Method and device for analyzing root causes of poor quality of Internet dedicated lines
By conducting big data analysis and model identification of DPI call list data of Internet dedicated lines, the problems of low coverage and large volatility of timed sampling data are solved, and accurate analysis and rapid positioning of the root causes of poor quality of Internet dedicated lines are achieved.
Patent Information
- Application Number
- CN202111142756.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-09-28
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2041-09-28
AI Technical Summary
In the existing Internet dedicated line quality poor analysis methods, the data coverage rate of timed sampling is low and the volatility is high, resulting in inaccuracy of the quality poor positioning results.
By conducting big data analysis on the DPI call sheet data of the Internet dedicated line, the quality index characteristics of the dedicated line dimension and multiple transmission dimensions are obtained, and the preset identification model is used to identify the quality special line and determine the root cause of the quality poor.
The accuracy and efficiency of the root cause analysis of poor quality is improved, and the quality of Internet dedicated lines can be more comprehensively reflected, and the transmission dimensions and root cause of poor quality are accurately determined.
Smart Images

Figure CN115883322B_ABST
Abstract
Description
Technical Field
[0001] This document relates to the field of computer technology, and in particular to a method and device for analyzing the root causes of poor quality of Internet dedicated lines. Background Art
[0002] At present, the root cause analysis of Internet dedicated line quality problems is mainly carried out by sampling the relevant data of Internet dedicated lines at regular intervals to analyze the network quality and locate and delimit the quality problems. However, due to the low data coverage and high volatility of the sampled data, this will inevitably affect the accuracy of the final quality problem location and delimitation results.
[0003] Therefore, there is an urgent need for a solution that can accurately analyze the root causes of poor quality of Internet dedicated lines. Summary of the invention
[0004] The purpose of the embodiments of this specification is to provide a method and device for analyzing the root causes of poor quality of Internet dedicated lines, so as to accurately analyze the root causes of poor quality of Internet dedicated lines.
[0005] In order to achieve the above purpose, the embodiments of this specification adopt the following technical solutions:
[0006] In a first aspect, a method for analyzing root causes of poor quality of an Internet dedicated line is provided, comprising:
[0007] Obtain the deep packet inspection (DPI) call bill data of the Internet dedicated line;
[0008] Performing big data analysis on the DPI call list data to obtain quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions;
[0009] Based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, identifying whether the Internet dedicated line is a poor quality dedicated line, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line;
[0010] If the Internet dedicated line is a poor quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into a preset transmission quality identification model to identify the poor quality transmission dimension corresponding to the Internet dedicated line, wherein the transmission quality identification model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line;
[0011] Based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension, the root cause of the poor quality of the Internet dedicated line is determined.
[0012] In a second aspect, a device for analyzing root causes of poor quality of an Internet dedicated line is provided, comprising:
[0013] A DPI call record acquisition unit is used to acquire the deep packet inspection DPI call record data of the Internet dedicated line;
[0014] An indicator feature acquisition unit, used to perform big data analysis on the DPI call list data to obtain quality indicator features of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions;
[0015] A dedicated line quality analysis unit, used to identify whether the Internet dedicated line is a poor quality dedicated line based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line;
[0016] a transmission dimension analysis unit, for inputting the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions into a preset transmission quality recognition model if the Internet dedicated line is a poor quality dedicated line, so as to identify the poor quality transmission dimension corresponding to the Internet dedicated line, wherein the transmission quality recognition model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line;
[0017] The poor quality root cause determination unit is used to determine the poor quality root cause of the Internet dedicated line based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension.
[0018] In a third aspect, an electronic device is provided, including:
[0019] processor;
[0020] a memory for storing instructions executable by the processor;
[0021] The processor is configured to execute the instructions to implement the method as described in the first aspect.
[0022] According to a fourth aspect, a computer-readable storage medium is provided. When instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the method described in the first aspect.
[0023] The scheme of the embodiments of this specification obtains the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions by performing big data analysis on the DPI call bill data of the Internet dedicated line. Since the DPI call bill data is continuous and complete full data, it better solves the problems of low coverage and high volatility of the timed sampling data, so that the obtained quality indicator characteristics of the dedicated line dimension and multiple transmission dimensions are more accurate and can more comprehensively reflect the dedicated line quality of the Internet dedicated line and the transmission quality of different dimensions; based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and the preset dedicated line quality identification model, it is identified whether the Internet dedicated line is a poor quality dedicated line. Since the quality indicator characteristics of the dedicated line dimension can It accurately and comprehensively reflects the quality of Internet dedicated lines. Through the quality indicator characteristics, it can be accurately identified whether there is a quality problem in the Internet dedicated line. If the Internet dedicated line is a poor quality line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into the preset transmission quality identification model to identify the poor quality transmission dimensions corresponding to the Internet dedicated line. Since the quality indicator characteristics of multiple transmission dimensions can reflect the transmission quality of the Internet dedicated line from multiple dimensions, these quality indicator characteristics can accurately locate the transmission dimensions that cause the poor quality problem of the Internet dedicated line. Further, based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimensions, the root cause of the poor quality of the Internet dedicated line can be quickly and accurately located. Furthermore, since the poor quality dedicated line identification model is pre-trained based on the quality indicator characteristics of the sample dedicated lines in the same dedicated line dimension and a first label that identifies whether it is a poor quality dedicated line, and the transmission quality identification model is pre-trained based on the quality indicator characteristics of the poor quality sample dedicated lines in the same multiple transmission dimensions and a second label that identifies its poor quality transmission dimension, it is only necessary to input corresponding data into the dedicated line quality identification model and the transmission quality identification model to automatically determine the root cause of the poor quality of the Internet dedicated line, thereby improving the efficiency of analyzing the root cause of the poor quality of the Internet dedicated line. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings:
[0025] Figure 1 A flowchart of a method for analyzing root causes of poor quality of an Internet dedicated line provided in accordance with an embodiment of the present specification;
[0026] Figure 2 A schematic diagram of the structure of a transmission quality identification model provided for one embodiment of this specification;
[0027] Figure 3 A partial structural diagram of a transmission quality identification model provided for one embodiment of this specification;
[0028] Figure 4 A schematic diagram of a device for analyzing root causes of poor quality of an Internet dedicated line provided in accordance with an embodiment of the present specification;
[0029] Figure 5 A schematic diagram of the structure of an electronic device provided for one embodiment of the present specification. DETAILED DESCRIPTION
[0030] In order to make the purpose, technical solutions and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in combination with the specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.
[0031] The terms "first", "second", etc. in this specification and claims are used to distinguish similar objects, and are not used to describe a particular order or precedence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of this specification can be implemented in an order other than those illustrated or described herein. In addition, "and / or" in this specification and claims means at least one of the connected objects, and the character " / " generally means that the objects associated with each other are in an "or" relationship.
[0032] In order to accurately analyze the root causes of poor quality of Internet dedicated lines, the embodiments of this specification provide a method for analyzing the root causes of poor quality of Internet dedicated lines. By performing big data analysis on the DPI call list data of the Internet dedicated line, the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions are obtained. Since the DPI call list data is continuous and complete full data, it better solves the problems of low coverage and high volatility of timed sampling data, so that the obtained quality indicator characteristics in the dedicated line dimension and multiple transmission dimensions are more accurate and can more comprehensively reflect the dedicated line quality of the Internet dedicated line and the transmission quality in different dimensions; based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and the preset dedicated line quality identification model, it is identified whether the Internet dedicated line is a poor quality dedicated line. Line, since the quality indicator characteristics of the dedicated line dimension can accurately and comprehensively reflect the dedicated line quality of the Internet dedicated line, the quality indicator characteristics can be used to accurately identify whether the Internet dedicated line has a quality problem; if the Internet dedicated line is a poor quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into the preset transmission quality identification model to identify the poor quality transmission dimensions corresponding to the Internet dedicated line. Since the quality indicator characteristics of multiple transmission dimensions can reflect the transmission quality of the Internet dedicated line from multiple dimensions, the transmission dimensions that cause the poor quality problem of the Internet dedicated line can be accurately located through these quality indicator characteristics. Further, based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimensions, the root cause of the poor quality of the Internet dedicated line can be quickly and accurately located. Furthermore, since the poor quality dedicated line identification model is pre-trained based on the quality indicator characteristics of the sample dedicated lines in the same dedicated line dimension and a first label that identifies whether it is a poor quality dedicated line, and the transmission quality identification model is pre-trained based on the quality indicator characteristics of the poor quality sample dedicated lines in the same multiple transmission dimensions and a second label that identifies its poor quality transmission dimension, it is only necessary to input corresponding data into the dedicated line quality identification model and the transmission quality identification model to automatically determine the root cause of the poor quality of the Internet dedicated line, thereby improving the efficiency of analyzing the root cause of the poor quality of the Internet dedicated line.
[0033] It should be understood that the root cause analysis method for poor quality of Internet dedicated lines provided in the embodiments of this specification can be executed by an electronic device or software installed in an electronic device, and can be specifically executed by a terminal device or a server device.
[0034] The technical solutions provided by the embodiments of this specification are described in detail below in conjunction with the accompanying drawings.
[0035] Please refer to Figure 1 , is a flow chart of a method for analyzing root causes of poor quality of Internet dedicated lines provided by an embodiment of this specification, and the method may include:
[0036] S102, obtaining DPI call record data of the Internet dedicated line.
[0037] Deep Packet Inspection (DPI) is a deep inspection technology based on data packets. It performs deep inspection on different network application layer loads (such as HTTP, DNS, etc.) and determines the legitimacy of the message by inspecting the payload of the message. In the embodiment of this specification, by deploying probes at the core network node of the Internet dedicated line exit, the DPI technology is used to perform protocol analysis on the traffic service content of the Internet dedicated line and generate DPI call bill data for the Internet dedicated line.
[0038] In the embodiments of this specification, in order to improve the quality of DPI bill data, thereby improving the accuracy of subsequent root cause analysis of poor quality of Internet dedicated lines, the collected DPI bill data can also be cleaned. Specifically, the DPI bill data can be quality checked to identify fields in the DPI bill data that have null values, illegal values, unreasonable values, etc., and these fields can be filtered out to achieve screening and optimization of the quality of DPI bill data.
[0039] S104, performing big data analysis on the DPI call record data to obtain quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions.
[0040] Specifically, the Extract Transform Load (ETL) technology can be used to perform big data analysis on the DPI call data, thereby obtaining the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions. The method of using ETL technology to perform big data analysis is a method known to those skilled in the art and will not be repeated here.
[0041] Of course, in actual applications, by performing big data analysis on DPI call record data, the quality indicator characteristics of Internet dedicated lines in other dimensions (such as province, city, industry, etc.) can also be obtained, and this specification does not limit this.
[0042] In the embodiments of this specification, the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension are used to characterize the quality of the Internet dedicated line in the dedicated line dimension. Specifically, the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension may include the indicator values of the Internet dedicated line in multiple dedicated line quality indicators, wherein the multiple dedicated line quality indicators may include but are not limited to: web page access success rate (web_succ), web page access delay (web_delay), download rate (dn_rate), video pause times (video_pause), video playback success rate (video_succ) and video first frame delay (video_fdelay), etc.
[0043] The quality indicator characteristics of the Internet dedicated line in a single transmission dimension are used to characterize the quality of the Internet dedicated line in that transmission dimension. In order to accurately and comprehensively reflect the end-to-end transmission quality of the Internet dedicated line, so as to more accurately locate the factors affecting the poor quality of the Internet dedicated line in the future, the multiple transmission dimensions may include but are not limited to: transmission performance dimension, service provider (SP) dimension, core network element dimension, home broadband optical line (Optical Line Terminal, OLT) dimension, optical network unit (Optical Network Unit, ONU) dimension, broadband remote access server (Broadband Remote Access Server, BRAS) dimension, etc.
[0044] S106, based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, identify whether the Internet dedicated line is a poor quality dedicated line.
[0045] The dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and the corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line.
[0046] In the embodiments of this specification, any appropriate method may be used to evaluate whether an Internet dedicated line is a poor quality dedicated line. In an optional implementation, the index values of the Internet dedicated line in multiple dedicated line quality indicators may be combined to evaluate the dedicated line quality difference score that characterizes the quality of the Internet dedicated line in the dedicated line dimension, and further determine whether the Internet dedicated line is a poor quality dedicated line based on the dedicated line quality difference score. Specifically, the above-mentioned dedicated line quality identification model may include a fitting function of the mapping relationship between the weighted sum of the index differences that characterize each dedicated line quality indicator and the dedicated line quality difference score, wherein the index difference of the dedicated line quality indicator is the difference between the index value of the dedicated line quality indicator and the corresponding index threshold. Accordingly, the above-mentioned S106 may include the following steps:
[0047] Step A1, determining a quality difference value of the Internet dedicated line based on the fitting function and the index values of the Internet dedicated line in multiple dedicated line quality indicators.
[0048] For example, the above fitting function is shown in the following formula (1):
[0049]
[0050] Wherein, F(x) represents the quality difference score of the Internet dedicated line; web_succ represents the web page access success rate of the Internet dedicated line, x0 represents the index threshold corresponding to the web page access success rate, and a0 represents the weight corresponding to the web page access success rate; web_delay represents the web page access delay of the Internet dedicated line, represents the index threshold x1 corresponding to the web page access delay, and a1 represents the weight corresponding to the web page access delay; dn_rate represents the download rate of the Internet dedicated line, x2 represents the index threshold corresponding to the download rate, and a2 represents the weight corresponding to the download rate; video_pause represents the number of video pauses of the Internet dedicated line, x3 represents the index threshold corresponding to the number of video pauses, and a3 represents the weight corresponding to the number of video pauses; video_succ represents the video playback success rate of the Internet dedicated line, x4 represents the index threshold corresponding to the video playback success rate, and a4 represents the weight corresponding to the video playback success rate; video_fdelay represents the video first frame delay of the Internet dedicated line, x5 represents the index threshold corresponding to the video first frame delay, and a5 represents the weight corresponding to the video first frame delay.
[0051] It should be noted that the indicator thresholds and weights corresponding to each dedicated line quality indicator can be set according to actual needs, or can be obtained by training the quality indicator features and corresponding first labels of a large number of sample dedicated lines in the dedicated line dimension. The latter method can enable the fitting function to more accurately evaluate the dedicated line quality difference value of the Internet dedicated line. The latter method is introduced below.
[0052] In another embodiment, before the above step S106, the above method provided in the embodiment of this specification may further include:
[0053] Step B1, obtaining a first training data set, wherein the first training data set includes quality indicator features of multiple sample dedicated lines in the dedicated line dimension and corresponding first labels, and the first label of the sample dedicated line is the actual dedicated line quality difference value of the sample dedicated line.
[0054] Step B2, repeatedly perform the following iterative operation on the first training data set multiple times until the preset iteration stop condition is met: based on the fitting function used in this iteration and the quality indicator characteristics of multiple sample dedicated lines in the dedicated line dimension, predict the dedicated line quality difference values of multiple sample dedicated lines in this iteration; based on the prediction results of the dedicated line quality difference values of multiple sample dedicated lines in this iteration and the first training data set, adjust the parameters in the fitting function used in this iteration to obtain the fitting function used for the next iteration, wherein the parameters of the fitting function include the weights and indicator thresholds corresponding to each dedicated line quality indicator.
[0055] It should be noted that the iteration stopping condition in the embodiments of this specification may be that the number of iterations reaches a preset number or the prediction error obtained by predicting the first training data set through the fitting function is less than a preset error threshold, etc., and this embodiment of this specification does not make specific limitations on this.
[0056] It can be understood that the quality indicator characteristics of multiple sample dedicated lines in the dedicated line dimension and the corresponding first labels are used as training data sets. Based on the deep learning algorithm, the training data sets are repeatedly iterated multiple times and the parameters of the fitting function are continuously optimized during the iteration process, so that the fitting function finally trained can be used to accurately evaluate the dedicated line quality difference value of the Internet dedicated line.
[0057] In the process of performing the above-mentioned iterative operation, the parameters in the fitting function used in each iteration can be adjusted according to the step size. In the embodiment of the present specification, the step size can be preset according to actual needs, or it can be dynamically adjusted at each iteration. Specifically, in an optional implementation, in the above-mentioned step B2, based on the prediction results of the quality difference values of the multiple sample dedicated lines in this iteration and the first training data set, the parameters in the fitting function used in this iteration are adjusted, including:
[0058] Step B11, for a single sample dedicated line, based on the differences between each indicator threshold in the fitting function used in this iteration and the predicted results of the dedicated line quality difference value of the sample dedicated line, determine the candidate step length corresponding to the sample dedicated line in the next iteration.
[0059] More specifically, the candidate step size corresponding to the sample dedicated line in the next iteration can be determined according to the following formula (2):
[0060]
[0061] Among them, d(X,Y) represents the candidate step size corresponding to the sample dedicated line in the next iteration, X k represents the indicator threshold corresponding to the kth dedicated line quality indicator of the sample dedicated line, n represents the number of dedicated line quality indicators, and Y k It indicates the quality difference value of the sample dedicated line predicted by the fitting function.
[0062] It can be understood that the candidate step size determined in this way comprehensively considers the impact of each dedicated line quality indicator threshold in the fitting function used in each iteration on the dedicated line quality difference value output by the fitting function, and subsequent parameter adjustments based on this candidate step size can enable the fitting function to quickly reach the predetermined iteration stop condition, so as to obtain a fitting function that can accurately evaluate the dedicated line quality difference value.
[0063] Step B12, determining a target step length for the next iteration based on candidate step lengths corresponding to the multiple sample dedicated lines in the next iteration.
[0064] Optionally, the smallest candidate step length can be selected from the candidate step lengths corresponding to multiple sample dedicated lines in the next iteration as the target step length for the next iteration, or the average of the candidate step lengths corresponding to multiple sample dedicated lines in the next iteration can be used as the target step length for the next iteration. The embodiments of this specification do not make specific limitations on this.
[0065] Step B13: adjusting the parameters of the fitting function used in the current iteration based on the target step size of the next iteration to obtain the fitting function used in the next iteration.
[0066] After determining the target step size for the next iteration, the parameters in the sum function used in this iteration can be adjusted according to the target step size to obtain new parameters, thereby obtaining the fitting function used in the next iteration.
[0067] It should be noted that, in practical applications, the target step size obtained based on the candidate step size determined based on the above formula (2) can be used to adjust the indicator thresholds corresponding to each dedicated line quality indicator in the fitting function, and the weights corresponding to each dedicated line quality indicator in the fitting function can be adjusted based on human experience.
[0068] Of course, it is worth noting that in addition to the above-mentioned method for training the fitting function, in some other optional implementations, any other appropriate deep learning algorithm can be used to train the fitting function, such as the K-nearest neighbor (KNN) algorithm, and the embodiments of this specification do not specifically limit this.
[0069] It can be understood that by adjusting the parameters in the fitting function applicable to each iteration in the above manner, the fitting function can quickly reach the predetermined iteration stop condition and obtain a fitting function that can accurately evaluate the dedicated line quality difference value, thereby further improving the efficiency and accuracy of the entire Internet dedicated line quality difference root cause analysis process.
[0070] Step A2: If the quality difference score of the Internet dedicated line exceeds a preset quality difference threshold, the Internet dedicated line is determined to be a poor quality dedicated line.
[0071] It can be understood that in the above implementation, the quality of the Internet dedicated line is quantitatively evaluated using a fitting function, and whether the Internet dedicated line is a poor quality dedicated line can be accurately identified based on the quantitative evaluation result.
[0072] S108, if the Internet dedicated line is a poor quality dedicated line, input the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions into a preset transmission quality identification model to identify the poor quality transmission dimension corresponding to the Internet dedicated line.
[0073] Among them, the transmission quality identification model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line.
[0074] In the embodiments of this specification, the transmission quality identification model can adopt any appropriate machine learning model. Considering that the input data is the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions, and these quality indicator characteristics have complex correlations, in an optional implementation, the transmission quality identification model can adopt a deep neural network (DNN), and its specific structure is as follows: Figure 2 As shown, the transmission quality identification model includes an input layer (inputlayer), multiple hidden layers (hidden layer1, hidden layer2 and hidden layer3) and an output layer (outputlayer).
[0075] With respect to the transmission quality identification model in the above S108, the embodiments of this specification also include a method for training the transmission quality identification model. It is worth noting that the training of the transmission quality identification model is performed in advance based on the quality indicator characteristics of the collected poor quality sample dedicated lines in multiple transmission dimensions and the second label indicating the poor quality transmission dimension. In the subsequent process of identifying the poor quality transmission dimension of the Internet dedicated line, it is not necessary to train the transmission quality identification model every time, or the transmission quality identification model can be periodically updated based on the quality indicator characteristics of the newly collected poor quality sample dedicated lines in multiple transmission dimensions and the corresponding second label.
[0076] Specifically, the training method for the transmission quality recognition model includes:
[0077] Step C1, obtaining a second training data set, wherein the second training data set includes quality indicator features of the poor quality sample dedicated lines in multiple transmission dimensions and corresponding second labels.
[0078] Step C2, predicting the poor quality transmission dimension corresponding to the poor quality sample dedicated line through the forward propagation algorithm based on the network structure parameters of the transmission quality identification model and the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions.
[0079] For example, Figure 3The following is a partial structural diagram of a transmission quality identification model. Taking the transmission quality identification model as an example, the first layer (Layer L1) is the input data, that is, the quality index characteristics of the poor quality sample dedicated line in each transmission dimension (including x1 to x3). For the network structure parameters of the second layer (Layer L2) to It can be determined by the following formulas (3) to (5):
[0080]
[0081]
[0082]
[0083] Similarly, the network structure parameters of each layer can be determined in turn, and the final hw,b(x) is the quality difference transmission dimension.
[0084] Step C3, based on the poor quality transmission dimension and the second label corresponding to the poor quality sample dedicated line, adjust the network structure parameters of the transmission quality identification model until the preset training stop condition is met.
[0085] Specifically, based on the poor quality transmission dimension and the second label corresponding to the poor quality sample line, the prediction error of the transmission quality identification model for the second training data set can be determined, and the network structure parameters of each layer of the transmission quality identification model can be adjusted based on the prediction error, and the above process can be repeated until the preset training stop condition is met.
[0086] Among them, the preset training stop condition can be reaching a preset number of iterations or the prediction error of the transmission quality identification model for the second training data set is less than a preset error threshold, etc., which can be set according to actual needs and is not limited in this embodiment of the present specification.
[0087] It can be understood that the quality indicator characteristics of the poor quality dedicated line samples in multiple transmission dimensions and the corresponding second labels are used as training data, and deep learning is performed on the training data to establish a transmission quality identification model, so that the established transmission quality identification model can be used to accurately identify the poor quality transmission dimensions that cause poor quality problems of Internet dedicated lines.
[0088] S110, determining a root cause of poor quality of the Internet dedicated line based on quality indicator characteristics of the Internet dedicated line in a corresponding poor quality transmission dimension.
[0089] In specific implementation, various methods known to those skilled in the art may be used to analyze the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension to determine the root cause of the poor quality of the Internet dedicated line.
[0090] The method for analyzing the root causes of poor quality of Internet dedicated lines provided in the embodiments of this specification performs big data analysis on the DPI call bill data of the Internet dedicated line to obtain the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions. Since the DPI call bill data is continuous and complete full data, it better solves the problems of low coverage and high volatility of the timed sampling data, so that the obtained quality indicator characteristics of the dedicated line dimension and multiple transmission dimensions are more accurate and can more comprehensively reflect the dedicated line quality of the Internet dedicated line and the transmission quality of different dimensions; based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and the preset dedicated line quality identification model, it identifies whether the Internet dedicated line is a poor quality dedicated line. Since the quality of the dedicated line dimension The quality indicator characteristics can accurately and comprehensively reflect the quality of the Internet dedicated line. Through the quality indicator characteristics, it is possible to accurately identify whether the Internet dedicated line has a quality problem. If the Internet dedicated line is a poor-quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into a preset transmission quality identification model to identify the poor-quality transmission dimensions corresponding to the Internet dedicated line. Since the quality indicator characteristics of multiple transmission dimensions can reflect the transmission quality of the Internet dedicated line from multiple dimensions, the transmission dimensions that cause the poor-quality problems of the Internet dedicated line can be accurately located through these quality indicator characteristics. Further, based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor-quality transmission dimensions, the root cause of the poor-quality problem of the Internet dedicated line can be quickly and accurately located. Furthermore, since the poor quality dedicated line identification model is pre-trained based on the quality indicator characteristics of the sample dedicated lines in the same dedicated line dimension and a first label that identifies whether it is a poor quality dedicated line, and the transmission quality identification model is pre-trained based on the quality indicator characteristics of the poor quality sample dedicated lines in the same multiple transmission dimensions and a second label that identifies its poor quality transmission dimension, it is only necessary to input corresponding data into the dedicated line quality identification model and the transmission quality identification model to automatically determine the root cause of the poor quality of the Internet dedicated line, thereby improving the efficiency of analyzing the root cause of the poor quality of the Internet dedicated line.
[0091] In addition, Figure 1 Corresponding to the method for analyzing the root cause of poor quality of Internet dedicated lines shown in the figure, the embodiment of this specification also provides a device for analyzing the root cause of poor quality of Internet dedicated lines. FIG. 400 is a schematic diagram of the structure of a device for analyzing the root cause of poor quality of Internet dedicated lines 400 provided in the embodiment of this specification, including:
[0092] A DPI call list acquisition unit 410 is used to acquire deep packet inspection DPI call list data of the Internet dedicated line;
[0093] The indicator feature acquisition unit 420 is used to perform big data analysis on the DPI call list data to obtain the quality indicator features of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions respectively;
[0094] A dedicated line quality analysis unit 430, configured to identify whether the Internet dedicated line is a poor quality dedicated line based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, wherein the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line;
[0095] The transmission dimension analysis unit 440 is used for inputting the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions into a preset transmission quality identification model to identify the poor quality transmission dimension corresponding to the Internet dedicated line if the Internet dedicated line is a poor quality dedicated line. The transmission quality identification model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line;
[0096] The poor quality root cause determining unit 450 is used to determine the poor quality root cause of the Internet dedicated line based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension.
[0097] The root cause analysis device for poor quality of Internet dedicated lines provided in the embodiments of this specification obtains the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions by performing big data analysis on the DPI call list data of the Internet dedicated line. Since the DPI call list data is continuous and complete full data, it better solves the problems of low coverage and high volatility of the timed sampling data, so that the obtained quality indicator characteristics in the dedicated line dimension and multiple transmission dimensions are more accurate and can more comprehensively reflect the dedicated line quality of the Internet dedicated line and the transmission quality in different dimensions; based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and the preset dedicated line quality identification model, it identifies whether the Internet dedicated line is a poor quality dedicated line. Since the quality indicator characteristics of the dedicated line dimension The quality indicator characteristics can accurately and comprehensively reflect the quality of the Internet dedicated line. Through the quality indicator characteristics, it is possible to accurately identify whether the Internet dedicated line has a quality problem. If the Internet dedicated line is a poor-quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into a preset transmission quality identification model to identify the poor-quality transmission dimensions corresponding to the Internet dedicated line. Since the quality indicator characteristics of multiple transmission dimensions can reflect the transmission quality of the Internet dedicated line from multiple dimensions, the transmission dimensions that cause the poor-quality problems of the Internet dedicated line can be accurately located through these quality indicator characteristics. Further, based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor-quality transmission dimensions, the root cause of the poor-quality problem of the Internet dedicated line can be quickly and accurately located. Furthermore, since the poor quality dedicated line identification model is pre-trained based on the quality indicator characteristics of the sample dedicated lines in the same dedicated line dimension and a first label that identifies whether it is a poor quality dedicated line, and the transmission quality identification model is pre-trained based on the quality indicator characteristics of the poor quality sample dedicated lines in the same multiple transmission dimensions and a second label that identifies its poor quality transmission dimension, it is only necessary to input corresponding data into the dedicated line quality identification model and the transmission quality identification model to automatically determine the root cause of the poor quality of the Internet dedicated line, thereby improving the efficiency of analyzing the root cause of the poor quality of the Internet dedicated line.
[0098] Optionally, the quality indicator feature of the Internet dedicated line in the dedicated line dimension includes the indicator values of the Internet dedicated line in multiple dedicated line quality indicators respectively, and the dedicated line quality identification model includes a fitting function that characterizes the mapping relationship between the weighted sum of the indicator differences of each of the dedicated line quality indicators and the dedicated line quality difference value, and the indicator difference of the dedicated line quality indicator is the difference between the indicator value of the dedicated line quality indicator and the corresponding indicator threshold;
[0099] The dedicated line quality analysis unit 430 includes:
[0100] A quality difference value determination subunit, used to determine the quality difference value of the Internet dedicated line based on the fitting function and the index values of the Internet dedicated line in multiple dedicated line quality indicators;
[0101] The poor quality dedicated line identification subunit is used to determine that the Internet dedicated line is a poor quality dedicated line if the dedicated line quality difference score of the Internet dedicated line exceeds a preset poor quality threshold.
[0102] Optionally, the device 400 further includes:
[0103] A first training data acquisition unit is used to acquire a first training data set before the dedicated line quality analysis unit 430 identifies whether the Internet dedicated line is a poor-quality dedicated line based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, wherein the first training data set includes the quality indicator characteristics of multiple sample dedicated lines in the dedicated line dimension and the corresponding first labels, and the first labels are actual dedicated line quality difference scores of the sample dedicated lines;
[0104] An iteration unit is used to repeatedly perform the following iteration operation on the first training data set for multiple times until a preset iteration stop condition is met:
[0105] Based on the fitting function used in this iteration and the quality indicator characteristics of the multiple sample dedicated lines in the dedicated line dimension, predict the dedicated line quality difference values of the multiple sample dedicated lines in this iteration;
[0106] Based on the prediction results of the quality differential values of the multiple sample dedicated lines in this iteration and the first training data set, the parameters in the fitting function used in this iteration are adjusted to obtain the fitting function used in the next iteration, wherein the parameters include the weights and indicator thresholds corresponding to each of the dedicated line quality indicators.
[0107] Optionally, the iteration unit is specifically used for:
[0108] For a single sample dedicated line, based on the differences between each indicator threshold in the fitting function used in this iteration and the prediction result of the dedicated line quality difference value of the sample dedicated line, determine the candidate step length corresponding to the sample dedicated line in the next iteration;
[0109] Determine a target step length for the next iteration based on the candidate step lengths corresponding to the multiple sample dedicated lines in the next iteration;
[0110] The parameters in the fitting function used in the current iteration are adjusted based on the target step size of the next iteration to obtain the fitting function used in the next iteration.
[0111] Optionally, the device 400 further includes:
[0112] A second training data acquisition unit is used to acquire a second training data set before the transmission dimension analysis unit 440 inputs the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions into a preset transmission quality recognition model to identify the poor quality transmission dimension corresponding to the Internet dedicated line, wherein the second training data set includes the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels;
[0113] A poor quality transmission prediction unit, configured to predict the poor quality transmission dimension corresponding to the poor quality sample dedicated line through a forward propagation algorithm based on the network structure parameters of the transmission quality identification model and the quality indicator characteristics of each of the poor quality sample dedicated lines in multiple transmission dimensions;
[0114] The network structure parameter adjustment unit is used to adjust the network structure parameters of the transmission quality identification model based on the poor quality transmission dimension and the second label corresponding to the poor quality sample dedicated line until the preset training stop condition is met.
[0115] Optionally, the multiple dedicated line quality indicators include: web page access success rate, web page access delay, download rate, video freeze times, video playback success rate and video first frame delay.
[0116] Optionally, the multiple transmission dimensions include: transmission performance dimension, service provider SP dimension, core network element dimension, home broadband optical line terminal OLT dimension, optical network unit ONU dimension, and broadband remote access server BRAS dimension.
[0117] Obviously, the Internet dedicated line poor quality root cause analysis device of the embodiment of this specification can be used as the above Figure 1 The execution subject of the Internet dedicated line poor quality root cause analysis method shown in the figure can realize the Internet dedicated line poor quality root cause analysis method in Figure 1 The functions realized are as follows. Since the principles are the same, they will not be described here.
[0118] Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of this specification. Figure 5 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.
[0119] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0120] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.
[0121] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming an Internet dedicated line poor quality root cause analysis device at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations:
[0122] Obtain the deep packet inspection (DPI) call record data of the Internet dedicated line;
[0123] Performing big data analysis on the DPI call list data to obtain quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions;
[0124] Based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, identifying whether the Internet dedicated line is a poor quality dedicated line, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line;
[0125] If the Internet dedicated line is a poor quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into a preset transmission quality identification model to identify the poor quality transmission dimension corresponding to the Internet dedicated line, wherein the transmission quality identification model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line;
[0126] Based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension, the root cause of the poor quality of the Internet dedicated line is determined.
[0127] The above is as in this manual Figure 1 The method performed by the root cause analysis device for poor quality of Internet dedicated lines disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in software form. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of this specification can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of this specification can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in a decoding processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.
[0128] It should be understood that the electronic device of the embodiment of this specification can implement the Internet dedicated line poor quality root cause analysis device in Figure 1 Functions of the embodiments shown. Since the principles are the same, the embodiments of this specification will not be described in detail here.
[0129] Of course, in addition to software implementation, the electronic device of this specification does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.
[0130] The embodiment of the present specification also provides a computer-readable storage medium, which stores one or more programs, wherein the one or more programs include instructions, which, when executed by a portable electronic device including a plurality of application programs, enable the portable electronic device to execute Figure 1 The method of the embodiment shown is specifically used to perform the following operations:
[0131] Obtain the deep packet inspection (DPI) call record data of the Internet dedicated line;
[0132] Performing big data analysis on the DPI call list data to obtain quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions;
[0133] Based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, identifying whether the Internet dedicated line is a poor quality dedicated line, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line;
[0134] If the Internet dedicated line is a poor quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into a preset transmission quality identification model to identify the poor quality transmission dimension corresponding to the Internet dedicated line, wherein the transmission quality identification model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line;
[0135] Based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension, the root cause of the poor quality of the Internet dedicated line is determined.
[0136] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0137] In short, the above description is only a preferred embodiment of this specification and is not intended to limit the protection scope of this specification. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification shall be included in the protection scope of this specification.
[0138] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions.
[0139] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0140] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0141] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
Claims
1. A method for analyzing the root causes of poor quality of Internet dedicated lines, characterized in that: include: Obtain the deep packet inspection (DPI) call bill data of the Internet dedicated line; Performing big data analysis on the DPI call list data to obtain quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions; Based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, identifying whether the Internet dedicated line is a poor quality dedicated line, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line; If the Internet dedicated line is a poor-quality dedicated line, the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions are input into a preset transmission quality identification model to identify the poor-quality transmission dimension corresponding to the Internet dedicated line, and the transmission quality identification model is trained based on the quality indicator characteristics of the poor-quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor-quality transmission dimension corresponding to the poor-quality sample dedicated line; wherein the multiple transmission dimensions include: transmission performance dimension, service provider SP dimension, core network element dimension, home broadband optical line terminal OLT dimension, optical network unit ONU dimension, broadband remote access server BRAS dimension; Based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension, the root cause of the poor quality of the Internet dedicated line is determined.
2. The method according to claim 1, characterized in that: The quality index characteristics of the Internet dedicated line in the dedicated line dimension include the index values of the Internet dedicated line in multiple dedicated line quality indicators respectively, the dedicated line quality identification model includes a fitting function that characterizes the mapping relationship between the weighted sum of the index differences of each of the dedicated line quality indicators and the dedicated line quality difference value, and the index difference of the dedicated line quality indicator is the difference between the index value of the dedicated line quality indicator and the corresponding index threshold; The identifying whether the Internet dedicated line is a poor quality dedicated line based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model includes: Determine a quality difference value of the Internet dedicated line based on the fitting function and the index values of the Internet dedicated line in a plurality of dedicated line quality indicators; If the dedicated line quality difference score of the Internet dedicated line exceeds a preset quality difference threshold, the Internet dedicated line is determined to be a poor quality dedicated line.
3. The method according to claim 2, characterized in that Before identifying whether the Internet dedicated line is a poor-quality dedicated line based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, the method further includes: Obtain a first training data set, where the first training data set includes quality indicator characteristics of multiple sample dedicated lines in a dedicated line dimension and corresponding first labels, where the first labels are actual dedicated line quality difference scores of the sample dedicated lines; Repeat the following iterations on the first training data set until a preset iteration stop condition is met: Based on the fitting function used in this iteration and the quality indicator characteristics of the multiple sample dedicated lines in the dedicated line dimension, predict the dedicated line quality difference values of the multiple sample dedicated lines in this iteration; Based on the prediction results of the quality difference values of the multiple sample dedicated lines in this iteration and the first training data set, the parameters in the fitting function used in this iteration are adjusted to obtain the fitting function used in the next iteration, wherein the parameters include the weights and indicator thresholds corresponding to each of the dedicated line quality indicators.
4. The method according to claim 3, characterized in that The adjusting the parameters in the fitting function used in the current iteration based on the prediction results of the quality difference values of the multiple sample dedicated lines in the current iteration and the first training data set to obtain the fitting function used in the next iteration includes: For a single sample dedicated line, based on the differences between each indicator threshold in the fitting function used in this iteration and the prediction result of the dedicated line quality difference value of the sample dedicated line, determine the candidate step length corresponding to the sample dedicated line in the next iteration; Determine a target step length for the next iteration based on the candidate step lengths corresponding to the multiple sample dedicated lines in the next iteration; The parameters in the fitting function used in the current iteration are adjusted based on the target step size of the next iteration to obtain the fitting function used in the next iteration.
5. The method according to claim 1, characterized in that Before inputting the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions into a preset transmission quality identification model to identify the poor quality transmission dimension corresponding to the Internet dedicated line, the method further includes: Acquire a second training data set, where the second training data set includes quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and corresponding second labels; Predicting the quality difference transmission dimension corresponding to the quality difference sample dedicated line based on the network structure parameters of the transmission quality identification model and the quality indicator characteristics of each of the quality difference sample dedicated lines in multiple transmission dimensions through a forward propagation algorithm; Based on the poor quality transmission dimension and the second label corresponding to the poor quality sample dedicated line, the network structure parameters of the transmission quality identification model are adjusted until a preset training stop condition is met.
6. The method according to claim 2, characterized in that The multiple dedicated line quality indicators include: web page access success rate, web page access delay, download rate, video freeze times, video playback success rate and video first frame delay.
7. A root cause analysis device for poor quality of Internet dedicated lines, characterized in that: include: A DPI call record acquisition unit is used to acquire the deep packet inspection DPI call record data of the Internet dedicated line; An indicator feature acquisition unit, used to perform big data analysis on the DPI call list data to obtain quality indicator features of the Internet dedicated line in the dedicated line dimension and multiple transmission dimensions; A dedicated line quality analysis unit, used to identify whether the Internet dedicated line is a poor quality dedicated line based on the quality indicator characteristics of the Internet dedicated line in the dedicated line dimension and a preset dedicated line quality identification model, wherein the dedicated line quality identification model is trained based on the quality indicator characteristics of the sample dedicated line in the dedicated line dimension and a corresponding first label, and the first label is used to indicate whether the sample dedicated line is a poor quality dedicated line; A transmission dimension analysis unit, for inputting the quality indicator characteristics of the Internet dedicated line in multiple transmission dimensions into a preset transmission quality recognition model if the Internet dedicated line is a poor quality dedicated line, so as to identify the poor quality transmission dimension corresponding to the Internet dedicated line, wherein the transmission quality recognition model is trained based on the quality indicator characteristics of the poor quality sample dedicated line in multiple transmission dimensions and the corresponding second labels, and the second label is used to indicate the poor quality transmission dimension corresponding to the poor quality sample dedicated line; wherein the multiple transmission dimensions include: transmission performance dimension, service provider SP dimension, core network element dimension, home broadband optical line terminal OLT dimension, optical network unit ONU dimension, broadband remote access server BRAS dimension; The poor quality root cause determination unit is used to determine the poor quality root cause of the Internet dedicated line based on the quality indicator characteristics of the Internet dedicated line in the corresponding poor quality transmission dimension.
8. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that: When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the method as claimed in any one of claims 1 to 6.
Citation Information
Patent Citations
Network analysis method and device
CN108768702A
Method and device for analyzing poor quality of home broadband user
CN112702224A
Rating network security posture and comparing network maliciousness
US20160021141A1
Method and device for training multi-label classification model
WO2019100724A1
