CN115550977A - Root cause positioning method and equipment for key performance index abnormity - Google Patents
Root cause positioning method and equipment for key performance index abnormity Download PDFInfo
- Publication number
- CN115550977A CN115550977A CN202211136228.4A CN202211136228A CN115550977A CN 115550977 A CN115550977 A CN 115550977A CN 202211136228 A CN202211136228 A CN 202211136228A CN 115550977 A CN115550977 A CN 115550977A
- Authority
- CN
- China
- Prior art keywords
- performance index
- root cause
- data
- key performance
- correlation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/04—Arrangements for maintaining operational condition
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The application provides a root cause positioning method and equipment for key performance index abnormity, which comprises the steps of selecting a marked original data set by using a machine learning model, selecting a plurality of important performance indexes, and screening out associated performance indexes which do not participate in key performance index calculation. Associating performance index data, marked key performance index data and known root cause performance index data in a plurality of periods before and after the key performance index of the wireless access network is abnormal to construct a root cause positioning data set; calculating the reason difference of the discrete associated performance index data in the root cause positioning data set to obtain the reason of the abnormal key performance index; and performing multiple correlation calculations on the continuous associated performance index data in the root cause positioning data set to obtain the root cause of the key performance index abnormality. Through the arrangement, the reason of the abnormal key performance index can be positioned from the discrete associated performance indexes, and the accuracy of the reason positioning is improved.
Description
Technical Field
The present application relates to communications technologies, and in particular, to a method and an apparatus for locating a root cause of an abnormal key performance indicator.
Background
The wireless access network is a wireless implementation system for transmitting telecommunication services, and Key Performance Indicators (KPI) of access class, maintenance class, mobility, quality class and the like represent the capability of the wireless access network for providing services, and can reflect the operation quality of the wireless access network. When key performance indexes such as low wireless call completing rate, high call drop rate and low switching success rate appear in an area covered by a wireless access network, the area can be judged as a quality difference area. Although the quality difference region can be identified through the abnormal key performance index, the reason of poor quality of the wireless access network is numerous, and the positioning of the quality difference root cause is difficult to realize.
In the prior art, through KPI root cause correlation analysis and detection, a Pearson or spearman similarity calculation function is used for calculating the correlation between the correlated KPI and a main KPI, and root cause analysis is performed, but the Pearson or spearman similarity calculation function is not suitable for discrete data, so that discrete correlation performance indexes are ignored in the root cause correlation analysis process, and the root cause analysis step is not complete, so that the root cause positioning is not accurate enough.
Disclosure of Invention
The application provides a root cause positioning method and equipment for key performance index abnormity, and aims to classify and calculate correlation performance indexes, use multi-step calculation during correlation analysis, and improve the accuracy of root cause positioning when the key performance indexes are abnormal.
An embodiment of the present application provides a root cause positioning method for a key performance index abnormality, including:
and selecting the marked original data set by using a machine learning model, selecting a plurality of important performance indexes, and screening the important performance indexes according to whether the important performance indexes participate in the calculation of the key performance indexes to obtain the associated performance indexes.
Associating performance index data, marked key performance index data and known root cause performance index data in a plurality of periods before and after the key performance index of the wireless access network is abnormal to construct a root cause positioning data set;
calculating the reason difference of the discrete associated performance index data in the root cause positioning data set to obtain the reason of the abnormal key performance index;
and performing multiple times of correlation calculation on the continuous correlation performance index data in the root cause positioning data set to obtain the root cause of the key performance index abnormality.
In an embodiment, performing multiple correlation calculations on continuous associated performance indicator data in a root cause positioning data set to obtain a root cause of a key performance indicator abnormality specifically includes:
carrying out correlation calculation on the continuous correlation performance index data in the root cause positioning data set and the marked key performance index data to obtain continuous correlation performance index data of which the correlation coefficient absolute value is greater than or equal to a height correlation threshold value, and setting the continuous correlation performance index data of which the correlation coefficient absolute value is greater than or equal to the height correlation threshold value as abnormal cause index data;
and performing correlation calculation on the known root cause performance index data and the abnormal cause index data in the root cause positioning data set to obtain the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to a highly correlated threshold value, wherein the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to the highly correlated threshold value is the root cause of the key performance index abnormality.
In an embodiment, the calculating a cause difference value of discrete associated performance indicator data in the root cause positioning data set to obtain a cause of the key performance indicator abnormality specifically includes:
calculating the difference value between the mean value of the discrete associated performance index data in a plurality of periods after the key performance index is abnormal and the mean value of the discrete associated performance index data in a plurality of periods before the key performance index is abnormal aiming at each discrete associated performance index data;
and taking the discrete correlation performance index with the difference value larger than the preset difference value threshold value as the reason of the abnormal key performance index.
In an embodiment, before obtaining the associated performance index, the method further includes:
constructing an original data set based on all counters, performance indexes and key performance index data in a wireless access network communication flow measured by key performance indexes in an observation time period;
marking key performance index data in the original data set according to the abnormal threshold, acquiring marked key performance index data, and acquiring a marked original data set;
in an embodiment, in a plurality of periods before and after an abnormality occurs in a key performance indicator of a radio access network, associating performance indicator data, marked key performance indicator data, and known root cause performance indicator data to construct a root cause location data set, specifically including:
performing data conversion on the associated performance index data, the marked key performance index data and the known root performance index data to obtain an associated performance index feature vector, a marked key performance index feature vector and a known root performance index feature vector;
performing data preprocessing on the associated performance index feature vector, the marked key performance index feature vector and the known root factor feature vector;
performing attribute marking on the associated performance index feature vector according to the data type, and constructing a root cause positioning data set based on the associated performance index feature vector, the marked key performance index feature vector and the known root cause feature vector; wherein the data types include discrete data and continuous data.
In an embodiment, the known root cause performance indicator data includes one or more combinations of uplink weak coverage, downlink weak coverage, over coverage, uplink interference strength, and downlink interference rate.
Another embodiment of the present application provides a root cause positioning apparatus for key performance index abnormality, including:
and the characteristic acquisition module is used for selecting the marked original data set by using a machine learning model, selecting a plurality of important performance indexes, and screening the important performance indexes according to whether the important performance indexes participate in the calculation of the key performance indexes to acquire the associated performance indexes.
The index acquisition module is used for associating performance index data, marked key performance index data and known root cause performance index data in a plurality of periods before and after the key performance index of the wireless access network is abnormal to construct a root cause positioning data set;
the processing module is used for calculating the reason difference of the discrete associated performance index data in the root cause positioning data set to acquire the reason of the abnormal key performance index;
and the processing module is used for carrying out multiple times of correlation calculation on the continuous correlation performance index data in the root cause positioning data set so as to obtain the abnormal root cause of the key performance index.
Another embodiment of the present application provides an electronic device, including: a processor, and a memory communicatively coupled to the processor;
the memory stores computer execution instructions;
the processor executes the computer-executable instructions stored in the memory to implement the root cause localization method for critical performance indicator anomalies provided by the above-described embodiments.
Yet another embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the method for locating a root cause of a critical performance indicator anomaly provided in the foregoing embodiment is implemented.
Yet another embodiment of the present application provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the method for root cause positioning of key performance indicator anomaly provided in the foregoing embodiments is implemented.
The application provides a root cause positioning method and equipment for key performance index abnormity, which collects all counters, performance indexes and key performance index statistical data in a wireless access network communication flow to generate an original data set, selects a plurality of most important characteristics in the performance indexes according to a machine learning model, eliminates the characteristics participating in key performance index calculation according to prior knowledge, and the rest characteristics are associated performance indexes; performing data preprocessing on the correlation performance indexes, the key performance indexes and the known root cause performance index data in a plurality of periods before and after the key performance indexes are abnormal to construct a root cause positioning data set; meanwhile, the data types of the associated performance indexes in the root cause positioning data set are considered, and classified calculation is carried out. And performing difference calculation on the discrete correlation performance indexes to obtain the failure reason of the abnormity of the key performance indexes, and performing two-step correlation calculation on the continuous correlation performance indexes to output the reason of the abnormity of the key performance indexes. When the key performance index is abnormal, the reason can be positioned from the discrete associated performance index, and the reason of the abnormal key performance index can be more accurately positioned according to the multi-step correlation calculation of the continuous key performance index.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and, together with the description, serve to explain the principles of the application.
Fig. 1 is a flowchart of a root cause locating method for key performance indicator abnormality according to an embodiment of the present application;
FIG. 2 is a flowchart of a method for root cause location of key performance indicator anomalies according to another embodiment of the present application;
FIG. 3 is a flowchart of a method for root cause location of key performance indicator anomalies according to yet another embodiment of the present application;
FIG. 4 is a schematic structural diagram of a root cause locating device for a critical performance indicator anomaly according to another embodiment of the present application;
fig. 5 is a schematic structural diagram of an electronic device according to yet another embodiment of the present application.
With the above figures, there are shown specific embodiments of the present application, which will be described in more detail below. These drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the inventive concepts to those skilled in the art by reference to specific embodiments.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the appended claims.
The key performance index can reflect the quality of the service provided by the wireless access network in the communication service process of the wireless access network, and when the key performance index is abnormal, the key performance index represents that the operation quality of the wireless access network in the area is not good, and the root cause of the poor operation quality needs to be positioned.
When a KPI root cause correlation analysis method is used for positioning, a Pearson or spearman similarity calculation function is only suitable for continuous data, when a correlation performance index is selected, the correlation performance index can be only screened out in a discrete mode for convenience of calculation, the existing analysis steps are not perfect, and the accuracy of the KPI root cause correlation analysis method for positioning the root cause is not high enough.
In order to solve the technical problems, the application provides a method and equipment for positioning root cause with abnormal key performance indexes, and aims to perform classified calculation on the associated performance indexes, use multistep calculation during correlation analysis and improve the accuracy of root cause positioning. The technical idea of the application is as follows: collecting all counters, performance indexes and key performance index statistical data in a wireless access network communication process to generate an original data set, selecting a plurality of important features by using a user-defined model for the original data set, and further screening the important features according to whether the important features participate in the calculation of the key performance indexes. And performing data preprocessing on the associated performance indexes, the key performance indexes and the known root cause performance index data in a plurality of periods before and after the key performance indexes are abnormal, taking the data as a root cause positioning data set, and calculating in a classified mode according to the data types in the root cause positioning data set. And performing difference calculation on the discrete correlation performance indexes to obtain failure reasons of the abnormal key performance indexes, and performing two-step correlation calculation on the continuous correlation performance indexes to output the reasons of the abnormal key performance indexes.
As shown in fig. 1, an embodiment of the present application provides a method for locating a root cause of an abnormal key performance index, where the method specifically includes the following steps:
s101, selecting the marked original data set by using a machine learning model, selecting a plurality of important performance indexes, screening the important performance indexes according to whether the important performance indexes participate in the calculation of the key performance indexes, and obtaining the associated performance indexes.
In this step, a machine learning model with a Feature Importance (Feature Importance) measure is selected, for example: in the random forest based on the decision tree learner, the performance indexes in the original data set are used as the input of the model, the marked key performance indexes in the original data set are used as the output of the model, the model is trained, and the random forest can evaluate the importance of each performance index by comparing the contribution of each performance index in each tree in the random forest and select a plurality of important performance indexes. Screening out performance indicators participating in the calculation of key performance indicators, such as: the success rate of establishing the RRC connection E-RAB = the radio access rate, and the two indexes of the success rate of establishing the RRC connection and the success rate of establishing the E-RAB need to be filtered out, so as to obtain the associated performance index.
S102, associating performance index data, marked key performance index data and known root cause performance index data in a plurality of periods before and after the key performance index of the wireless access network is abnormal, and constructing a root cause positioning data set.
In this step, the associated performance index data can be obtained by selecting through the custom model and screening out various performance indexes participating in the calculation of the key performance index. The key performance indicator data is labeled according to its anomaly threshold. When the key performance index of the wireless access network is abnormal, three different types of data, namely associated performance index data, marked key performance index data and known root cause performance index data in a plurality of periods before and after the abnormal key performance index of the wireless access network, are obtained and are used for constructing a root cause positioning data set.
S103, calculating the reason difference of the discrete associated performance index data in the root cause positioning data set, and acquiring the reason of the abnormal key performance index.
The Pearson or spearman similarity calculation function is not suitable for the discrete associated performance index data, so that the reason of the abnormal key performance index is obtained by adopting a difference value calculation mode aiming at the discrete associated performance index data.
And S104, carrying out multiple times of correlation calculation on the continuous correlation performance index data in the root cause positioning data set to obtain the abnormal root cause of the key performance index.
In the step, the continuous associated performance index data and the key performance index data are subjected to correlation calculation to obtain the reason performance index data with the highest correlation, and the reason-tracing reasoning is completed. And performing correlation calculation on the cause performance index data with the highest correlation and the known root cause performance index data, calculating the root cause with the highest correlation, and acquiring the root cause with abnormal key performance indexes.
In the technical scheme, various performance indexes in the communication flow of the wireless access network record the operation condition of the wireless access network, and the associated performance indexes can be screened from the operation condition by judging whether to participate in the calculation of the key performance indexes through a machine learning model. And when the key performance index of the wireless access network is abnormal, constructing a root cause positioning data set based on the associated performance index data, the marked key performance index data and the known root cause performance index data in a plurality of periods before and after the abnormality. And dividing the data type of the associated performance index in the root cause positioning data set into a continuous type and a discrete type, performing difference calculation on the discrete type associated performance index data to obtain the abnormal reason of the key performance index, and performing multiple correlation calculations on the continuous type associated performance index data to obtain the root reason of the abnormal key performance index.
As shown in fig. 2, an embodiment of the present application provides a root cause positioning method for key performance index abnormality, where before obtaining a correlation performance index, the method further includes the following steps:
s201, based on all counters, performance indexes and key performance index data in the wireless access network communication process measured by the key performance indexes in the observation time period, an original data set is constructed.
In this step, the performance statistics of the radio access network are generated by a large number of counters, which are incremented or decremented by different events and signaling messages in the communication flow, and the operating conditions of the radio access network are recorded. And various performance indexes are obtained through different counter calculations. The performance index capable of reflecting the operation quality of the wireless access network is called a key performance index and comprises access type, maintenance type, mobility, quality type, service integrity, utilization rate, availability and service type indexes. After the key performance index is abnormal, an original data set can be constructed according to the counter, the performance index and the key performance index in a plurality of periods before and after the abnormality.
S202, marking the key performance indexes in the original data set according to the abnormal threshold value, obtaining the marked key performance indexes, and obtaining the marked original data set.
In this step, different key performance indicators correspond to different abnormal threshold values, when the attribute value of the key performance indicator is greater than or less than the abnormal threshold value, the key performance indicator is marked, the key performance indicator with the abnormal attribute value is marked as 1, and the key performance indicator with the normal attribute value is marked as 0. And marking all key performance indexes in the original data set, and then obtaining the marked original data set.
In the above technical solution, an original data set is constructed based on original performance indicators in a plurality of weeks before and after the key performance indicator is abnormal in the wireless access network communication flow. And marking and distinguishing the key performance indexes in the original data set according to the abnormal threshold value to obtain a marked original data set.
An embodiment of the present application provides a root cause positioning method for key performance index abnormality, which associates performance index data, marked key performance index data and known root cause performance index data in multiple periods before and after the key performance index of a radio access network is abnormal, and constructs a root cause positioning data set, specifically including:
the known root cause performance index data comprises one or more combinations of uplink weak coverage, downlink weak coverage, over coverage, uplink interference strength and downlink interference rate.
The known root cause performance index data comprises all possible root causes of the key performance index abnormity, and the root cause of the key performance index abnormity needs to be accurately positioned.
And performing data conversion on the associated performance index data, the marked key performance index data and the known root cause performance index data to obtain an associated performance index feature vector, a marked key performance index feature vector and a known root cause feature vector.
And storing each item of data in the form of a feature vector so as to facilitate difference calculation and correlation calculation. For example: the associated performance indicator feature vector is x 1 (i) The known root cause feature vector is x 2 (i) The marked key performance indicator feature vector is y (i) 。
And performing data preprocessing on the associated performance index feature vector, the marked key performance index feature vector and the known root cause feature vector.
The characteristic vectors of all null values are deleted, missing values in the characteristic vectors are filled by using the mean value of the characteristic vectors, the characteristic vectors containing percentage are converted into real numbers, and standard deviation standardization processing is carried out on the characteristic vectors.
Performing attribute marking on the associated performance index feature vector according to the data type, and constructing a root cause positioning data set based on the associated performance index feature vector, the marked key performance index feature vector and the known root cause feature vector; wherein the data types include discrete data and continuous data.
The correlation performance index feature vector is divided into a continuous type and a discrete type, the continuous type correlation performance index feature vector comprises 9 categories including uplink poor quality, downlink poor quality, uplink error code, downlink error code, RRC reestablishment, time delay, PRB utilization rate, signaling channel utilization rate and switching, and each category comprises a plurality of continuous type feature vectors. The discrete correlation performance index feature vector is the abnormal reason of signaling establishment failure, and comprises RRC connection establishment failure times (UE has no response), E _ RAB establishment failure times (core network problems and transport layer problems) and RRC establishment congestion times caused by limited radio resources; the abnormal E _ RAB release times (radio layer problem, core network problem, base station problem and handover problem), the same-frequency handover failure times, the UE context release times of eNodeB _ handover failure, the UE release caused by the loss of the wireless connection of the UE, the license limited times of connected users and the service quit time of an LTE cell.
In the technical scheme, the associated performance index data, the marked key performance index data and the known root cause performance index data are converted into a form of a characteristic vector, the characteristic vector is subjected to data preprocessing, the associated performance index characteristic vector is subjected to attribute marking according to discrete data and continuous data, and a root cause positioning data set is constructed according to the associated performance index characteristic vector, the marked key performance index characteristic vector and the known root cause characteristic vector.
An embodiment of the present application provides a root cause positioning method for key performance index abnormality, which performs cause difference calculation on discrete index data in a root cause positioning data set, and obtains a cause of the key performance index abnormality, specifically including:
calculating the difference value between the mean value of the discrete associated performance index data in a plurality of periods after the key performance index is abnormal and the mean value of the discrete associated performance index data in a plurality of periods before the key performance index is abnormal aiming at each discrete associated performance index data; and taking the discrete correlation performance index with the difference value larger than the preset difference value threshold value as the reason of the abnormal key performance index.
For example: the number of abnormal release of the E _ RAB (radio layer problem) and the average number of RRC connection are discrete eigenvectors, in 3 periods before and after the occurrence of the abnormality of the key performance index, the average value of the number of abnormal release of the E _ RAB (radio layer problem) in the 3 periods after the occurrence of the abnormality of the key performance index is 170.67, the average value of the number of abnormal release of the E _ RAB (radio layer problem) in the 3 periods before the occurrence of the abnormality of the key performance index is 1.33, and the difference value between the two is 169.34. The average number of RRC connections in the last 3 cycles was 15.77, the average number of RRC connections in the first 3 cycles was 17.89, and the difference between the two was-2.12. The preset difference threshold is zero, and in the two discrete type characteristic value vectors, only the difference between the last 3 periods and the first 3 periods of the abnormal E _ RAB release times is greater than zero, so that the abnormal E _ RAB release times are output, and the abnormal E _ RAB release times (wireless layer problems) are the reasons of the abnormal key performance indexes in the discrete type associated performance indexes.
In the above technical solution, cause difference calculation is performed on the discrete associated performance index data in the root cause positioning data set, the mean value of the discrete associated performance index data in a plurality of periods after occurrence of the key performance index abnormality is used as a reduced number, the discrete associated performance index data in a plurality of periods before occurrence of the key performance index abnormality is used as a reduced number, the difference value of the mean value of the discrete associated performance index data in a plurality of periods before occurrence of the key performance index abnormality is calculated, and when the difference value is greater than a preset difference threshold value, the corresponding discrete associated performance index is the cause of the key performance index abnormality.
As shown in fig. 3, an embodiment of the present application provides a method for locating a root cause of an abnormal key performance indicator, which performs multiple correlation calculations on continuous associated performance indicator data in a root cause location data set to obtain a root cause of an abnormal key performance indicator, and specifically includes:
s301, carrying out correlation calculation on the continuous correlation performance index data in the root cause positioning data set and the marked key performance index data to obtain continuous correlation performance index data with a correlation coefficient absolute value larger than or equal to a height correlation threshold value, and setting the continuous correlation performance index data with the correlation coefficient absolute value larger than the height correlation threshold value as abnormal cause index data.
In this step, a pearson correlation coefficient is calculated between the consecutive correlation performance indicator data and the marked key performance indicator data, the pearson correlation coefficient value being between-1 and 1. When the Pearson correlation coefficient is positive, the two data are in positive correlation, and when the Pearson correlation coefficient is negative, the two data are in negative correlation. And setting a corresponding high correlation threshold value according to the type of the continuous correlation performance index, wherein when the absolute value of the Pearson correlation coefficient is greater than or equal to the high correlation threshold value and the continuous correlation performance index is in positive correlation or negative correlation with the marked key performance index data, the continuous correlation performance index is the reason of the abnormal key performance index. And setting one eigenvector with the largest absolute value of the Pearson correlation coefficient of each category in the continuous correlation performance index as abnormal reason index data.
S302, performing correlation calculation on the known root cause performance index data and the abnormal cause index data in the root cause positioning data set to obtain the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to a highly correlated threshold value, wherein the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to the highly correlated threshold value is the root cause of the key performance index abnormality.
In this step, the communication logical link is divided into two directions, i.e., an uplink direction and a downlink direction, and a pearson correlation coefficient between the root cause performance index data known in the same link direction and the abnormality cause index data calculated in S301 is calculated. When the absolute value of the pearson correlation coefficient is greater than or equal to the high correlation threshold and the known root cause performance index data and the abnormal cause index data are in positive correlation or negative correlation, the corresponding known root cause performance index data are the root causes of the key performance index abnormality.
For example: in 3 periods before and after the call drop rate of the key performance index is abnormal, the uplink initial HARQ retransmission rate in the uplink error code category is greater than the Pearson correlation coefficient of the continuous correlation performance index and the key performance index and is in positive correlation, namely the uplink initial HARQ retransmission rate in the uplink error code category is the reason of the highest abnormal correlation of the key performance index; the known root cause performance index data and the uplink initial HARQ retransmission ratio feature vector in the uplink error code category are subjected to correlation calculation, and the sampling point proportion of TA more than or equal to 0 and less than 0.5Km and the proportion of the periodic reference signal receiving power more than or equal to-105 dBm are calculated to be the root cause of the abnormal key performance index.
In the technical scheme, two-step correlation calculation is carried out on continuous correlation performance index data in the root cause positioning data set, the correlation between the continuous correlation performance index and the key performance index is calculated in the first step, the reason with the highest abnormal correlation with the key performance index is calculated, and the reason tracing reasoning is completed. And secondly, calculating the correlation between the known root cause performance index and the associated performance index with the highest correlation, calculating the root cause with the highest correlation, and acquiring the root cause with abnormal key performance indexes.
As shown in fig. 4, an embodiment of the present application provides a root cause locating apparatus 100 for a critical performance indicator abnormality, including:
the feature obtaining module 101 selects the marked original data set by using a machine learning model, selects a plurality of important performance indicators, and filters the plurality of important performance indicators according to whether to participate in the calculation of the key performance indicators, so as to obtain associated performance indicators.
The index acquisition module 102 is configured to associate performance index data, marked key performance index data, and known root cause performance index data in multiple cycles before and after an abnormality occurs in a key performance index of a radio access network, and construct a root cause location data set;
the processing module 103 is used for calculating a reason difference value of the discrete associated performance index data in the root cause positioning data set to obtain the reason of the abnormal key performance index;
the processing module 103 performs multiple correlation calculations on the continuous associated performance index data in the root cause positioning data set, and obtains the root cause of the key performance index abnormality.
In an embodiment, the feature obtaining module 101 is specifically configured to:
constructing an original data set based on all counters, performance indexes and key performance index data in a wireless access network communication flow measured by key performance indexes in an observation time period;
marking the key performance index data in the original data set according to the abnormal threshold value to obtain marked key performance index data;
and selecting the marked original data set by using a machine learning model, selecting a plurality of important performance indexes, and screening the plurality of important performance indexes according to whether the marked original data set participates in the calculation of the key performance indexes to obtain the associated performance indexes.
In an embodiment, the index obtaining module 102 is specifically configured to:
performing data conversion on the associated performance index data, the marked key performance index data and the known root performance index data to obtain an associated performance index feature vector, a marked key performance index feature vector and a known root performance index feature vector;
performing data preprocessing on the associated performance index feature vector, the marked key performance index feature vector and the known root factor feature vector;
performing attribute marking on the associated performance index feature vector according to the data type, and constructing a root cause positioning data set based on the associated performance index feature vector, the marked key performance index feature vector and the known root cause feature vector; wherein the data types include discrete type data and continuous type data.
In an embodiment, the processing module 103 is specifically configured to:
calculating the difference value between the mean value of the discrete correlation performance index data in a plurality of periods after the occurrence of the abnormity and the mean value of the discrete correlation performance index data in a plurality of periods before the occurrence of the abnormity aiming at each discrete correlation performance index data;
and taking the discrete correlation performance index with the difference value larger than the preset difference value threshold value as the reason of the abnormal key performance index.
In an embodiment, the processing module 103 is specifically configured to:
carrying out correlation calculation on the continuous correlation performance index data in the root cause positioning data set and the marked key performance index data to obtain continuous correlation performance index data of which the correlation coefficient absolute value is greater than or equal to a height correlation threshold value, and setting the continuous correlation performance index data of which the correlation coefficient absolute value is greater than or equal to the height correlation threshold value as abnormal cause index data;
and performing correlation calculation on the known root cause performance index data and the abnormal cause index data in the root cause positioning data set to obtain the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to the high correlation threshold, wherein the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to the high correlation threshold is the root cause of the abnormal key performance index.
As shown in fig. 5, an embodiment of the present application provides an electronic device 200, where the electronic device 200 includes a memory 201 and a processor 202.
Wherein the memory 201 is used for storing computer instructions executable by the processor;
the processor 202, when executing computer instructions, performs the steps of the method in the embodiments described above. Reference may be made in particular to the description relating to the method embodiments described above.
Alternatively, the memory 201 may be separate or integrated with the processor 202. When the memory 201 is provided separately, the electronic device further includes a bus for connecting the memory 201 and the processor 202.
The embodiment of the present application further provides a computer-readable storage medium, in which computer instructions are stored, and when the processor executes the computer instructions, the steps in the method in the foregoing embodiment are implemented.
Embodiments of the present application further provide a computer program product, which includes computer instructions, and when the computer instructions are executed by a processor, the computer instructions implement the steps of the method in the above embodiments.
Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It will be understood that the present application is not limited to the precise arrangements that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (10)
1. A root cause positioning method for key performance index abnormity is characterized by comprising the following steps:
selecting a marked original data set by using a machine learning model, selecting a plurality of important performance indexes, and screening the important performance indexes according to whether the important performance indexes participate in the calculation of the key performance indexes to obtain associated performance indexes;
constructing a root cause positioning data set based on the associated performance index data, the marked key performance index data and the known root cause performance index data in a plurality of periods before and after the key performance index of the wireless access network is abnormal;
calculating a reason difference value of the discrete associated performance index data in the root cause positioning data set to obtain the reason of the abnormal key performance index;
and performing multiple times of correlation calculation on the continuous correlation performance index data in the root cause positioning data set to obtain the root cause of the key performance index abnormality.
2. The root cause positioning method according to claim 1, wherein performing multiple correlation calculations on continuous associated performance indicator data in the root cause positioning data set to obtain a root cause with an abnormal key performance indicator includes:
performing correlation calculation on the continuous correlation performance index data in the root cause positioning data set and the marked key performance index data to obtain continuous correlation performance index data of which the correlation coefficient absolute value is greater than or equal to a high correlation threshold value, and setting the continuous correlation performance index data of which the correlation coefficient absolute value is greater than or equal to the high correlation threshold value as abnormal cause index data;
and performing correlation calculation on the known root cause performance index data and the abnormal cause index data in the root cause positioning data set to obtain known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to a highly correlated threshold, wherein the known root cause performance index data of which the correlation coefficient absolute value is greater than or equal to the highly correlated threshold is a root cause of the abnormal key performance index.
3. The root cause positioning method according to claim 1, wherein performing cause difference calculation on discrete associated performance indicator data in the root cause positioning dataset to obtain the cause of the key performance indicator abnormality specifically includes:
calculating the difference value between the mean value of the discrete correlation performance index data in a plurality of periods after the key performance index is abnormal and the mean value of the discrete correlation performance index data in a plurality of periods before the key performance index is abnormal aiming at each discrete correlation performance index data;
and taking the discrete correlation performance index with the difference value larger than the preset difference value threshold value as the reason of the abnormal key performance index.
4. The root cause localization method according to claim 1 or 2, wherein before obtaining the associated performance indicators, the method further comprises:
constructing an original data set based on all counters, performance indexes and key performance index data in a wireless access network communication flow measured by key performance indexes in an observation time period;
and marking the key performance index data in the original data set according to an abnormal threshold value, acquiring marked key performance index data, and acquiring a marked original data set.
5. The method according to claim 1 or 2, wherein the constructing of the root cause location data set based on the associated performance indicator data, the marked key performance indicator data and the known root cause performance indicator data in a plurality of periods before and after the occurrence of the anomaly in the key performance indicator of the radio access network specifically comprises:
performing data conversion on the associated performance index data, the marked key performance index data and the known root performance index data to obtain an associated performance index feature vector, a marked key performance index feature vector and a known root performance index feature vector;
performing data preprocessing on the correlation performance index feature vector, the marked key performance index feature vector and the known root cause feature vector;
performing attribute marking on the correlation performance index feature vector according to the data type, and constructing a root cause positioning data set based on the correlation performance index feature vector, the marked key performance index feature vector and the known root cause feature vector; wherein the data types include discrete data and continuous data.
6. The method of claim 1, wherein the known root cause performance indicator data comprises one or more combinations of uplink weak coverage, downlink weak coverage, over coverage, uplink interference strength, and downlink interference rate.
7. A root cause positioner that key performance index is unusual, its characterized in that includes:
the characteristic acquisition module is used for selecting the marked original data set by using a machine learning model, selecting a plurality of important performance indexes, and screening the important performance indexes according to whether the important performance indexes participate in the calculation of the key performance indexes to acquire associated performance indexes;
the index acquisition module is used for constructing a root cause positioning data set based on the associated performance index data, the marked key performance index data and the known root cause performance index data in a plurality of periods before and after the key performance index of the wireless access network is abnormal;
the processing module is used for calculating the reason difference of the discrete associated performance index data in the root cause positioning data set to obtain the reason of the abnormal key performance index;
and the processing module is used for carrying out multiple times of correlation calculation on the continuous correlation performance index data in the root cause positioning data set so as to obtain the root cause with abnormal key performance indexes.
8. An electronic device, comprising: a processor, and a memory communicatively coupled to the processor;
the memory stores computer execution instructions;
the processor executes the memory-stored computer-executable instructions to implement the method of root cause localization of key performance indicator anomalies as claimed in any one of claims 1 to 6.
9. A computer-readable storage medium having stored thereon computer-executable instructions for implementing the method for root cause localization of key performance indicator anomalies of any one of claims 1 to 6 when executed by a processor.
10. A computer program product, comprising a computer program which, when executed by a processor, implements the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211136228.4A CN115550977A (en) | 2022-09-19 | 2022-09-19 | Root cause positioning method and equipment for key performance index abnormity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211136228.4A CN115550977A (en) | 2022-09-19 | 2022-09-19 | Root cause positioning method and equipment for key performance index abnormity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115550977A true CN115550977A (en) | 2022-12-30 |
Family
ID=84727547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211136228.4A Pending CN115550977A (en) | 2022-09-19 | 2022-09-19 | Root cause positioning method and equipment for key performance index abnormity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115550977A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024164765A1 (en) * | 2023-02-10 | 2024-08-15 | 中兴通讯股份有限公司 | System exception causal relation acquisition method and apparatus, and system exception root cause localization method and apparatus |
-
2022
- 2022-09-19 CN CN202211136228.4A patent/CN115550977A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024164765A1 (en) * | 2023-02-10 | 2024-08-15 | 中兴通讯股份有限公司 | System exception causal relation acquisition method and apparatus, and system exception root cause localization method and apparatus |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101964985B (en) | Coverage and capacity self-optimization device of self-organization network in LTE/LTE-A and method thereof | |
| WO2017041406A1 (en) | Failure positioning method and device | |
| WO2016090961A1 (en) | Method and device for network associations analysis | |
| CN112243249B (en) | LTE new access anchor point cell parameter configuration method and device under 5G NSA networking | |
| US9756518B1 (en) | Method and apparatus for detecting a traffic suppression turning point in a cellular network | |
| EP3183903B1 (en) | Received signal strength based interferer classification of cellular network cells | |
| CN107947968B (en) | Method and device for processing network quality complaint information | |
| Mfula et al. | Adaptive root cause analysis for self-healing in 5G networks | |
| US20230198640A1 (en) | Channel state information values-based estimation of reference signal received power values for wireless networks | |
| CN111263389A (en) | Method and device for automatically positioning Volte voice quality problem | |
| CN115550977A (en) | Root cause positioning method and equipment for key performance index abnormity | |
| CN111343647B (en) | Method, apparatus, device and medium for user perception evaluation | |
| CN110875825B (en) | Fault judgment method and device | |
| CN112584407B (en) | LTE user complaint qualitative method and device based on space-time combination | |
| KR20210068106A (en) | Data transmission method and device | |
| CN114698016A (en) | Network fault detection method and device for poor-quality cell | |
| CN115589599A (en) | Method and device for transmitting information | |
| Abd El-Aziz et al. | Cell outage detection and degradation classification based on alarms and kpi’s correlation | |
| CN114710801B (en) | Processing method and device of evaluation parameters of network performance, storage medium and equipment | |
| Khatib et al. | Modelling LTE solved troubleshooting cases | |
| EP4150861B1 (en) | Determining cell upgrade | |
| KR102724600B1 (en) | Method for monitoring wireless network quality and apparatus therefor | |
| WO2022149149A1 (en) | Artificial intelligence with dynamic causal model for failure analysis in mobile communication network | |
| Cilínio et al. | Explainable Fault Analysis in Mobile Networks: A SHAP-Based Supervised Clustering Approach | |
| JP7083130B2 (en) | Communication management system and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |