transmission speed by IQR statistical method, and evaluating the trend of connection quality by ARIMA event sequence analysis method;

[0067] S10.3. Based on the real-time analysis results, diagnose network faults through fault tree analysis and provide repair suggestions;

[0068] S10.4. Use the random forest algorithm to achieve fault prediction, automatic repair and dynamic resource allocation.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. In the 5G wireless base station operation and maintenance management system and method, the stream processing framework Apache Kafka can process a large amount of real-time data streams, including signal strength, data transmission speed, and connection quality. It has low latency characteristics and can quickly process and forward data, which is particularly important for scenarios that require real-time monitoring. Based on these monitoring data, the system can quickly identify changes in network performance. By responding to changes in network performance in real time, the operation and maintenance team can adjust the network configuration in a timely manner, such as dynamically allocating resources to cope with sudden traffic, optimizing signal strength, and adjusting connection quality parameters, thereby improving overall network performance and user experience.

[0071] 2. In the 5G wireless base station operation and maintenance management system and method, fault tree analysis can help identify various combinations of events that lead to faults, thereby providing targeted repair suggestions; the random forest algorithm can accurately predict possible faults and automatically trigger the repair process, while dynamically adjusting resource allocation according to network load. Through fault tree analysis and random forest algorithms, the operation and maintenance management system can automatically detect, diagnose and repair faults, greatly reducing the need for manual intervention and improving operation and maintenance efficiency. Dynamic resource allocation can ensure that resources are reasonably allocated and avoid resource waste. It can also ensure rapid response when network load changes and improve resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 It is the overall flow chart of the present invention;

[0073] The meaning of each number in the figure is:

[0074] 1. Real-time monitoring module; 2. Real-time analysis module; 3. Fault diagnosis module; 4. Automated operation and maintenance module. DETAILED DESCRIPTION

[0075] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example 1

[0076] See also Figure 1 As shown, a 5G wireless base station operation and maintenance management system is provided, including

[0077] Real-time monitoring module 1 monitors the 5G network performance indicators of the base station in real time through the stream processing framework Apache Kafka. The performance indicators include signal strength, data transmission speed, and connection quality;

[0078] In this embodiment, the real-time monitoring module 1 monitors the 5G network performance indicators in real time through the stream processing framework Apache Kafka. Kafka can process a large amount of real-time data streams, which is very suitable for the large amount of performance indicator data generated in the 5G network. It has low latency characteristics and can quickly process and forward data, which is particularly important for scenarios that require real-time monitoring. It can be easily expanded by adding more servers to handle the growing amount of data. The data retention time can be set as needed, so as to flexibly manage storage space and data access. It provides strong security features, including user authentication and access control, to ensure that only authorized consumers can access the data, including the following steps:

[0079] S2.1. Deploy Kafka server. 5G network performance indicators include signal strength, data transmission speed and connection quality. Create different Topics for each performance indicator.

[0080] S2.2. Deploy data collectors at base stations to regularly collect signal strength, data transmission speed, and connection quality data, and send them to corresponding Kafka Topics;

[0081] S2.3. Use the Kafka consumer API to subscribe to the previously created Topics and read the messages, which are the original 5G network performance data.

[0082] The real-time analysis module 2 calculates the average value of the signal strength through the window function, detects the abnormal value of the data transmission speed through the IQR statistical method, and evaluates the trend of the connection quality through the ARIMA event sequence analysis method;

[0083] In this embodiment, the real-time analysis module 2 includes a data unit, a transmission unit, and a connection unit;

[0084] The data unit uses a sliding window function to aggregate signal strength data and calculate the average value, reduces the overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal by weighted averaging and introducing environmental factors to optimize the window function;

[0085] The transmission unit uses the IQR statistical method to identify outliers in data transmission speed;

[0086] The connection unit uses the ARIMA time series analysis method to evaluate the changing trend of connection quality over time.

[0087] Among them, the data unit uses a window function to aggregate signal strength data and calculate the average value. The overlap between windows is reduced by dynamically adjusting the window size and step size, and the signal is smoothed by weighted averaging and introducing environmental factors to optimize the window function. The introduction of environmental factors such as topography, building distribution, climate conditions, traffic conditions, and human interference can better adapt to the characteristics of signals changing with the environment, thereby obtaining more accurate smoothing results. By combining outlier detection and processing, GIS data integration, dynamic adjustment of window size, and adaptive weight functions, the window function can be further optimized to better adapt to different geographical locations and the influence of external environmental factors. This optimization can improve the effect of signal strength smoothing, thereby improving the performance and reliability of the entire 5G wireless base station operation and maintenance management system, specifically:

[0088] The specific calculation of the average value before window function optimization is:

[0089]

[0090] in, Indicates signal strength data, , Represents the average value of all signal strength values ​​in the window, Indicates the window size, Indicates the index of the signal strength data point currently being processed. Used to iterate over the signal strength data points within a window.

[0091] The specific calculation of the average value after the window function is optimized is:

[0092] S4.1. Use the IQR method to detect outliers in the signal strength data. For the detected outliers, use the linear interpolation method to replace them with the average of the neighboring values.

[0093] Among them, in linear interpolation, when there is a straight line between two known data points, any unknown value between the two points can be estimated by this straight line. When it comes to the processing of missing data, linear interpolation can be used to fill in the missing values ​​by taking the average of adjacent known values ​​to replace the missing values;

[0094] S4.2, obtain GIS data such as base station location coordinates, topography, and building distribution; monitor the environmental conditions around the base station through sensors, including temperature, humidity, and wind speed; assign an environmental factor to each base station based on the degree of impact of the environmental conditions on the base station;

[0095] S4.3, calculate the rate of change of signal strength over time, evaluate the stability of signal strength according to the rate of change, and dynamically adjust the window size according to the stability;

[0096] S4.4. Dynamically adjust the weights according to the changes in signal strength and external environmental conditions, combine outlier processing, GIS data integration, dynamic adjustment of window size and adaptive weight function to form the final optimized window function.

[0097] The optimized window function is:

[0098]

[0099] in,

[0100]

[0101]

[0102]

[0103] Indicates the smoothed time series signal at time The value of Indicates the dynamically adjusted window size; Indicates that the window function is at time The weight of Represents the time series signal at time The value of represents the linear combination function based on environmental factors; Indicates the change threshold; Indicates the current window size; Indicates the minimum window size; Indicates the maximum window size; Indicates time The rate of change of the signal; Represents the time series signal at time The value of Represents the time series signal at time The value of represents the dynamically adjusted step size, Indicates the current step length;

[0104] in, Used to update the step size and determine the starting position of the next window, that is, the starting position of the next window should be the end position of the current window plus the adjusted step size ; Consider the impact of various environmental factors (such as topography, building distribution, climate conditions, traffic conditions, and human interference) on signal strength, quantify them as influencing factors of the window function, and output a value in the form of a linear combination, specifically:

[0105]

[0106] express The weight of express The weight of express The weight of express The weight of express The weight of .

[0107] The introduction of makes signal smoothing not only rely on the value of the signal itself, but also considers the influence of environmental conditions on signal quality, thereby improving the accuracy and reliability of the smoothing result.

[0108] In this embodiment, the transmission unit uses the IQR statistical method to identify outliers of the data transmission speed. The IQR method is not affected by extreme values ​​because it is only based on the middle 50% of the data (i.e., the distance between the second and third quartiles), thereby improving the robustness of anomaly detection. The IQR method is based on simple statistical concepts and is easy to understand and implement. By adjusting the IQR multiple (usually 1.5 or 3.0), the range of outliers can be determined according to specific circumstances. The IQR method provides an intuitive way to identify outliers, that is, values ​​outside the first quartile minus 1.5 times the IQR or the third quartile plus 1.5 times the IQR are considered outliers, including the following steps:

[0109] S5.1. Collect the data transmission speed records over a period of time as a data set;

[0110] S5.2, sort the data in ascending order;

[0111] S5.3. Calculate the first quartile Q1, where Q1 is the value at the 25% position among all the data, and calculate the third quartile Q3, where Q3 is the value at the 75% position among all the data;

[0112] S5.4. Calculate the interquartile range (IQR). ;

[0113] S5.5. Define outlier limits; lower limit: , upper limit: ; Any data point below the lower limit is considered an outlier, and any data point above the lower limit is considered an outlier;

[0114] S5.6. Traverse each data point in the data set and mark those values ​​below the lower limit and above the upper limit as outliers.

[0115] In this embodiment, the connection unit uses the ARIMA time series analysis method to evaluate the changing trend of the connection quality over time. The "I" (integral) part in the ARIMA model allows the non-stationary time series to be differentiated, so that the data becomes stable, which is a prerequisite for effective time series analysis. Through differential processing, the trend and seasonal effects can be eliminated, so that the model can better fit the data. The ARIMA model can flexibly adjust the parameters of the autoregressive (AR), differential (I) and moving average (MA) parts to adapt to different types of time series data. The ARIMA model can generate predictions for future time points, which is very useful for evaluating the changing trend of the connection quality. The steps include:

[0116] S6.1. Collect data about the quality of the connection, including packet loss, latency, and jitter, which usually vary over time.

[0117] S6.2. Draw a time series graph to intuitively understand the changing trend of connection quality and calculate basic statistics, including mean, standard deviation, minimum and maximum values;

[0118] S6.3. Use the ADF test to check whether the data is stationary. If not, perform differential processing until the data is stationary.

[0119] Among them, the ADF test is used to check whether the data is stationary. The ADF test is mainly used to test whether the time series has a unit root (that is, whether the time series has the characteristics of a random walk). If a unit root exists, the time series is non-stationary. Detecting a unit root means that there may be a long-term random trend in the original data, which needs to be eliminated by means such as differentiation. It provides a clear statistical basis to determine whether the time series is stationary. By comparing the ADF statistic with the critical value, or the calculated p-value with the significance level (for example, 0.05), you can decide whether to reject the null hypothesis (the time series contains a unit root). The specific steps are:

[0120] S6.31. Construct ADF regression model:

[0121]

[0122] in, express The first difference of is the level value lagged by one period; is the lagged difference term; is the autoregressive coefficient; is a constant term; and is the coefficient of the difference term; is the order of the difference;

[0123] S6.32. Select the number of lag periods Perform ADF regression and calculate ADF statistics;

[0124] S6.33. Compare the calculated ADF statistic with the critical value T. If the ADF statistic is less than the critical value, the data is considered to be stationary; if the ADF statistic is greater than the critical value, the data is considered to be non-stationary.

[0125] S6.4. Determine the model parameters based on the ACF and PACF graphs: p (number of autoregressive terms), d (order of differences), q (number of moving average terms). The p value can be roughly determined from the maximum lag point of the PACF graph, and the q value can be roughly determined from the maximum lag point of the ACF graph. The order of differences d is usually determined by the unit root test.

[0126] S6.5. Check whether the residual sequence of the fitted ARIMA model is white noise to ensure the validity of the model;

[0127] S6.6. Use the determined ARIMA model to predict future connection quality, and the prediction results are used to evaluate the trend of connection quality over time.

[0128] Fault diagnosis module 3 diagnoses network faults through fault tree analysis based on real-time analysis results. Fault tree analysis provides a structured method to identify and analyze all basic events that may cause network failures. It uses graphical representation to decompose complex system failures into a series of logic gates (such as AND gates, OR gates, etc.) and basic events for easy understanding and communication. Through the fault tree, it is possible to clearly see how different fault paths affect the entire network system. It allows the probability of each basic event to be estimated, and the probability of a specific failure mode can be inferred from these probabilities. The following steps are included:

[0129] S8.1. Determine the specific network failure to be analyzed: 5G base station suffered DoS attack resulting in service interruption as the top event;

[0130] S8.2. List the direct causes of the top event, including low signal strength, abnormal data transmission, and decreased connection quality;

[0131] S8.3. Use OR gates to represent the logical relationship between basic events and construct a complete fault tree structure;

[0132] The fault tree structure consists of top events, basic events, and logic gates (OR gates). The top event is that the 5G base station is attacked by DoS, resulting in service interruption; the basic events are: low signal strength, abnormal data transmission, and decreased connection quality; the logic gates are: the logical relationship between the basic events represented by the OR gate;

[0133] S8.4. Collect historical network security related information, including historical data on signal strength, historical data on data transmission speed, and historical data on connection quality;

[0134] S8.5. Based on historical data, calculate the probability of occurrence of each basic event, including the probability of using historical data to calculate the signal strength below the threshold R, the probability of using historical data to calculate the transmission speed abnormality, and the probability of using historical data to calculate the connection quality below the standard T. Evaluate which basic event has the greatest impact on the top event and determine the key fault point; find the minimum set of basic events that cause the top event to occur, and these sets represent different fault paths;

[0135] S8.6. Based on the results of the fault tree analysis, adjust the network architecture, including adjusting the signal coverage to improve signal strength, real-time traffic monitoring and anomaly detection mechanism, timely discovering data transmission speed anomalies, and optimizing the connection quality monitoring mechanism.

[0136] Automated operation and maintenance module 4 uses the random forest algorithm to achieve fault prediction, automatic repair and dynamic resource allocation. Random forest is a powerful machine learning algorithm that can process high-dimensional data and improve prediction accuracy by integrating multiple decision trees. Random forest can quickly predict faults and immediately start the automatic repair process, reducing the need for manual intervention. It can dynamically adjust resource allocation according to the current workload and predicted fault probability, thereby improving resource utilization. The following steps are included:

[0137] S9.1. Collect historical data related to the fault, including signal strength, data transmission speed, and connection quality, extract features reflecting the status of the equipment and potential failure modes from the raw data, including statistical indicators (such as mean and variance) and time series characteristics (such as trend and periodicity), and divide the data set into a training set and a test set;

[0138] S9.2. Train the random forest model using the training set data, select hyperparameters, including the number of trees q and the maximum depth e, and use cross-validation to evaluate the performance of the model and adjust the hyperparameters.

[0139] S9.3. Use the trained model to predict new data to identify potential failure modes;

[0140] S9.4. Based on the results of fault prediction, define a set of automatic repair strategies. When the model predicts a potential fault, the corresponding repair strategy is automatically triggered;

[0141] The repair strategy is as follows: when certain types of failures occur, try to recover by remotely restarting the service or device;

[0142] S9.5. Dynamically adjust resource allocation based on prediction results. When high load is predicted, more resources can be allocated in advance. Example 2

[0143] The difference between Example 2 of the present invention and Example 1 is that this example introduces the 5G wireless base station operation and maintenance management method used by the above-mentioned 5G wireless base station operation and maintenance management system.

[0144] A 5G wireless base station operation and maintenance management method, based on the above-mentioned 5G wireless base station operation and maintenance management system, comprises the following steps:

[0145] S10.1. Real-time monitoring of 5G network performance indicators of base stations through the stream processing framework Apache Kafka, including signal strength, data transmission speed, and connection quality;

[0146] S10.2, calculating the average value of signal strength by window function, detecting abnormal values ​​of data transmission speed by IQR statistical method, and evaluating the trend of connection quality by ARIMA event sequence analysis method;

[0147] S10.3. Based on the real-time analysis results, diagnose network faults through fault tree analysis and provide repair suggestions;

[0148] S10.4. Use the random forest algorithm to achieve fault prediction, automatic repair and dynamic resource allocation.

[0149] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A 5G wireless base station operation and maintenance management system, characterized in that: include: A real-time monitoring module (1), wherein the real-time monitoring module (1) monitors the 5G network performance indicators of the base station in real time through the stream processing framework Apache Kafka; A real-time analysis module (2), wherein the real-time analysis module (2) calculates the average value of the signal strength through a window function, detects abnormal values ​​of the data transmission speed through an IQR statistical method, evaluates the trend of the connection quality through an ARIMA event sequence analysis method, reduces overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal through weighted averaging and introducing environmental factors to optimize the window function; The optimized window function is: ； in, ； ； ； Indicates the smoothed time series signal at time The value of Indicates the dynamically adjusted window size; Indicates that the window function is at time The weight of Represents the time series signal at time The value of represents the linear combination function based on environmental factors; Indicates the change threshold; Indicates the current window size; Indicates the minimum window size; Indicates the maximum window size; Indicates time The rate of change of the signal; Represents the time series signal at time The value of Represents the time series signal at time The value of represents the dynamically adjusted step size, represents the current step length; where Used to update the step size and determine the starting position of the next window, that is, the starting position of the next window should be the end position of the current window plus the adjusted step size ; A fault diagnosis module (3), wherein the fault diagnosis module (3) diagnoses network faults through fault tree analysis based on real-time analysis results and provides repair suggestions; An automated operation and maintenance module (4), wherein the automated operation and maintenance module (4) utilizes a random forest algorithm to achieve fault prediction, automatic repair, and dynamic resource allocation.

2. The 5G wireless base station operation and maintenance management system according to claim 1, characterized in that: The real-time monitoring module (1) monitors the 5G network performance indicators of the base station in real time through the stream processing framework Apache Kafka, including the following steps: S2.

1. Deploy Kafka server. 5G network performance indicators include signal strength, data transmission speed and connection quality. Create different Topics for each performance indicator. S2.

2. Deploy data collectors at base stations to regularly collect signal strength, data transmission speed, and connection quality data, and send them to corresponding Kafka Topics; S2.

3. Use the Kafka consumer API to subscribe to the previously created Topics and read the messages.

3. The 5G wireless base station operation and maintenance management system according to claim 1, characterized in that: The real-time analysis module (2) comprises a data unit, a transmission unit, and a connection unit; The data unit uses a sliding window function to aggregate signal strength data and calculate the average value, reduces the overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal by weighted averaging and introducing environmental factors to optimize the window function; The transmission unit uses the IQR statistical method to identify outliers in data transmission speed; The connection unit uses the ARIMA time series analysis method to evaluate the changing trend of connection quality over time.

4. The 5G wireless base station operation and maintenance management system according to claim 3, characterized in that: The data unit uses a window function to aggregate signal strength data and calculates the average value, reduces overlap between windows by dynamically adjusting the window size and step size, and smoothes the signal by weighted averaging and introducing environmental factors to optimize the window function, specifically: The specific calculation of the average value before window function optimization is: ； in, Indicates signal strength data, Represents the average value of all signal strength values ​​in the window, Indicates the window size, Indicates the index of the signal strength data point currently being processed. Used to traverse the signal strength data points within the window; The specific calculation of the average value after the window function is optimized is: S4.

1. Use the IQR method to detect outliers in the signal strength data. For the detected outliers, use the linear interpolation method to replace them with the average of the neighboring values. S4.2, obtain GIS data such as base station location coordinates, topography, and building distribution; monitor the environmental conditions around the base station through sensors, including temperature, humidity, and wind speed; assign an environmental factor to each base station based on the degree of impact of the environmental conditions on the base station; S4.3, calculate the rate of change of signal strength over time, evaluate the stability of signal strength according to the rate of change, and dynamically adjust the window size according to the stability; S4.

4. Dynamically adjust the weights according to the changes in signal strength and external environmental conditions, combine outlier processing, GIS data integration, dynamic adjustment of window size and adaptive weight function to form the final optimized window function.

5. The 5G wireless base station operation and maintenance management system according to claim 4, characterized in that: The transmission unit uses an IQR statistical method to identify abnormal values ​​of data transmission speed, including the following steps: S5.

1. Collect the data transmission speed records over a period of time as a data set; S5.2, sort the data in ascending order; S5.

3. Calculate the first quartile Q1, where Q1 is the value at the 25% position among all the data, and calculate the third quartile Q3, where Q3 is the value at the 75% position among all the data; S5.

4. Calculate the interquartile range (IQR). ; S5.

5. Define outlier limits; lower limit: , upper limit: ; Any data point below the lower limit is considered an outlier, and any data point above the lower limit is considered an outlier; S5.

6. Traverse each data point in the data set and mark those values ​​below the lower limit and above the upper limit as outliers.

6. The 5G wireless base station operation and maintenance management system according to claim 5, characterized in that: The connection unit uses the ARIMA time series analysis method to evaluate the change trend of connection quality over time, including the following steps: S6.

1. Collect data about the quality of the connection, including packet loss, latency, and jitter; S6.

2. Draw a time series graph to intuitively understand the changing trend of connection quality and calculate basic statistics, including mean, standard deviation, minimum and maximum values; S6.

3. Use the ADF test to check whether the data is stationary. If not, perform differential processing until the data is stationary. S6.

4. Determine the model parameters based on the ACF and PACF graphs: number of autoregressive terms, number of difference orders, and number of moving average terms; S6.

5. Check whether the residual sequence of the fitted ARIMA model is white noise; S6.

6. Use the determined ARIMA model to predict future connection quality, and the prediction results are used to evaluate the trend of connection quality over time.

7. The 5G wireless base station operation and maintenance management system according to claim 6, characterized in that: In S6.3, the specific steps of using the ADF test to check whether the data is stationary are: S6.31, construct ADF regression model; S6.

32. Select the number of lag periods Perform ADF regression and calculate ADF statistics; S6.

33. Compare the calculated ADF statistic with the critical value T. If the ADF statistic is less than the critical value, the data is considered to be stationary; if the ADF statistic is greater than the critical value, the data is considered to be non-stationary.

8. The 5G wireless base station operation and maintenance management system according to claim 1, characterized in that: The fault diagnosis module (3) diagnoses network faults through fault tree analysis based on the real-time analysis results, including the following steps: S8.

1. Determine the specific network failure to be analyzed: 5G base station suffered DoS attack resulting in service interruption as the top event; S8.

2. List the direct causes of the top event, including low signal strength, abnormal data transmission, and decreased connection quality; S8.

3. Use OR gates to represent the logical relationship between basic events and construct a complete fault tree structure; S8.

4. Collect historical network security related information, including historical data on signal strength, historical data on data transmission speed, and historical data on connection quality; S8.

5. Based on historical data, calculate the probability of occurrence of each basic event, including the probability of using historical data to calculate the signal strength below the threshold R, the probability of using historical data to calculate the transmission speed abnormality, and the probability of using historical data to calculate the connection quality below the standard T. Evaluate which basic event has the greatest impact on the top event and determine the key failure point; find the minimum set of basic events that cause the top event to occur; S8.

6. Based on the results of the fault tree analysis, adjust the network architecture, including adjusting the signal coverage to improve signal strength, real-time traffic monitoring and anomaly detection mechanism, timely discovering data transmission speed anomalies, and optimizing the connection quality monitoring mechanism.

9. The 5G wireless base station operation and maintenance management system according to claim 1, characterized in that: The automated operation and maintenance module (4) uses a random forest algorithm to achieve fault prediction, automatic repair, and dynamic resource allocation, including the following steps: S9.

1. Collect historical data related to the fault, including signal strength, data transmission speed, and connection quality, extract features reflecting the status of the device and potential failure modes from the raw data, and divide the data set into a training set and a test set; S9.

2. Train the random forest model using the training set data, select hyperparameters, including the number of trees q and the maximum depth e, and use cross-validation to evaluate the performance of the model and adjust the hyperparameters. S9.

3. Use the trained model to predict new data to identify potential failure modes; S9.

4. Based on the results of fault prediction, define a set of automatic repair strategies. When the model predicts a potential fault, the corresponding repair strategy is automatically triggered; S9.

5. Dynamically adjust resource allocation based on prediction results.

10. A 5G wireless base station operation and maintenance management method, based on the 5G wireless base station operation and maintenance management system according to any one of claims 1 to 9, characterized in that: The steps include: S10.

1. Real-time monitoring of 5G network performance indicators of base stations through the stream processing framework Apache Kafka, including signal strength, data transmission speed, and connection quality; S10.2, calculating the average value of signal strength by window function, detecting abnormal values ​​of data transmission speed by IQR statistical method, and evaluating the trend of connection quality by ARIMA event sequence analysis method; S10.

3. Based on the real-time analysis results, diagnose network faults through fault tree analysis and provide repair suggestions; S10.

4. Use the random forest algorithm to achieve fault prediction, automatic repair and dynamic resource allocation.

Images