Positioning method and device with poor user perception quality, electronic equipment and storage medium

Abstract

The application provides a positioning method and device for user perceived quality difference, electronic equipment and storage medium, and relates to the technical field of communication. The method comprises the following steps: acquiring multiple candidate quality difference perception indexes contained in multiple bills of numbers of users; determining a quality difference perception index from the multiple candidate quality difference perception indexes according to a correlation coefficient of the multiple candidate quality difference perception indexes; performing convergent calculation on the quality difference perception indexes contained in the multiple bills of the numbers of users respectively to obtain a perception index table of the numbers of users; and positioning a resident cell of the user perceived quality difference based on the perception index table of the numbers of users.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a positioning method, apparatus, electronic device, and storage medium with poor user-perceived positioning quality. Background Technology

[0002] With the rapid development of mobile communication networks, mobile communication has permeated every aspect of users' lives, from clothing and food to housing and transportation. Users' XDR (X Detailed Record) call details record various service categories such as instant messaging, online shopping, video, music, games, cloud storage, and live streaming. Each service category can be further subdivided into hundreds of specific application names, resulting in an exceptionally large amount of XDR call detail record data.

[0003] In related technologies, user-perceived quality defects can be identified based on users' XDR call detail record (CDR) data. However, because XDR CDR data has numerous fields and an exceptionally large volume, a big data platform needs to be built to store and process the XDR CDR data, consuming significant storage space and computing resources. Furthermore, identifying user-perceived quality defects based on users' XDR CDR data suffers from problems such as short data retention time and low query efficiency. Summary of the Invention

[0004] In view of the above problems, embodiments of this application provide a positioning method, apparatus, electronic device, and storage medium with poor user-perceived positioning quality, so as to overcome the above problems or at least partially solve the above problems.

[0005] A first aspect of this application provides a method for locating user-perceived poor quality, including: Obtain multiple candidate quality perception indicators from multiple call detail records (CDRs) for each user number; Based on the cross-correlation coefficients of the multiple candidate quality defect perception indicators, the quality defect perception indicator is determined from the multiple candidate quality defect perception indicators; The perception indicators of poor quality contained in multiple call detail records of each user number are aggregated and calculated to obtain the perception indicator table of each user number. Based on the perception index table for each user number, locate the cell where the user perception quality is poor.

[0006] Optionally, the step of locating the cells with poor user perception quality based on the perception index table for each user number includes: Obtain the perceived quality difference threshold for each user number; Based on the perceived quality difference threshold of each user number and the perceived index table of each user number, multiple user numbers with poor perceived quality are identified from the user numbers, and the perceived index table of the multiple user numbers with poor perceived quality is identified as the perceived quality difference table. Based on the proportion of user numbers with poor perceived quality to the total number of user numbers in each cell, multiple cells with poor perceived quality are identified.

[0007] Optionally, it also includes: The dwell locations in the plurality of perception quality poor tables are matched with the plurality of perception quality poor cells; Based on the matching results, the corresponding poor-quality cells for each of the multiple user numbers with poor perceived quality are determined.

[0008] Optionally, the step of clustering multiple perception quality poor tables to determine multiple perception quality poor cells includes: Cluster the multiple perceptual quality difference tables to determine the perceptual grid corresponding to each of the multiple perceptual quality difference tables; The perception quality of each perception grid is determined based on the proportion of the perception quality difference table corresponding to each perception grid to the perception index table corresponding to that perception grid. Perception grids with perception quality below a threshold are identified as poor perception quality grids.

[0009] Optionally, it also includes: Based on the dwell time in the perception index table of each user number in the perception quality poor grid and the building types contained in the perception quality poor grid, the grid type of the perception quality poor grid is determined. The grid type includes at least one or more of the following: workplace grid, residential grid, and transit route grid.

[0010] Optionally, obtaining the perceived quality difference threshold for each user number includes: Separate the perception time series of each user number from the perception index table of each user number; Decompose the perceived time series of each user number to obtain the random fluctuation time series of each user number; Obtain the variance of the random fluctuation time series of each user number; The perceived quality difference threshold for each user number is determined based on the variance of the random fluctuation time series of each user number.

[0011] Optionally, it also includes: Obtain a network quality report, which includes at least one or more of the following: cell operating parameters, user-level MR data, cell-level KPI performance indicators, and user-level CDR data; By matching the occurrence time, user number, and cell of the perceived quality defect in the perceived quality defect table with the network quality report, the cause of the quality defect can be obtained.

[0012] Optionally, the quality perception indicators include at least one or more of the following: latency indicators, traffic indicators, packet loss rate indicators, out-of-order rate indicators, and perception rate indicators. The step of aggregating and calculating the perceived quality issues contained in multiple call detail records (CDRs) for each user number to obtain a perceived quality index table for each user number includes: For each user number, the latency-related indicators, packet loss rate indicators, out-of-order rate indicators, and perceived rate-related indicators contained in the multiple call detail records of that user number are averaged respectively, and / or the traffic-related indicators contained in the multiple call detail records of that user number are summed to obtain the perceived indicator table for that user number.

[0013] A second aspect of this application provides a positioning device with poor user-perceived positioning quality, comprising: The acquisition module is used to acquire multiple candidate quality difference perception indicators contained in multiple call detail records of each user number; The determination module is used to determine the quality defect perception index from the multiple candidate quality defect perception indices based on the cross-correlation coefficient of the multiple candidate quality defect perception indices. The calculation module is used to aggregate and calculate the quality difference perception indicators contained in multiple call detail records of each user number to obtain the perception indicator table of each user number. The positioning module is used to locate the cells with poor user perception quality based on the perception index table of each user number.

[0014] A third aspect of this application provides an electronic device, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the user-perceived quality difference positioning method as described in the first aspect.

[0015] A fourth aspect of this application provides a computer-readable storage medium that, when instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform the user-perceived quality difference positioning method as described in the first aspect.

[0016] A fifth aspect of this application provides a computer program product, including a computer program that, when executed by a processor, implements the user-perceived quality difference positioning method as described in the first aspect.

[0017] The embodiments of this application have the following advantages: In this embodiment, multiple candidate quality-poor perception indicators are obtained from multiple call detail records (CDRs) for each user number. Based on the cross-correlation coefficients of these candidate indicators, a quality-poor perception indicator is determined from them. The quality-poor perception indicators contained in each of the multiple CDRs for each user number are aggregated and calculated to obtain a perception indicator table for each user number. Based on the perception indicator table for each user number, the cell where the user experiences quality-poor perception is located is identified. In this way, quality-poor perception indicators closely related to the user's perceived quality-poor condition are identified from the complex CDR data. Aggregating and calculating these indicators to obtain the perception indicator table simplifies the data in the CDRs, thereby saving storage space and computing resources. Furthermore, because the perception indicator table has a relatively small data volume, it can have a longer retention time. In addition, querying only the perception indicator table significantly improves query efficiency. Therefore, using the perception indicator table to locate the cell where the user experiences quality-poor perception effectively improves efficiency. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the steps of a user-perceived poor quality positioning method in an embodiment of this application; Figure 2 This is a schematic diagram of the decomposition of the perceived time series of the user number in an embodiment of this application; Figure 3 This is a flowchart illustrating a user-perceived poor quality positioning method according to an embodiment of this application; Figure 4 This is a block diagram of a positioning device with poor user perception quality according to an embodiment of this application. Detailed Implementation

[0020] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0021] In related technologies, methods for locating poor user perception quality mainly include user feedback methods and methods using call detail records (CDRs) to locate poor user perception quality. User feedback methods have several technical problems, including low correlation between actual user perception and on-site CQT (Call Quality Test) tests or community KPIs (Key Performance Indicators), high investment of human and material resources, long optimization time and low efficiency, and a lag in perception optimization work that prevents proactive user care in advance.

[0022] Using call detail records (CDRs) to pinpoint user-perceived quality issues has the following problems: The original CDRs involve numerous fields, making it difficult for business personnel to prioritize checking which directly related user perception metrics when encountering such issues; the large volume of original CDRs limits the platform's shared layer capacity, allowing only 3 days of data retention; and the large data volume also leads to low query efficiency and long waiting times; the lack of a defined standard for poor perception quality and a quality filtering mechanism makes it impossible to quickly and accurately identify and analyze CDRs with poor perception quality from a single user's full CDR data.

[0023] The user-perceived poor quality location method proposed in this application's embodiments filters quality perception indicators that can directly reflect user perception, formulates a perception quality perception threshold for each user number, aggregates perception data by dimensions such as user, cell, service, and time (hour / day / week / month), and introduces a machine learning clustering algorithm to mark repeated quality quality problems in adjacent cells within the same time period within the statistical period, thereby quickly locating user-perceived poor quality problems.

[0024] Reference Figure 1 As shown, a flowchart illustrating the steps of a user-perceived poor quality positioning method in an embodiment of this application is illustrated. Figure 1 As shown, the method for locating poor user-perceived quality may specifically include the following steps: Step S11: Obtain multiple candidate quality difference perception indicators contained in multiple call detail records of each user number.

[0025] This document organizes user plane data tables for 4G and 5G data services. It includes HTTP (Hypertext Transfer Protocol), HTTPS (Hypertext Transfer Protocol Secure), and Video tables for 4G data services, and HTTP, HTTPS, and Video tables for 5G data services. Since the data for 4G and 5G data services is collected by different core network elements from different sources, these data are stored in separate tables. The HTTP table stores call detail records (CDRs) based on the HTTP protocol (unencrypted), the HTTPS table stores CDRs based on the HTTPS protocol (encrypted), and the Video table stores CDRs in video format. The CDRs in the Video table are extracted from the HTTP table and have the URL (Uniform Resource Locator) suffix ending in video format. Preprocessing of the data in each table allows for the removal of perceptual fields with significant data gaps, poor consistency, or large fluctuations through data observation.

[0026] Candidate quality-poor perception indicators are metrics closely related to user perception. These can be determined based on experience, by identifying the worst-quality indicators across multiple complaint call detail records (CDRs) within the same time period, or by comparing indicators across multiple complaint and non-complaint CDRs. Once candidate quality-poor perception indicators are determined, multiple CDRs for each user number can be retrieved to identify these indicators.

[0027] Candidate metrics for perceiving poor quality can include metrics for browsing services and metrics for evaluating video services. Metrics for browsing services include: TCP (Transmission Control Protocol) pipeline quality (e.g., TCP handshake latency, TCP uplink / downlink retransmission rate), webpage first-screen loading latency, and service duration (e.g., latency for green code pop-up in QR code scanning services). Metrics for evaluating video services include: video rate, video bitrate, and video buffering duration.

[0028] Step S12: Determine the quality poor perception index from the multiple candidate quality poor perception indices based on the cross-correlation coefficients of the multiple candidate quality poor perception indices.

[0029] The perceived quality index is an indicator that affects user perception. It can be determined by calculating the cross-correlation coefficients of multiple candidate perceived quality indices, and identifying those with cross-correlation coefficients greater than a threshold value. This threshold value can be set according to specific circumstances, and can be 0.6.

[0030] Step S13: Aggregate and calculate the quality difference perception indicators contained in the multiple call detail records of each user number to obtain the perception indicator table of each user number.

[0031] For multiple call detail records (CDRs) for each user number, the perceived quality issues contained in each CDR can be aggregated and calculated to obtain a perception index table for that user number. The perception index table stores the indices obtained after the aggregation calculation.

[0032] The perception indicator table can be stored in the columnar database HBase. After aggregating and calculating multiple call detail records (CDRs), the amount of CDR data is compressed to about 1 / 30, the storage time of user-level CDRs can be extended to 1 month (the original storage time was 3 days), and the query speed is reduced from 100 minutes to 10 seconds.

[0033] Optionally, the perceived quality issues metrics may include at least one or more of the following: latency metrics, traffic metrics, packet loss rate metrics, out-of-order rate metrics, and perceived speed metrics. Aggregating and calculating the perceived quality issues metrics contained in the call detail records (CDRs) of a user number may include: averaging the latency metrics contained in multiple CDRs of the user number; averaging the packet loss rate metrics contained in multiple CDRs of the user number; averaging the out-of-order rate metrics contained in multiple CDRs of the user number; averaging the perceived speed metrics contained in multiple CDRs of the user number; and summing the traffic metrics contained in multiple CDRs of the user number.

[0034] By performing different convergence calculations on different quality difference perception indicators, the resulting indicators can be made to conform to the characteristics of each indicator, thereby obtaining an accurate perception indicator table.

[0035] Step S14: Based on the perception index table of each user number, locate the cell where the user perception quality is poor.

[0036] The user number perception index table stores indicators that affect user perception. Therefore, based on the user number perception index table, it is possible to determine which user numbers have poor user perception quality. Then, based on the dwell location and dwell time in the perception index table of user numbers with poor user perception quality, the dwell cell with poor user perception quality can be located.

[0037] Furthermore, compared to locating cells with poor user perception based on the complex data in the original call detail records, the indicators in the perception indicator table in this embodiment are directly related to user perception. Therefore, locating cells with poor perception based on the perception indicator table can avoid interference caused by indicators unrelated to user perception, thereby improving the accuracy of positioning.

[0038] The technical solution adopted in this application identifies quality perception indicators closely related to user perceived quality issues from the complex data of call detail records (CDRs). These quality perception indicators are then aggregated and calculated to obtain a perception indicator table, thus simplifying the data in the CDRs and saving storage space and computing resources. Furthermore, because the perception indicator table has a relatively small data volume, it can have a longer retention time. In addition, querying only the perception indicator table significantly improves query efficiency; therefore, using the perception indicator table to locate the cells where users experience poor perceived quality can effectively improve efficiency.

[0039] Optionally, based on the above technical solution, to locate the cell with poor perceived quality based on the perception index table of each user number, it is necessary to first determine whether the user number has poor perceived quality according to the user number's perception index table. Therefore, it is necessary to first obtain the perceived quality poor threshold for each user number.

[0040] Optionally, for each user number, the perception time series of the user number can be separated from the user number's perception index table. The value of each quality-poor perception index for the user number can be obtained second by second, thus obtaining the user number's perception time series. Since user numbers may not have business activities at every moment, there may be missing quality-poor perception indicators for some moments. The KNN (k-nearest neighbor) algorithm from machine learning can be used to find several historical data points that are most similar (proximity in time and geographical location) to the missing points to fill in the missing values. If missing values ​​still exist, linear interpolation methods can be used to supplement them.

[0041] The perceived index of a user's phone number is essentially a stationary time series. From an additive model perspective, this series is composed of three time series with different properties. The perceived time series of a user's phone number can be represented by the following formula: Y(t) = T(t) + S(t) + R(t), where Y(t) represents the perceived time series, T(t) represents the trend time series, S(t) represents the periodic time series, and R(t) represents the random fluctuation time series. Therefore, the random fluctuation time series of the user's phone number can be obtained from the perceived time series.

[0042] The STL (Seasonal and Trend decomposition using Loess) algorithm can be used to decompose the perceived time series of user numbers, obtaining a random fluctuation time series of user numbers. The STL algorithm can use a weekly sliding statistical window to decompose the perceived time series of user numbers, obtaining a random fluctuation time series of user numbers (unpredictable white noise with a mean of 0).

[0043] The variance σ of the random fluctuation time series of a user's phone number can be calculated. Based on the variance of the random fluctuation time series, the perceived quality difference threshold for that user's phone number can be determined. Optionally, 4 times σ can be taken as the perceived quality difference threshold for the user's phone number. The variance σ of the random fluctuation time series can be calculated using the following formula: Where n is the number of samples in the weekly sliding statistical window. These are the values ​​for each sample. It is the sample mean.

[0044] Figure 2 This is a schematic diagram illustrating the decomposition of the perceived time series of a user number according to an embodiment of this application. After decomposing the perceived time series of the user number using the STL algorithm, a trend time series, a periodic time series, and a random fluctuation time series can be obtained. After obtaining the random fluctuation time series, a 4x variance line (perceived quality difference threshold) can be determined. Based on the perceived quality difference threshold, poor quality call detail records (CDRs) with perceived abrupt changes at time points A, B, and C can be filtered out.

[0045] In this way, for each user number, the perceived quality difference threshold can be intelligently determined, realizing the determination of the user-level perceived quality difference threshold.

[0046] After determining the perceived quality difference threshold for user numbers, user numbers with poor perceived quality can be identified from multiple user numbers based on the perceived quality difference threshold and the indicators in the user number perception indicator table. Accordingly, the perception indicator table for user numbers with poor perceived quality can be defined as the perceived quality difference table. The perceived quality difference table is defined for user numbers whose traffic exceeds a fixed value and whose percentage of poor quality is greater than a fixed percentage. The user numbers corresponding to the perceived quality difference table are then identified as user numbers with poor perceived quality.

[0047] For example, a perception quality poor table could be defined as a table showing traffic exceeding 1Mb and a poor quality percentage exceeding 30%. Traffic exceeding 1Mb ensures that the service is initiated by the user, thus avoiding the analysis of backend services that do not affect the user's perception.

[0048] Multiple user numbers with poor perceived quality are aggregated to determine the cell where each user number is located. For each cell, the total number of user numbers in that cell and the number of user numbers with poor perceived quality within that cell are determined. The cell is classified as a poor perceived quality cell based on the proportion of user numbers with poor perceived quality to the total number of user numbers in that cell. If the proportion of user numbers with poor perceived quality to the total number of user numbers in a cell exceeds a certain threshold, then the cell is identified as a poor perceived quality cell.

[0049] In this way, based on the perception index table of each user number, the resident cells with poor user perception quality can be quickly and accurately located.

[0050] Optionally, in order to determine the cell to which each user number with poor perceived quality belongs, after locating the cell where the user is staying with poor perceived quality, the user's location in each table of poor perceived quality can be obtained, and the location in each table of poor perceived quality can be matched with the cell with poor perceived quality. Thus, based on the matching results, the cell corresponding to each user number with poor perceived quality can be determined.

[0051] In this way, the cell to which each user's number with poor perceived quality belongs is determined, achieving user-level positioning based on perceived quality.

[0052] Optionally, the user number's perception index table can be mapped to each perception grid. Clustering the user number's perception quality poor table can determine the perception grid corresponding to each perception quality poor table. The number of perception index tables in a perception grid and the number of quality poor indicators in that grid are obtained. Based on the proportion of perception quality poor tables to the total number of perception index tables corresponding to that grid, the perception quality of that grid can be determined. A higher proportion of perception quality poor tables to the total number of perception index tables corresponding to a grid indicates a lower quality grid.

[0053] Perception grids with perceived quality below a threshold can be identified as poor-quality grids. The threshold can be set based on actual conditions. Perception quality can be characterized by the proportion of the poor-quality table corresponding to a given grid to the total perceived index table corresponding to that grid. For example, a perception grid with a poor-quality table of less than or equal to 10% of its corresponding perceived index table can be identified as a good-quality grid; a perception grid with a poor-quality table of less than 10% to 30% of its corresponding perceived index table can be identified as a medium-quality grid; and a perception grid with a poor-quality table of less than 30% of its corresponding perceived index table can be identified as a good-quality grid with poor-quality perception.

[0054] Thus, after determining the perceived quality difference grid, the location of the perceived quality difference can be pinpointed to the location of the grid. The location of the quality difference grid can be considered as the location of its centroid. By determining the location of the centroid of the quality difference grid, the user's quality difference area can be quickly located.

[0055] Optionally, the method for determining the sensing grid can be: obtaining the dwell position and dwell time in the sensing index table for each user number; and determining multiple sensing grids according to the dwell position and dwell time in the sensing index table for each user number. Machine learning clustering algorithms can be used to cluster sensing index tables that satisfy adjacent dwell times and adjacent dwell positions, thus dividing the data into multiple sensing grids. Machine learning clustering algorithms can use K-means clustering, K-means++ clustering, hierarchical clustering, DBSCAN clustering, or other clustering algorithms selected by the user. The distance metric used in the clustering algorithm is Euclidean distance.

[0056] Sensing grids possess both temporal and location attributes. Based on the dwell location and dwell time in multiple sensing indicator tables corresponding to the sensing grid, the grid category can be determined. Sensing grids are typically categorized into three types: workplace grids, residential grids, and traffic route grids. Specifically, workplace grids generally have dwell locations in office buildings and similar locations, with dwell times typically during the day; residential grids generally have dwell locations in residential buildings and similar locations, with dwell times typically at night; and traffic route grids generally exhibit greater variation in both dwell location and dwell time.

[0057] The grid type of the perception quality poor grid can be determined based on the dwell time in the perception index table of each user number in the perception quality poor grid and the building types contained in the perception quality poor grid. The grid type includes at least one or more of the following: workplace grid, residential grid, and transit grid.

[0058] After determining multiple sensing grids, the dwell positions and dwell times corresponding to multiple sensing quality difference tables can be matched with the time and position of multiple sensing grids to determine the sensing grids corresponding to the sensing quality difference tables.

[0059] In this way, multiple sensing grids can be determined based on the dwell position and dwell time of multiple sensing index tables, thereby dividing each sensing index table and sensing quality difference table into each sensing grid.

[0060] Optionally, based on the above technical solution, after the location of poor user-perceived quality is achieved, in order to facilitate troubleshooting by staff, the cause of poor user-perceived quality can also be determined.

[0061] Network quality reports can be obtained, which may include at least one or more of the following: cell engineering parameters, user-level MR (Measurement Report) data, cell-level KPI performance indicators, and user-level CDR (Call Detail Records) data.

[0062] By using the MR quality poor standard, the causes of poor coverage perception can be located; by using the cell KPI quality poor standard, the causes of poor cell load / interference can be located; by using the causes of abnormal events such as RRC (Radio Resource Control) reconstruction / Erab (Evolved Radio Access Bearer) drop / handover failure in user-level CDR, the causes can be located and associated with user perception.

[0063] By matching the time of occurrence, user number, and cell where the perceived quality is poor with the network quality report, the cause of the poor quality can be determined.

[0064] In this way, the cause of the quality problem can be identified, which helps staff to quickly eliminate the cause of the quality problem in the community, thereby improving the user experience.

[0065] Alternatively, as an example, Figure 3 This is a flowchart illustrating a method for locating user-perceived poor quality according to an embodiment of this application. The method for locating user-perceived poor quality mainly includes a perception data preparation stage S1, a perception data processing stage S2, a poor quality location stage S3, and a poor quality cause association stage S4.

[0066] During the data preparation phase, user plane data tables for 4G and 5G data services are compiled. Through data observation, perception fields with significant data gaps, poor consistency, or large fluctuations are removed, and quality-poor perception indicators are identified. Emerging service identifiers are captured through packet sniffing, forming an emerging service identifier library. Call detail records (CDRs) with key business characteristics are extracted from the host field of the HTTPS table or the URL field of the HTTP table. Quality-poor perception indicators belonging to emerging services can be labeled with emerging service identifiers, allowing for greater focus on these indicators during subsequent processing.

[0067] During the perception data processing phase, based on the identified quality-poor perception indicators, call detail record (CDR) data is aggregated and calculated according to dimensions such as user number, hour, network type (4G / 5G), cell ECI (E-UTRAN Cell Identifier), and service type (service_type, categorized by major service category and sub-categories). Daily, weekly, and monthly aggregation tables are then created based on the hourly aggregation table to facilitate statistical analysis of user-level perception data across the entire network over long statistical periods. User-level CDRs are marked with emerging service identifiers, and user-level perception indicators are aggregated based on these new service identifiers. The previously scattered HTTP, HTTPS, and Video tables for 4G / 5G users are merged and integrated to obtain a user-level perception indicator table. Queries are then performed within this user-level perception indicator table, eliminating the need to query the three perception tables separately and then merge the results, thus improving query efficiency.

[0068] The previous article disclosed the aggregation calculation of the perceived quality difference index of each call detail record based on the user number dimension, but similar methods can be used to aggregate and calculate the index based on other dimensions.

[0069] In the quality defect localization phase, the STL algorithm is used to decompose the user's perceived time series, and four times the standard deviation of the randomly fluctuating time series is taken to determine the user-level perceived quality defect threshold. Cells with user call detail records (CDRs) exceeding 1Mb and having a quality defect rate greater than 30% are screened for quality defects. A machine learning clustering algorithm is introduced to statistically analyze the cells with user CDRs in the same time period within a certain period, defining duplicate user quality defect cells to quickly locate user quality defect issues.

[0070] During the quality degradation cause correlation phase, by correlating user numbers, service cells, and service time information of the perceived quality degradation grid, corresponding provincial 4 / 5G cell parameters, user-level MR data, cell-level KPI performance indicators, and user-level CDR data can be queried. Using user-level MR quality degradation standards (RSRP average less than -115dBm / RSRP excellent rate less than 80% / RSRQ average less than -14dB / RSRQ excellent rate less than 80%), coverage-related quality degradation causes (weak coverage or overlapping coverage) are identified. Using cell KPI quality degradation standards (RRC connected users greater than 200 / downlink PRB utilization greater than 80% / uplink PRB utilization greater than 30% / CQI excellent rate less than 80%), cell load / interference-related quality degradation causes are identified. Using abnormal events such as RRC reconstruction / Erab disconnection / handover failure in user-level CDRs, radio interface signaling quality degradation causes are identified.

[0071] The technical solution adopted in this application has the following advantages: 1. No need to regularly obtain a large number of user survey samples; only a small amount of user feedback data is needed to verify the accuracy of the perceived poor quality raster. 2. A brand-new user-level perception model and supporting data processing flow were designed, reducing call detail record (CDR) capacity by more than 30 times and CDR query time by more than 300 times, achieving long-term storage and fast query of user-level perception data. 3. Separating the user-level perception time series allows for the assessment of perception degradation issues at the hourly level for a single user; 4. Accurately locate clusters of poor perception quality for a single user (hourly / cell-level) and automatically associate the causes of poor perception quality (coverage-related, capacity-related, interference-related, equipment alarm-related). 5. The sensing grid adopts a binary master form of dwell time and dwell location, eliminating the need for latitude and longitude backfilling of sensing data.

[0072] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this application are not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of this application.

[0073] Figure 4 This is a block diagram of a positioning device with poor user-perceived quality according to an embodiment of this application, such as... Figure 4 As shown, the device includes: an acquisition module, a determination module, a calculation module, and a positioning module, wherein: The acquisition module is used to acquire multiple candidate quality difference perception indicators contained in multiple call detail records of each user number; The determination module is used to determine the quality defect perception index from the multiple candidate quality defect perception indices based on the cross-correlation coefficient of the multiple candidate quality defect perception indices. The calculation module is used to aggregate and calculate the quality difference perception indicators contained in multiple call detail records of each user number to obtain the perception indicator table of each user number. The positioning module is used to locate the cells with poor user perception quality based on the perception index table of each user number.

[0074] Optionally, the positioning module includes: A threshold acquisition unit is used to acquire the perception quality difference threshold for each user number. The poor quality user determination unit is used to determine multiple poor quality user numbers from the user numbers based on the perceived poor quality threshold of each user number and the perceived index table of each user number, and to determine the perceived index table of the multiple poor quality user numbers as a perceived poor quality table. The poor quality cell identification unit is used to identify multiple poor quality cells based on the proportion of user numbers with poor perceived quality to the total number of user numbers in that cell.

[0075] Optionally, it also includes: The matching module is used to match the dwell locations in the plurality of poor perception quality tables with the plurality of poor perception quality cells; The poor-quality cell determination module is used to determine the poor-quality cells corresponding to each of the multiple perceived poor-quality user numbers based on the matching results.

[0076] Optionally, the poor-quality cell determination unit includes: Clustering subunit is used to cluster the plurality of perceptual quality difference tables and determine the perceptual grid corresponding to each of the plurality of perceptual quality difference tables. The proportion determination subunit is used to determine the perception quality of each perception grid according to the proportion of the perception quality difference table corresponding to each perception grid to the perception index table corresponding to that perception grid. The comparison sub-unit is used to identify perceptual grids with perceptual quality below a threshold as perceptual quality poor grids.

[0077] Optionally, it also includes: The type determination unit is used to determine the grid type of the perception quality poor grid based on the dwell time in the perception index table of each user number in the perception quality poor grid and the building type contained in the perception quality poor grid. The grid type includes at least one or more of the following: workplace grid, residential grid, and transit route grid.

[0078] Optionally, the threshold acquisition unit includes: A separation subunit is used to separate the perception time series of each user number from the perception index table of each user number; The decomposition subunit is used to decompose the perceived time series of each user number to obtain the random fluctuation time series of each user number; The variance acquisition subunit is used to acquire the variance of the random fluctuation time series of each user number; The threshold determination subunit is used to determine the perceived quality difference threshold of each user number based on the variance of the random fluctuation time series of each user number.

[0079] Optionally, it also includes: The report acquisition module is used to acquire network quality reports, which include at least one or more of the following: cell operating parameters, user-level MR data, cell-level KPI performance indicators, and user-level CDR data. The association module is used to associate the occurrence time, user number, and stationed cell of the perceived quality defect in the perceived quality defect table with the network quality report to obtain the cause of the quality defect.

[0080] Optionally, the quality perception indicators include at least one or more of the following: latency indicators, traffic indicators, packet loss rate indicators, out-of-order rate indicators, and perception rate indicators. The computing module includes: The calculation unit is used to perform average processing on the latency-related indicators, packet loss rate indicators, out-of-order rate indicators, and perceived rate-related indicators contained in multiple call detail records (CDRs) for each user number, and / or to perform sum processing on the traffic-related indicators contained in multiple CDRs for each user number, to obtain a perceived indicator table for the user number.

[0081] It should be noted that the system implementation and the method implementation are similar, so the description is relatively simple. For relevant details, please refer to the method implementation.

[0082] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0083] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0084] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, electronic devices, and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart... Figure 1One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0085] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction set implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0087] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0088] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0089] The above provides a detailed description of a user-perceived poor positioning method, apparatus, electronic device, and storage medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for locating poor user-perceived quality, characterized in that, include: Obtain multiple candidate quality perception indicators from multiple call detail records (CDRs) for each user number; Based on the cross-correlation coefficients of the multiple candidate quality defect perception indicators, the quality defect perception indicator is determined from the multiple candidate quality defect perception indicators; Using the user number as the core dimension, cross-call detail record aggregation calculation is performed on the quality difference perception indicators contained in multiple call detail records of each user number to obtain the perception indicator table of each user number. Based on the perception index table of each user number, the method for locating the cell with poor perceived quality includes: obtaining the perception quality poor threshold for each user number; determining multiple user numbers with poor perceived quality from the user numbers according to the perception quality poor threshold and the perception index table of each user number, and determining the perception index table of the multiple user numbers with poor perceived quality as a perception quality poor table; and determining multiple cells with poor perceived quality according to the proportion of user numbers with poor perceived quality in each cell to the total number of user numbers in that cell. The step of obtaining the perception quality difference threshold for each user number includes: separating the perception time series of each user number from the perception index table of each user number; decomposing the perception time series of each user number using the STL algorithm to obtain the random fluctuation time series of each user number; obtaining the variance of the random fluctuation time series of each user number; and using the variance of the random fluctuation time series of each user number as a preset multiple as the perception quality difference threshold for each user number.

2. The method according to claim 1, characterized in that, Also includes: The dwell locations in the plurality of perception quality poor tables are matched with the plurality of perception quality poor cells; Based on the matching results, the corresponding poor-quality cells for each of the multiple user numbers with poor perceived quality are determined.

3. The method according to claim 1, characterized in that, Clustering multiple perception quality poor tables identified several perception quality poor cells, including: Cluster the multiple perceptual quality difference tables to determine the perceptual grid corresponding to each of the multiple perceptual quality difference tables; The perception quality of each perception grid is determined based on the proportion of the perception quality difference table corresponding to each perception grid to the perception index table corresponding to that perception grid. Perception grids with perception quality below a threshold are identified as poor perception quality grids.

4. The method according to claim 3, characterized in that, Also includes: Based on the dwell time in the perception index table of each user number in the perception quality poor grid and the building types contained in the perception quality poor grid, the grid type of the perception quality poor grid is determined. The grid type includes at least one or more of the following: workplace grid, residential grid, and transit route grid.

5. The method according to any one of claims 1-4, characterized in that, Also includes: Obtain a network quality report, which includes at least one or more of the following: cell operating parameters, user-level MR data, cell-level KPI performance indicators, and user-level CDR data; By matching the occurrence time, user number, and cell of the perceived quality defect in the perceived quality defect table with the network quality report, the cause of the quality defect can be obtained.

6. The method according to any one of claims 1-4, characterized in that, The quality perception indicators include at least one or more of the following: latency indicators, traffic indicators, packet loss rate indicators, out-of-order rate indicators, and perception rate indicators. The step of aggregating and calculating the perceived quality issues contained in multiple call detail records (CDRs) for each user number to obtain a perceived quality index table for each user number includes: For each user number, the latency-related indicators, packet loss rate indicators, out-of-order rate indicators, and perceived rate-related indicators contained in the multiple call detail records of that user number are averaged respectively, and / or the traffic-related indicators contained in the multiple call detail records of that user number are summed to obtain the perceived indicator table for that user number.

7. A positioning device with poor user-perceived positioning quality, characterized in that, include: The acquisition module is used to acquire multiple candidate quality difference perception indicators contained in multiple call detail records of each user number; The determination module is used to determine the quality defect perception index from the multiple candidate quality defect perception indices based on the cross-correlation coefficient of the multiple candidate quality defect perception indices. The calculation module is used to perform cross-call detail record aggregation calculation on the quality difference perception indicators contained in multiple call detail records of each user number, with the user number as the core dimension, to obtain the perception indicator table of each user number. The positioning module is used to locate the user's cell with poor perceived quality based on the perception index table of each user number, including: obtaining the perception quality poor threshold for each user number; determining multiple user numbers with poor perceived quality from the user numbers according to the perception quality poor threshold and the perception index table of each user number, and determining the perception index table of the multiple user numbers with poor perceived quality as a perception quality poor table; determining multiple cells with poor perceived quality according to the proportion of user numbers with poor perceived quality in each cell to the total number of user numbers in that cell; The step of obtaining the perception quality difference threshold for each user number includes: separating the perception time series of each user number from the perception index table of each user number; decomposing the perception time series of each user number using the STL algorithm to obtain the random fluctuation time series of each user number; obtaining the variance of the random fluctuation time series of each user number; and using the variance of the random fluctuation time series of each user number as a preset multiple as the perception quality difference threshold for each user number.

8. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the user-perceived quality difference positioning method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the user-perceived quality difference positioning method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the user-perceived quality difference positioning method as described in any one of claims 1 to 6.

Images