WO2023116178A1 - 无线侧网络根因定位方法、运行控制装置及存储介质 - Google Patents

无线侧网络根因定位方法、运行控制装置及存储介质 Download PDF

Info

Publication number
WO2023116178A1
WO2023116178A1 PCT/CN2022/127442 CN2022127442W WO2023116178A1 WO 2023116178 A1 WO2023116178 A1 WO 2023116178A1 CN 2022127442 W CN2022127442 W CN 2022127442W WO 2023116178 A1 WO2023116178 A1 WO 2023116178A1
Authority
WO
WIPO (PCT)
Prior art keywords
kpi
value
kqi
predicted
root cause
Prior art date
Application number
PCT/CN2022/127442
Other languages
English (en)
French (fr)
Inventor
冯媛
杨翌晨
邵敏峰
李益刚
王东强
潘越洋
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2023116178A1 publication Critical patent/WO2023116178A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06—Management of faults, events, alarms or notifications
    • H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
    • H04W16/18—Network planning tools
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the present application relates to the technical field of network operation and maintenance, and in particular to a wireless-side network root cause location method, an operation control device, and a storage medium.
  • the daily operation and maintenance network of mobile operators includes wireless side cells, 4G and 5G core network elements.
  • the problems existing in the network are analyzed and evaluated.
  • Engineering implementation and other technical means can improve the network experience of mobile terminal users.
  • KPI Key Performance Indication
  • the goal of current system optimization is to improve the KPI on the wireless side, but the relationship between the KPI on the wireless side and the actual perception of users is not clear.
  • Embodiments of the present application propose a wireless-side network root cause location method, an operation control device, and a storage medium.
  • an embodiment of the present application provides a method for locating the root cause of the wireless side network, including: obtaining a preset mapping function model, wherein the mapping function model is used to establish a key performance indicator KPI value and a key quality indicator KQI value
  • the mapping relationship between obtain the KPI data of the target cell, the KPI data includes a plurality of KPI values; according to the mapping function model, calculate and obtain a plurality of KQI prediction values corresponding to the KPI values, wherein the KQI
  • the predicted value is used to reflect the impact of the corresponding KPI value on user perception; according to the mapping function model, multiple distribution-based KPI degradation inflection points are obtained; and the root cause KPI value is determined according to the KQI predicted value and the KPI degradation inflection point.
  • the embodiment of the present application provides an operation control device, including at least one control processor and a memory for communicating with the at least one control processor; the memory stores information that can be processed by the at least one control processor.
  • An instruction executed by a device the instruction is executed by the at least one control processor, so that the at least one control processor can execute the method for locating the root cause of the wireless side network as described in the embodiment of the first aspect above.
  • the embodiments of the present application provide a computer-readable storage medium, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to make the computer perform the above-mentioned first aspect embodiment.
  • the wireless side network root cause location method is used to make the computer perform the above-mentioned first aspect embodiment.
  • FIG. 1 is a flow chart of a method for locating the root cause of the wireless side network provided by an embodiment of the present application
  • FIG. 2 is a flow chart of a method for locating the root cause of the wireless side network provided by another embodiment of the present application;
  • FIG. 3 is a flow chart of a method for locating the root cause of the wireless side network provided by another embodiment of the present application.
  • FIG. 4 is a flow chart of a method for locating the root cause of the wireless side network provided by another embodiment of the present application.
  • FIG. 5 is a flow chart of a method for locating the root cause of the wireless side network provided by another embodiment of the present application.
  • FIG. 6 is a schematic diagram of a mapping function provided by another embodiment of the present application.
  • Fig. 7 is a schematic diagram of the KPI value of the target cell provided by another embodiment of the present application.
  • FIG. 8 is a schematic diagram of the ranking results of KQI prediction values provided by another embodiment of the present application.
  • FIG. 9 is a schematic diagram of marking and grouping KPI values according to preset service tags according to another embodiment of the present application.
  • Fig. 10 is a schematic diagram of a TCP index of a certain cell in Longsang Temple provided by another embodiment of the present application;
  • Fig. 11 is a schematic diagram of the root cause location results of Longsangsi Community provided by another embodiment of the present application.
  • Fig. 12 is a schematic structural diagram of an operation control device provided by another embodiment of the present application.
  • Embodiments of the present application provide a wireless-side network root cause location method, an operation control device, and a storage medium, which can quickly locate the root cause and help improve the user's actual experience perception.
  • the embodiment of the first aspect of the present application provides a method for locating the root cause of the network on the wireless side, including but not limited to steps S110 to S150:
  • Step S110 Obtain a preset mapping function model, wherein the mapping function model is used to establish a mapping relationship between KPI values and KQI values;
  • KQI Key Quality Indicators
  • KQI Key Quality Indicators
  • Step S120 Acquiring KPI data of the target cell, where the KPI data includes a plurality of KPI values;
  • the root cause of the target cell is identified, and KPI data of the target cell is first obtained, wherein the KPI data includes multiple KPI values.
  • KPI data can include the average uplink RLC SDU delay (ms) of the cell, PDCCH channel CCE occupancy rate, uplink bler, uplink HARQ retransmission ratio, uplink CCE allocation failure rate, MAC layer uplink residual block error rate, uplink
  • KQI data can include video effective download rate (kbps), video quality, video MOS, video TCP downlink RTT average delay, video TCP connection confirmation average delay, TCP connection success rate, web page Index values such as TCP uplink retransmission rate and webpage TCP downlink retransmission rate.
  • Step S130 According to the mapping function model, calculate and obtain a plurality of KQI prediction values corresponding to the KPI value, wherein the KQI prediction value is used to reflect the influence of the corresponding KPI value on user perception;
  • mapping function model includes multiple mapping functions.
  • the KPI value is input into the mapping function, and the corresponding KQI prediction value can be calculated.
  • the KQI prediction value can reflect the impact of the KPI value on user perception, which is convenient for rapid positioning.
  • Step S140 According to the mapping function model, a plurality of distribution-based KPI deterioration inflection points are obtained;
  • KPI degradation inflection points can be obtained, wherein the KPI degradation inflection point is a KPI critical value for measuring user perception, which can reflect the user's acceptance of network quality. It has been verified that the inflection point of KPI degradation combines the coverage of wireless base stations and user behavior differences in different cities or provinces, and has more practical guiding significance in the optimization of wireless networks.
  • Step S150 Determine the root cause KPI value according to the predicted KQI value and the KPI deterioration inflection point.
  • the root cause KPI value can be quickly located and user perception can be effectively reflected.
  • a targeted wireless index optimization strategy can be formulated, which is conducive to improving the actual experience of users perception.
  • multiple KQI prediction values can be obtained according to the KPI value of the target cell, and the KQI prediction value can reflect
  • the impact of KPI values on user perception is convenient for quickly locating KPI values with high influence.
  • multiple distribution-based KPI degradation inflection points can be obtained, and the root cause KPI value can be quickly located according to KQI prediction values and KPI degradation inflection points.
  • a targeted wireless index optimization strategy can be formulated, which is conducive to improving the user's actual experience perception.
  • the root cause KPI value is determined according to the KQI prediction value and the KPI degradation inflection point, including but not limited to steps S210 to step S230:
  • Step S210 Obtain a predicted KPI set according to the KQI predicted value and the KPI deterioration inflection point, wherein the predicted KPI set includes multiple first KPI predicted values that cause user perception abnormality, and the first KPI predicted value is a KQI greater than the preset KQI service threshold The KPI value corresponding to the predicted value or the KPI value smaller than the KPI deterioration inflection point;
  • Step S220 Mark and group the first predicted KPI values according to the preset business label, and determine the second predicted KPI value belonging to the root cause indicator group of concern;
  • Step S230 Determine the root cause KPI value according to the second predicted KPI value.
  • the KPI degradation inflection point is the KPI critical value for measuring user perception.
  • a KPI value smaller than the KPI degradation inflection point indicates that the KPI value is worse than the KPI critical value. Unacceptable to the user, the user is dissatisfied with the current network quality.
  • the KQI prediction value can reflect the impact of the KPI value on user perception. If the KQI prediction value is greater than the preset KQI service threshold, it can reflect that the KPI value corresponding to the KQI prediction value causes abnormal user perception.
  • the predicted KPI set that causes user perception abnormality can be determined, wherein the predicted KPI set includes multiple first KPI predicted values , the first KPI predicted value is the KPI value corresponding to the KQI predicted value greater than the KQI service threshold or the KPI value smaller than the KPI degradation inflection point.
  • the root cause indicators can be divided into basic KPIs such as capacity, coverage, and interference, performance KPIs such as access, delay, and rate, and quality KPIs such as retransmission and quality.
  • the remaining indicators are used as auxiliary root cause indicators.
  • root cause indicators can be improved during the wireless optimization process.
  • the root cause can be focused on the wireless optimization.
  • the expert analysis process is simplified, and combined with reference indicators, targeted wireless index optimization strategies can be formulated.
  • the first KPI predicted value is greater than the KPI value corresponding to the KQI predicted value of the preset KQI service threshold; in step S230, according to the second predicted KPI value Determine the root cause KPI value, including but not limited to step S310 and step S320:
  • Step S310 Obtain the abnormal influence degree of each second predicted KPI value
  • Step S320 Determine the root cause KPI value according to the influence degree of the abnormality.
  • the KPI prediction value is greater than the KPI value corresponding to the KQI prediction value of the KQI service threshold, it means that the KPI value corresponding to the KQI prediction value is an indicator that causes abnormal user perception.
  • the first KPI predicted value is marked and grouped to determine the second predicted KPI value.
  • the second predicted KPI value is the root cause index of concern, and the abnormal influence degree of each second predicted KPI value on user perception is further calculated. According to the abnormal influence degree, it can be Identify the root cause KPI value that mainly causes the poor KQI quality of the cell, so as to realize the location of the wireless root cause that affects user perception.
  • step S310 the abnormal influence degree of each second predicted KPI value is obtained in step S310, including but not limited to step S410 and step S420:
  • Step S410 Calculate the deviation difference between the predicted KQI value and the KQI service threshold
  • Step S420 Calculate the abnormal influence degree of the second predicted KPI value according to the deviation value.
  • the KPI value corresponding to the KQI forecast value greater than the KQI business threshold is determined, that is, the first KPI forecast value, and the second forecast KPI belonging to the root cause index group of concern is further determined value, calculate the difference between the KQI prediction value corresponding to each second predicted KPI value and the KQI business threshold, and obtain the deviation difference, obtain the deviation sum by accumulating all the deviation differences, and calculate the deviation difference corresponding to the second KPI prediction value Divide by the sum of the deviations to obtain the abnormal influence degree of the second predicted KPI value, and the network optimization personnel can optimize the indicators with high influence degree according to the ranking of the abnormal influence degree.
  • the multiple KQI predicted values are sorted from large to small.
  • the mapping function model Based on the mapping function model, input the KPI value of the target cell dimension to obtain multiple corresponding KQI prediction values, and sort the multiple KQI prediction values from large to small, combined with the unilateral characteristics of the KQI business threshold and index distribution, it can be quickly determined The KQI predicted value greater than the KQI service threshold, so as to determine the first KPI predicted value.
  • the KPI value corresponding to the KQI predicted value greater than the KQI service threshold is recorded as the first KPI predicted value, and the second KPI predicted value is further filtered out, And calculate the abnormal influence degree of each second KPI predicted value, obtain the abnormal influence degree ranking of the second KPI predicted value, so as to quickly locate the root cause KPI value.
  • the mapping function model includes multiple mapping functions; in step S140, according to the mapping function model, multiple distribution-based KPI degradation inflection points are obtained, including the following steps:
  • the preset KQI service thresholds are respectively input into multiple mapping functions, and multiple distribution-based KPI degradation inflection points are obtained through function inverse solution.
  • the preset mapping function model is obtained in step S110, including but not limited to steps S510 to S550:
  • Step S510 Obtain wireless side data and core network side data, wherein the wireless side data includes basic KPI data, and the core network side data includes basic KQI data;
  • the data on the wireless side adopts the soft acquisition method, and the uu port signaling is reported to the data analysis system by the data acquisition system.
  • the data on the wireless side includes time information, location information, self-busy time, and basic KPI data, core network side data
  • the user plane information is obtained mainly through hard probes
  • the core network side data includes time information, behavior information, location information, and basic KQI data.
  • Step S520 Perform data cleaning and data association on the wireless side data and the core network side data to obtain the full amount of data
  • Step S530 Calculate the correlation coefficient between the basic KPI data and the basic KQI data according to the full amount of data
  • Step S540 Determine the target KPI set that affects the target KQI value according to the correlation coefficient, wherein the target KPI set includes multiple target KPI values;
  • Step S550 According to the target KQI value and the target KPI value, construct a mapping function model through a logistic regression algorithm, wherein the mapping function model includes a plurality of functions for establishing a mapping relationship between KPI values and KQI values.
  • the correlation coefficient represents the fitting degree of the KQI value and the KPI value
  • the target KPI value that affects the target KQI value can be determined according to the calculated correlation coefficient.
  • the embodiment of the present application uses the Spearman correlation coefficient, but it is not limited thereto, and other correlation algorithms may also be used.
  • mapping function model in the embodiment of the present application is calculated by a logistic regression algorithm, and can also be solved by function approximation methods such as least squares method and polynomial fitting.
  • step S540 according to the correlation coefficient, determine the target KPI set that affects the target KQI value, including the following steps:
  • the correlation coefficient between the basic KPI data and the basic KQI data is calculated according to the full amount of data.
  • the target KPI set that affects the target KQI value is determined, and the mapping function model is constructed by a logistic regression algorithm. . As shown in Figure 6, taking the target KQI value as the average delay of TCP connection confirmation as an example, the target KPI value with weak correlation is obtained.
  • Example 1 Locate the root cause of the wireless side network in a community in Jiangjia Village, Qufu, Shandong province.
  • network indicators include wireless side data and core network side data
  • wireless side data includes time information, location information, self-busy time, and basic KPI data
  • core network side data includes time information, behavior information, and location information , and basic KQI data.
  • the wireless-side data and core network-side data cleaned in step 2 were aggregated at the cell dimension, aggregated and summarized at the hourly granularity, and the data samples from the whole province of Shandong were sampled during self-busy hours to obtain the full amount of data.
  • the Spearman correlation coefficient of the basic KPI data and basic KQI data is calculated for the full amount of data after associated sampling in step 3. For the index pair with a correlation coefficient greater than 0.2, it is determined that there is a weak correlation, which is clear for business scenarios The target KQI value of , and obtain the weakly correlated target KPI value respectively.
  • the target KPI values include: wireless connection rate, wireless disconnection rate, RRC connection re-establishment rate, MAC layer uplink residual block error rate, MAC layer downlink residual block error rate, Uplink HARQ retransmission ratio, downlink HARQ retransmission ratio, CQI excellent rate, downlink single-stream ratio, uplink QPSK coding ratio, downlink QPSK coding ratio, average cell downlink RLC SDU delay, cell uplink RLC SDU average delay, RRC connection
  • the maximum number of connected users, downlink shared channel PRB utilization rate, PDCCH channel CCE occupancy rate, cell uplink UE Throughput, cell downlink UE Throughput, and downlink DTX ratio respectively take KPI as input x, and take the average TCP connection confirmation delay as output y, Solve the x-y mapping function through the logistic regression algorithm, so as to obtain multiple mapping functions, refer to FIG. 6 .
  • the KQI service threshold is set to 50ms, and 50ms is brought into the mapping function, and multiple KPI degradation inflection points are obtained by reverse calculation.
  • the statistical data of the wireless side indicators of the cell (the KPI value of the target cell) is input into the mapping function, and the KQI prediction value is obtained.
  • the community GLZ*****46R1_Dashiziyuan Huiqiao Town, Changqing District, TCP two-three handshake delay (average TCP connection confirmation delay) is abnormal, take each KPI value in Figure 7 as x, and enter the value diagram respectively In the corresponding mapping function in 6, multiple KQI prediction values are obtained, and the KQI prediction values are sorted from large to small.
  • the KQI prediction value sorting result is KPI value vs. TCP two-three handshake delay index order of influence. It should be noted that the figure only shows part of the data, and the actual data is not limited thereto.
  • the KPI value that satisfies the condition is recorded as the first KPI predicted value and Form the predicted KPI set Y1, wherein the first KPI predicted value is the KPI value that causes the user to perceive abnormality, and mark the first KPI predicted value corresponding to the predicted KPI set Y1 according to the preset business label, refer to Figure 9, and according to the mark Carry out grouping, use capacity, coverage, and interference indicators as priority root cause indicators, and other indicators as auxiliary root cause indicators.
  • the difference between the KQI predicted value corresponding to the predicted KPI value and the KQI business threshold is obtained to obtain the deviation difference DIFF, and the deviation difference DIFF corresponding to the second KPI predicted value is divided by the deviation sum (all deviation differences are accumulated) to obtain the second Predict the abnormal influence degree of KPI value.
  • the root cause KPI value of the abnormal TCP two-three handshake delay is: the path loss distribution of uplink service mode users between partitions Proportion, influence degree 51.64%; TA greater than 27 proportion, influence degree 48.35%, auxiliary root cause indicators are: uplink HARQ retransmission ratio, uplink bler.
  • Example 2 The effective rate of video download in a community of Longsang Temple is low
  • the root cause positioning results obtained based on the mapping function model are: one type of root cause-coverage type, two types of root cause-the path loss distribution of uplink service state users between partitions ( >135) accounted for 60.36% of the impact; TA greater than 27 accounted for 39.64% of the contribution; it can be seen that the main cause of the poor quality problem is the coverage problem.
  • the effective rate of video downloading has increased from 10.89Mbps to 23.48Mbps, exceeding the perception baseline by 15Mbps, realizing a closed-loop perception; the average cache time of video playback has improved by 262ms; web pages and instant messaging services have also improved to varying degrees.
  • the evaluation of the optimization effect of the quality KPI layer shows that the improvement of the KPIs of the overall quality layer is relatively obvious, such as: the proportion of uplink QPSK is improved by 36%; the uplink HARQ retransmission rate is 16%.
  • the uplink/downlink rate index has improved significantly, the uplink rate has improved by 1.8Mbps, the downlink rate has improved by 30.9Mbps; the connection rate index has improved by 3.94%;
  • the average delay of TCP connection confirmation (ms) and the average RTT delay of TCP downlink have improved significantly, and the improvement range is about 50ms.
  • This application is guided by user perception, and can provide a clear wireless optimization index for user perception.
  • the mapping function model between KPI value and KQI value is abstracted, based on the quantitative mapping function model, based on the preset KQI
  • the service threshold is used to obtain the inflection point of KPI degradation based on the distribution.
  • the KQI prediction value is calculated to obtain the proportion of abnormal influence of the corresponding KPI value, and priority is given to solving indicators with high proportion of influence.
  • the root cause can be focused on the operable range of wireless optimization, which facilitates the formulation of targeted wireless index optimization strategies, and is conducive to improving the user's actual experience perception .
  • the embodiment of the second aspect of the present application provides an operation control device 1200, including at least one control processor 1210 and a memory 1220 for communicating with at least one control processor 1210; the control processor 1210 and The memory 1220 may be connected via a bus or in other ways.
  • An example of a connection via a bus is shown in FIG. 12 .
  • the memory 1220 stores instructions executable by at least one control processor 1210.
  • a control processor 1210 can execute the method for locating the root cause of the wireless side network in the embodiment of the first aspect above, for example, execute the above-described method steps S110 to S150 in FIG. 1 , method steps S210 to S230 in FIG. 2 , and FIG. 3 Method steps S310 and S320 in FIG.
  • mapping function model between the abstracted KPI value and KQI value
  • multiple KQI prediction values can be obtained according to the KPI value of the target cell, and the KQI prediction value can reflect the KPI value on user perception. Influence, it is convenient to quickly locate KPI values with high influence.
  • multiple KPI degradation inflection points based on distribution can be obtained.
  • a targeted wireless index optimization strategy can be formulated, which is conducive to improving the user's actual experience perception.
  • the embodiment of the third aspect of the present application provides a computer-readable storage medium, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions can be used to make the computer execute the wireless-side network root Due to the positioning method, for example, the method steps S110 to S150 in FIG. 1 described above, the method steps S210 to S230 in FIG. 2 , the method steps S310 and S320 in FIG. 3 , the method steps S410 and S420 in FIG. 4 , And the method steps S510 to S550 in FIG. 5 .
  • mapping function model By obtaining the mapping function model between the abstracted KPI value and KQI value, on the basis of the mapping function model, multiple KQI prediction values can be obtained according to the KPI value of the target cell, and the KQI prediction value can reflect the KPI value on user perception. Influence, it is convenient to quickly locate KPI values with high influence.
  • multiple KPI degradation inflection points based on distribution can be obtained.
  • a targeted wireless index optimization strategy can be formulated, which is conducive to improving the user's actual experience perception.
  • the embodiment of the present application includes: acquiring a preset mapping function model, wherein the mapping function model is used to establish a mapping relationship between a key performance indicator KPI value and a key quality indicator KQI value; acquiring KPI data of a target cell, the The KPI data includes a plurality of KPI values; according to the mapping function model, a plurality of KQI prediction values corresponding to the KPI values are calculated, wherein the KQI prediction values are used to reflect the influence of the corresponding KPI value on user perception; According to the mapping function model, a plurality of distribution-based KPI deterioration inflection points are obtained; and a root cause KPI value is determined according to the KQI prediction value and the KPI deterioration inflection point.
  • multiple KQI prediction values can be obtained according to the KPI value of the target cell, through KQI prediction
  • the value can reflect the impact of KPI values on user perception, which is convenient for quickly locating KPI values with high influence.
  • multiple distribution-based KPI deterioration inflection points can be obtained, and the root Based on the KPI value, using the root cause KPI value as an optimization index can formulate a targeted wireless index optimization strategy, which is conducive to improving the user's actual experience perception.
  • Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk DVD or other optical disk storage, magnetic cartridges, tape, magnetic disk storage or other magnetic storage devices, or can be used in Any other medium that stores desired information and that can be accessed by a computer.
  • communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media .

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本申请公开了一种无线侧网络根因定位方法、运行控制装置及存储介质，该方法包括：获取预设的映射函数模型，其中，映射函数模型用于建立关键性能指标KPI值和关键质量指标KQI值之间的映射关系(S110)；获取目标小区的KPI数据，KPI数据包括多个KPI值(S120)；根据映射函数模型，计算得到多个与KPI值对应的KQI预测值，其中，KQI预测值用于反映对应的KPI值对用户感知的影响(S130)；根据映射函数模型，得到多个基于分布的KPI劣化拐点(S140)；根据KQI预测值和KPI劣化拐点确定根因KPI值(S150)。

Description

无线侧网络根因定位方法、运行控制装置及存储介质
相关申请的交叉引用
本申请基于申请号为202111571366.0、申请日为2021年12月21日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此引入本申请作为参考。
技术领域
本申请涉及网络运维技术领域，尤其涉及一种无线侧网络根因定位方法、运行控制装置及存储介质。
背景技术
移动运营商日常运维网络包括无线侧小区，4G、5G核心网网元，通过分析移动通讯系统中控制面及用户面信令，对网络中存在的问题进行分析和评估，通过扩容、参数优化、工程实施等技术手段，可以提高移动终端用户的网络使用体验。目前在移动运营商无线运维领域，无线侧指标有上百个，针对无线侧关键性能指标(Key Performance Indication，KPI)专项优化的工具有具体的体系方法。当前系统优化的目标指向无线侧KPI提升，而无线侧KPI与用户的实际感知之间的关系是不明确的，当出现用户投诉或网络专项优化任务时，网络优化人员通常依据已有系统进行小区无线侧指标分析，并根据专家经验进行参数的调整，但优化操作能否提升用户的使用感知取决于专家经验，通常需要几轮优化迭代尝试，难以快速准确定位到异常根因，无法有效针对用户特定体验感知进行优化操作。
发明内容
本申请实施例提出一种无线侧网络根因定位方法、运行控制装置及存储介质。
第一方面，本申请实施例提供一种无线侧网络根因定位方法，包括：获取预设的映射函数模型，其中，所述映射函数模型用于建立关键性能指标KPI值和关键质量指标KQI值之间的映射关系；获取目标小区的KPI数据，所述KPI数据包括多个KPI值；根据所述映射函数模型，计算得到多个与所述KPI值对应的KQI预测值，其中，所述KQI预测值用于反映对应的KPI值对用户感知的影响；根据所述映射函数模型，得到多个基于分布的KPI劣化拐点；根据所述KQI预测值和所述KPI劣化拐点确定根因KPI值。
第二方面，本申请实施例提供一种运行控制装置，包括至少一个控制处理器和用于与所述至少一个控制处理器通信连接的存储器；所述存储器存储有可被所述至少一个控制处理器执行的指令，所述指令被所述至少一个控制处理器执行，以使所述至少一个控制处理器能够执行如上第一方面实施例所述的无线侧网络根因定位方法。
第三方面，本申请实施例提供一种计算机可读存储介质，所述计算机可读存储介质存储有计算机可执行指令，所述计算机可执行指令用于使计算机执行如上第一方面实施例所述的无线侧网络根因定位方法。
本申请的其它特征和优点将在随后的说明书中阐述，并且，部分地从说明书中变得显而易见，或者通过实施本申请而了解。本申请的目的和其他优点可通过在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
附图说明
附图用来提供对本申请技术方案的进一步理解，并且构成说明书的一部分，与本申请的实施例一起用于解释本申请的技术方案，并不构成对本申请技术方案的限制。
下面结合附图和实施例对本申请进一步地说明；
图1是本申请的一个实施例提供的无线侧网络根因定位方法的流程图；
图2是本申请的另一个实施例提供的无线侧网络根因定位方法的流程图；
图3是本申请的另一个实施例提供的无线侧网络根因定位方法的流程图；
图4是本申请的另一个实施例提供的无线侧网络根因定位方法的流程图；
图5是本申请的另一个实施例提供的无线侧网络根因定位方法的流程图；
图6是本申请的另一个实施例提供的映射函数的示意图；
图7是本申请的另一个实施例提供的目标小区的KPI值的示意图；
图8是本申请的另一个实施例提供的KQI预测值的排序结果示意图；
图9是本申请的另一个实施例提供的KPI值按照预设业务标签进行标记分组的示意图；
图10是本申请的另一个实施例提供的龙桑寺某小区TCP指标的示意图；
图11是本申请的另一个实施例提供的龙桑寺小区根因定位结果的示意图；
图12是本申请的另一个实施例提供的运行控制装置的结构示意图。
具体实施方式
本部分将详细描述本申请的实施例，本申请之若干实施例在附图中示出，附图的作用在于用图形补充说明书文字部分的描述，使人能够直观地、形象地理解本申请的每个技术特征和整体技术方案，但其不能理解为对本申请保护范围的限制。
在本申请的描述中，如果有描述到第一、第二只是用于区分技术特征为目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量或者隐含指明所指示的技术特征的先后关系。需要说明的是，虽然在装置示意图中进行了功能模块划分，在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于装置中的模块划分，或流程图中的顺序执行所示出或描述的步骤。
本申请的描述中，除非另有明确的限定，设置、安装、连接等词语应做广义理解，所属技术领域技术人员可以结合技术方案的内容合理确定上述词语在本申请中的含义。
下面结合附图，对本申请实施例作进一步阐述。
目前在移动运营商无线运维领域，无线侧指标有上百个，针对无线侧KPI专项优化的工具有具体的体系方法。当前系统优化的目标指向无线侧KPI提升，而无线侧KPI与用户的实际感知之间的关系是不明确的，当出现用户投诉或网络专项优化任务时，网络优化人员通常依据已有系统进行小区无线侧指标分析，并根据专家经验进行参数的调整，但优化操作能否提升用户的使用感知取决于专家经验，通常需要几轮优化迭代尝试，难以快速准确定位到异常根因，无法有效针对用户特定体验感知进行优化操作。
本申请实施例提供一种无线侧网络根因定位方法、运行控制装置及存储介质，能够快速定位根因，有利于提升用户的实际体验感知。
如图1所示，本申请的第一方面实施例提供一种无线侧网络根因定位方法，包括但不限于步骤S110至步骤S150：
步骤S110：获取预设的映射函数模型，其中，映射函数模型用于建立KPI值和KQI值之间的映射关系；
需要说明的是，当前网络系统优化的目标主要指向无线KPI提升，而无线KPI与用户的实际感知中间的映射关系是不明确的。关键质量指标(Key Quality Indicators，KQI)是针对不同业务提出的贴近用户感受的业务质量参数，能够反映小区无线信号质量，通过结合网优专家的经验，获取预设的映射函数模型，能够得到KPI值和KQI值之间的映射关系，从而能够根据KQI值确定关联的KPI值，通过将根因定位流程抽象为数学建模，为提升用户感知的网络优化提供有效的技术支撑工具。
步骤S120：获取目标小区的KPI数据，KPI数据包括多个KPI值；
当目标小区反映存在质差问题，对目标小区进行根因识别，首先获取目标小区的KPI数据，其中，KPI数据包括多个KPI值。
需要说明的是，KPI数据可以包括小区上行RLC SDU平均时延(ms)、PDCCH信道CCE占用率、上行bler、上行HARQ重传比率、上行CCE分配失败率、MAC层上行残留误块率、上行使用256QAM表的平均MCS等指标值，KQI数据可以包括视频有效下载速率(kbps)、视频质量、视频MOS、视频TCP下行RTT平均时延、视频TCP连接确认平均时延、TCP连接成功率、网页TCP上行重传率、网页TCP下行重传率等指标值。
步骤S130：根据映射函数模型，计算得到多个与KPI值对应的KQI预测值，其中，KQI预测值用于反映对应的KPI值对用户感知的影响；
需要说明的是，映射函数模型包括多个映射函数，将KPI值输入映射函数中，能够计算得到对应的KQI预测值,KQI预测值可以反映KPI值对用户感知的影响，便于快速定位影响度高的KPI值。对于N个KPI值X＝[x kpi1,x kpi2,...x kpin]获得N个KQI预测值Y＝[y kqi1,y kqi2,...y kqin]。
步骤S140：根据映射函数模型，得到多个基于分布的KPI劣化拐点；
根据映射函数模型，能够得到多个KPI劣化拐点，其中，KPI劣化拐点为衡量用户感知的KPI临界值，能够反映用户对网络质量的接受程度。经验证，KPI劣化拐点的结合了不同地市或省份的无线基站覆盖情况、用户行为差异，在无线网络的优化工作中具有更贴合实际的指导意义。
步骤S150：根据KQI预测值和KPI劣化拐点确定根因KPI值。
结合KQI预测值和KPI劣化拐点可以快速定位根因KPI值，有效地反映用户感知，通过将根因KPI值作为优化指标，可以制定针对性地制定无线指标优化策略，有利于提升用户的实际体验感知。
在本实施例中，通过获取抽象出KPI值和KQI值之间的映射函数模型，在映射函数模型的基础上，根据目标小区的KPI值可以获得多个KQI预测值，通过KQI预测值可以反映KPI值对用户感知的影响，便于快速定位影响度高的KPI值，另外，基于映射函数模型，得到多个基于分布的KPI劣化拐点，根据KQI预测值和KPI劣化拐点可以快速定位根因KPI值，将根因KPI值作为优化指标，可以制定针对性的无线指标优化策略，有利于提升用户的实际体 验感知。
如图2所示，在上述的无线侧网络根因定位方法中，步骤S150中根据KQI预测值和KPI劣化拐点确定根因KPI值，包括但不限于步骤S210至步骤S230：
步骤S210：根据KQI预测值和KPI劣化拐点得到预测KPI集合，其中，预测KPI集合包括多个引发用户感知异常的第一KPI预测值，第一KPI预测值为大于预设的KQI业务门限的KQI预测值所对应的KPI值或者为小于KPI劣化拐点的KPI值；
步骤S220：根据预设业务标签对第一KPI预测值进行标记分组，确定属于关注根因指标组别的第二预测KPI值；
步骤S230：根据第二预测KPI值确定根因KPI值。
KPI劣化拐点为衡量用户感知的KPI临界值，KPI值小于KPI劣化拐点表示KPI值劣于KPI临界值，可以理解的是，若KPI值小于KPI劣化拐点，则反映此KPI值下的业务感知是用户不能接受的，用户对当前的网络质量不满意。另外，KQI预测值可以反映KPI值对用户感知的影响，若KQI预测值大于预设的KQI业务门限，可以反映此KQI预测值所对应的KPI值引发用户感知异常。
需要说明的是，通过判断KQI预测值是否大于KQI业务门限，以及判断KPI值是否小于KPI劣化拐点，可以确定引发用户感知异常的预测KPI集合，其中，预测KPI集合包括多个第一KPI预测值，第一KPI预测值为大于KQI业务门限的KQI预测值所对应的KPI值或者为小于KPI劣化拐点的KPI值。
根据预设业务标签对第一KPI预测值做定性标签，并将第一KPI预测值进行分组，确定属于关注根因指标组别的第二预测KPI值。根据预设业务标签可以将根因指标分为容量、覆盖、干扰等基础KPI，接入、时延、速率等性能KPI，重传、质量等质量KPI等类别，将容量、覆盖、干扰类指标作为优先关注根因指标，其余指标作为辅助根因指标，通过对第一KPI预测值做定性标签，确定属于关注根因指标组别的第二预测KPI值，从第二预测KPI值中能够定位到具体的根因KPI值。
需要说明的是，关注根因指标表示此类指标在无线优化过程中可得到改善，通过对引发用户感知异常的第一KPI预测值进行分层分域标签，可以将根因聚焦在无线优化的可操作范围内，简化了专家分析过程，同时结合参考指标，能够制定针对性的无线指标优化策略。
如图3所示，在上述的无线侧网络根因定位方法中，第一KPI预测值为大于预设的KQI业务门限的KQI预测值所对应的KPI值；步骤S230中根据第二预测KPI值确定根因KPI值，包括但不限于步骤S310和步骤S320：
步骤S310：获取每个第二预测KPI值的异常影响度；
步骤S320：根据异常影响度确定根因KPI值。
需要说明的是，若第一KPI预测值为大于KQI业务门限的KQI预测值所对应的KPI值，则表示KQI预测值所对应的KPI值为引发用户感知异常的指标，根据预设业务标签对第一KPI预测值进行标记分组，确定第二预测KPI值，第二预测KPI值均为关注根因指标，进一步计算每个第二预测KPI值对用户感知的异常影响度，根据异常影响度可以识别主要引发小区KQI质差的根因KPI值，从而能够实现影响用户感知的无线根因定位。
如图4所示，在上述的无线侧网络根因定位方法中，步骤S310中获取每个第二预测KPI值的异常影响度，包括但不限于步骤S410和步骤S420：
步骤S410：计算KQI预测值和KQI业务门限的偏离差值；
步骤S420：根据偏离差值计算第二预测KPI值的异常影响度。
需要说明的是，通过计算出多个KQI预测值，确定大于KQI业务门限的KQI预测值所对应的KPI值，即第一KPI预测值，进一步确定属于关注根因指标组别的第二预测KPI值，计算每个第二预测KPI值对应的KQI预测值和KQI业务门限的差值，得到偏离差值，通过将所有偏离差值累加得到偏离总和，将第二KPI预测值对应的偏离差值除以偏离总和即得到第二预测KPI值的异常影响度，网络优化人员根据异常影响度的排序能够对影响度高的指标进行优化处理。
在上述的无线侧网络根因定位方法中，在步骤S210中根据KQI预测值和KPI劣化拐点得到预测KPI集合之前，将多个KQI预测值按照从大到小排序。
基于映射函数模型，输入目标小区维度的KPI值，得到多个对应的KQI预测值，将多个KQI预测值按照从大到小排序，结合KQI业务门限及指标分布的单边特性，能够快速确定大于KQI业务门限的KQI预测值，从而确定第一KPI预测值。
在一实施例中，通过将多个KQI预测值按照从大到小排序，将大于KQI业务门限的KQI预测值对应的KPI值记为第一KPI预测值，进一步筛选出第二KPI预测值，并计算每个第二KPI预测值的异常影响度，获取第二KPI预测值的异常影响度排名，从而快速定位根因KPI值。
在上述的无线侧网络根因定位方法中，映射函数模型包括多个映射函数；步骤S140中根据映射函数模型，得到多个基于分布的KPI劣化拐点，包括以下步骤：
将预设的KQI业务门限分别输入至多个映射函数中，通过函数反解得到多个基于分布的KPI劣化拐点。
需要说明的是，基于映射函数模型y KQI＝f(x KPI),结合预设的KQI业务门限y’ KQI，通过函数反解，可以得到多个基于分布的KPI劣化拐点x’ KPI＝f -1(y’ KQI)。基于映射函数{KQI 1＝f1(KPI 1),KQI 1＝f2(KPI 2),...,KQI 1＝fn(KPI n)}，输入相应的KQI业务门限，通过函数反解，可以得到多个KPI劣化拐点，通过获取基于用户感知的KPI劣化拐点，能够有效对目标小区的KPI值进行分析。
如图5所示，在上述的无线侧网络根因定位方法中，步骤S110中获取预设的映射函数模型，包括但不限于步骤S510至步骤S550：
步骤S510：获取无线侧数据和核心网侧数据，其中，无线侧数据包括基础KPI数据，核心网侧数据包括基础KQI数据；
需要说明的是，无线侧数据采用软采方式，uu口信令由数据采集系统上报至数据分析系统，无线侧数据包括时间信息、位置信息、自忙时，以及基础KPI数据，核心网侧数据主要通过硬探针获取用户面信息，核心网侧数据包含时间信息、行为信息、位置信息，以及基础KQI数据。
步骤S520：对无线侧数据和核心网侧数据进行数据清洗和数据关联，得到全量数据；
需要说明的是，获取基础数据后，对无线侧数据和核心网侧数据中的有效数据进行数据清洗，包括规范化、枚举化，并进行小区维度的指标聚集，进一步进行数据关联和采样，得到全量数据，便于后续数据分析。
步骤S530：根据全量数据，计算基础KPI数据和基础KQI数据的相关性系数；
步骤S540：根据相关性系数，确定影响目标KQI值的目标KPI集合，其中，目标KPI集合包括多个目标KPI值；
步骤S550：根据目标KQI值和目标KPI值，通过逻辑回归算法构建映射函数模型，其中，映射函数模型包括多个用于建立KPI值和KQI值之间的映射关系的函数。
基于关联后的全量数据，计算基础KPI数据和基础KQI数据的相关性系数，根据相关性系数的计算结果，确定影响用户感知的目标KQI值的目标KPI集合，即确定对目标KQI值有波及影响的目标KPI值，后续分别以目标KPI值为输入x，以目标KQI值为输出y，通过逻辑回归算法，求解x-y映射函数，从而构建出映射函数模型。
需要说明的是，相关性系数表征KQI值和KPI值的拟合程度，根据计算出的相关性系数，可以确定影响目标KQI值的目标KPI值。本申请实施例采用的是斯皮尔曼(spearman)相关性系数，但不限于此，还可以采用其他相关性算法。
此外，本申请实施例的映射函数模型通过逻辑回归算法计算得到，还可以使用最小二乘法、多项式拟合等函数逼近方式进行求解。
在上述的无线侧网络根因定位方法中，步骤S540中根据相关性系数，确定影响目标KQI值的目标KPI集合，包括以下步骤：
当相关性系数大于预设系数，确定影响目标KQI值的目标KPI集合。
在一实施例中，根据全量数据计算基础KPI数据和基础KQI数据的相关性系数，当相关性系数大于预设系数，则确定影响目标KQI值的目标KPI集合，通过逻辑回归算法构建映射函数模型。如图6所示，以目标KQI值是TCP连接确认平均时延为例，获取存在弱相关的目标KPI值，对于相关性系数大于0.2的指标对，判定存在弱相关，确定出影响TCP连接确认平均时延的目标KPI集合{KPI 1，KPI 2，...,KPI i}，分别以KPI i为输入x，以TCP连接确认平均时延为输入y，求解得到多个x-y映射函数。
为了更清楚阐述本申请的无线侧网络根因定位方法，以下将用两个实施例作进一步介绍。
实施例一：山东省曲阜姜家村某小区无线侧网络根因定位。
1、网络指标采集：网络指标包括无线侧数据和核心网侧数据，无线侧数据包括时间信息、位置信息、自忙时，以及基础KPI数据，核心网侧数据包含时间信息、行为信息、位置信息，以及基础KQI数据。
2、原始数据清洗
对步骤1的原始数据中的有效信息进行清洗，除去异常数据。
3、数据关联及采样
将步骤2清洗后的无线侧数据和核心网侧数据，分别进行小区维度的指标聚集，做小时粒度的聚集汇总统计，针对山东全省数据样本，进行自忙时数据采样，得到全量数据。
4、构建映射函数模型
如图6所示，对步骤3关联采样后的全量数据计算基础KPI数据和基础KQI数据的斯皮尔曼相关性系数，对于相关性系数大于0.2的指标对，判定存在弱相关，对于业务场景明确的目标KQI值，分别获取弱相关目标KPI值。以目标KQI值是TCP连接确认平均时延为例，目标KPI值包括：无线接通率、无线掉线率、RRC连接重建比率、MAC层上行残留误块率、MAC层下行残留误块率、上行HARQ重传比率、下行HARQ重传比率、CQI优良率、下行单流占比、上行QPSK编码比例、下行QPSK编码比例、小区下行RLC SDU平均时延、小区上行RLC SDU 平均时延、RRC连接最大连接用户数、下行共享信道PRB利用率、PDCCH信道CCE占用率、小区上行UE Throughput、小区下行UE Throughput、下行DTX比例，分别以KPI作为输入x，以TCP连接确认平均时延作为输出y，通过逻辑回归算法，求解x-y映射函数，从而得到多个映射函数，参照图6。
5、计算KPI劣化拐点
以KQI值为TCP二三次握手时延为例，该KQI业务门限设定为50ms，将50ms带入映射函数中，反向求解得到多个KPI劣化拐点。
6、计算KPI值的异常影响度
如图6至图8所示，针对特定小区业务感知异常情况，将小区的无线侧指标统计数据(目标小区的KPI值)输入到映射函数，并得到KQI预测值。如小区GLZ*****46R1_长清区大柿子园汇侨城TCP二三次握手时延(TCP连接确认平均时延)异常，将图7中各个KPI值作为x，分别输入值图6中对应的映射函数中，得到多个KQI预测值，并将KQI预测值按照从大到小排序，参照图8，获得KQI预测值排序结果则为KPI值对TCP二三次握手时延指标的影响度排序。需要说明的是，图中仅示出部分数据，实际数据并不限于此。
7、无线根因定位
如图9所示，当目标小区的KPI值满足小于KPI劣化拐点或者KPI值对应的KQI预测值大于预设的KQI业务门限(50ms)，将满足条件的KPI值记为第一KPI预测值并形成预测KPI集合Y1，其中，第一KPI预测值为引发用户感知异常的KPI值，针对预测KPI集合Y1对应的第一KPI预测值，按照预设业务标签进行标记，参照图9，并根据标记进行分组，将容量、覆盖、干扰类指标作为优先关注根因指标，其余指标作为辅助根因指标，将属于关注根因指标组别的KPI值记为第二预测KPI值，通过进一步计算第二预测KPI值对应的KQI预测值和KQI业务门限的差值，得到偏离差值DIFF，并将第二KPI预测值对应的偏离差值DIFF除以偏离总和(所有偏离差值累加得到)得到第二预测KPI值的异常影响度。
从而可以分析到样例小区GLZ*****46R1_长清区大柿子园汇侨城，TCP二三次握手时延异常的根因KPI值为：分区间上行业务态用户的路损分布占比，影响度51.64％；TA大于27的占比，影响度48.35％，辅助根因指标为：上行HARQ重传比率、上行bler。
实施例二：龙桑寺某小区的视频下载有效速率低
如图10和图11所示，获取山东省龙桑寺某小区数据，该小区TCP指标情况参照图10。可以看到TCP连接确认平均时延及TCP下行平均RTT时延偏高，定界视频下载速率低，主要是无线侧网络的问题，非核心网/SP问题，对该小区进行根因定位，根因定位步骤参照上述实施例一中步骤1到7，基于映射函数模型得到的根因定位结果为：一类根因-覆盖类，二类根因-分区间上行业务态用户的路损分布(>135)占比，影响占比60.36％；TA大于27的占比贡献度39.64％；可见，质差问题主因是覆盖问题。
通过对覆盖问题优化改善后，各层的KPI/KQI均发生改善：
视频下载有效速率由10.89Mbps提升至23.48Mbps，超过感知基线15Mbps，实现了感知闭环；视频播放平均缓存时长改善262ms；页面和即时通讯业务也有不同程度的改善。
质量KPI层优化效果评估，整体各项质量层KPI改善较为明显，如：上行QPSK占比改善36％；上行HARQ重传比率16％。
性能KPI层优化效果评估，上/下行速率指标改善明显，上行速率改善1.8Mbps，下行速 率改善30.9Mbps；接通率指标改善3.94％；
基础TCP层优化效果评估，TCP连接确认平均时延(ms)和TCP下行平均RTT时延改善较为明显，改善幅度50ms左右。
本申请以用户感知为导向，可以针对用户感知给出明确的无线优化指标。将网优专家经验与统计学及机器学习算法相结合，通过模型构建及逻辑回归，抽象出KPI值与KQI值之间的映射函数模型，在量化的映射函数模型基础上，基于预设的KQI业务门限，得到基于分布的KPI劣化拐点，另外，基于映射函数模型，计算出KQI预测值，获得对应的KPI值异常影响占比，优先解决影响度占比高的指标对于网络优化具体工作的实施有重要的指导意义，通过对预设业务标签对KPI值进行标记，可以将根因聚焦在无线优化的可操作范围内，便于制定针对性的无线指标优化策略，有利于提升用户的实际体验感知。
如图12所示，本申请的第二方面实施例提供一种运行控制装置1200，包括至少一个控制处理器1210和用于与至少一个控制处理器1210通信连接的存储器1220；控制处理器1210和存储器1220可以通过总线或者其他方式连接，图12中示出通过总线连接的例子，存储器1220存储有可被至少一个控制处理器1210执行的指令，指令被至少一个控制处理器1210执行，以使至少一个控制处理器1210能够执行如上第一方面实施例的无线侧网络根因定位方法，例如，执行以上描述的图1中的方法步骤S110至S150、图2中的方法步骤S210至S230、图3中的方法步骤S310和S320、图4中的方法步骤S410和S420、以及图5中的方法步骤S510至S550。通过获取抽象出KPI值和KQI值之间的映射函数模型，在映射函数模型的基础上，根据目标小区的KPI值可以获得多个KQI预测值，通过KQI预测值可以反映KPI值对用户感知的影响，便于快速定位影响度高的KPI值，另外，基于映射函数模型，得到多个基于分布的KPI劣化拐点，根据KQI预测值和KPI劣化拐点可以快速定位根因KPI值，将根因KPI值作为优化指标，可以制定针对性的无线指标优化策略，有利于提升用户的实际体验感知。
本申请的第三方面实施例提供一种计算机可读存储介质，计算机可读存储介质存储有计算机可执行指令，计算机可执行指令可以用于使计算机执行如上第一方面实施例的无线侧网络根因定位方法，例如，执行以上描述的图1中的方法步骤S110至S150、图2中的方法步骤S210至S230、图3中的方法步骤S310和S320、图4中的方法步骤S410和S420、以及图5中的方法步骤S510至S550。通过获取抽象出KPI值和KQI值之间的映射函数模型，在映射函数模型的基础上，根据目标小区的KPI值可以获得多个KQI预测值，通过KQI预测值可以反映KPI值对用户感知的影响，便于快速定位影响度高的KPI值，另外，基于映射函数模型，得到多个基于分布的KPI劣化拐点，根据KQI预测值和KPI劣化拐点可以快速定位根因KPI值，将根因KPI值作为优化指标，可以制定针对性的无线指标优化策略，有利于提升用户的实际体验感知。
本申请实施例包括：获取预设的映射函数模型，其中，所述映射函数模型用于建立关键性能指标KPI值和关键质量指标KQI值之间的映射关系；获取目标小区的KPI数据，所述KPI数据包括多个KPI值；根据所述映射函数模型，计算得到多个与所述KPI值对应的KQI预测值，其中，所述KQI预测值用于反映对应的KPI值对用户感知的影响；根据所述映射函数模型，得到多个基于分布的KPI劣化拐点；根据所述KQI预测值和所述KPI劣化拐点确定根因KPI值。根据本申请实施例提供的方案，通过获取抽象出KPI值和KQI值之间的映射函数模型，在映射函数模型的基础上，根据目标小区的KPI值可以获得多个KQI预测值，通过KQI 预测值可以反映KPI值对用户感知的影响，便于快速定位影响度高的KPI值，另外，基于映射函数模型，得到多个基于分布的KPI劣化拐点，根据KQI预测值和KPI劣化拐点可以快速定位根因KPI值，将根因KPI值作为优化指标，可以制定针对性的无线指标优化策略，有利于提升用户的实际体验感知。
本领域普通技术人员可以理解，上文中所公开方法中的全部或某些步骤、系统可以被实施为软件、固件、硬件及其适当的组合。某些物理组件或所有物理组件可以被实施为由处理器，如中央处理器、数字信号处理器或微处理器执行的软件，或者被实施为硬件，或者被实施为集成电路，如专用集成电路。这样的软件可以分布在计算机可读介质上，计算机可读介质可以包括计算机存储介质或非暂时性介质和通信介质或暂时性介质。如本领域普通技术人员公知的，术语计算机存储介质包括在用于存储信息诸如计算机可读指令、数据结构、程序模块或其他数据的任何方法或技术中实施的易失性和非易失性、可移除和不可移除介质。计算机存储介质包括但不限于RAM、ROM、EEPROM、闪存或其他存储器技术、CD-ROM、数字多功能盘DVD或其他光盘存储、磁盒、磁带、磁盘存储或其他磁存储装置、或者可以用于存储期望的信息并且可以被计算机访问的任何其他的介质。此外，本领域普通技术人员公知的是，通信介质通常包含计算机可读指令、数据结构、程序模块或者诸如载波或其他传输机制之类的调制数据信号中的其他数据，并且可包括任何信息递送介质。
上面结合附图对本申请实施例作了详细说明，但是本申请不限于上述实施例，在所述技术领域普通技术人员所具备的知识范围内，还可以在不脱离本申请宗旨的前提下作出各种变化。

Claims (10)

  1. 一种无线侧网络根因定位方法，包括：
    获取预设的映射函数模型，其中，所述映射函数模型用于建立关键性能指标KPI值和关键质量指标KQI值之间的映射关系；
    获取目标小区的KPI数据，所述KPI数据包括多个KPI值；
    根据所述映射函数模型，计算得到多个与所述KPI值对应的KQI预测值，其中，所述KQI预测值用于反映对应的KPI值对用户感知的影响；
    根据所述映射函数模型，得到多个基于分布的KPI劣化拐点；
    根据所述KQI预测值和所述KPI劣化拐点确定根因KPI值。
  2. 根据权利要求1所述的无线侧网络根因定位方法，其中，所述根据所述KQI预测值和所述KPI劣化拐点确定根因KPI值，包括：
    根据所述KQI预测值和所述KPI劣化拐点得到预测KPI集合，其中，所述预测KPI集合包括多个引发用户感知异常的第一KPI预测值，所述第一KPI预测值为大于预设的KQI业务门限的所述KQI预测值所对应的KPI值或者为小于所述KPI劣化拐点的KPI值；
    根据预设业务标签对所述第一KPI预测值进行标记分组，确定属于关注根因指标组别的第二预测KPI值；
    根据所述第二预测KPI值确定根因KPI值。
  3. 根据权利要求2所述的无线侧网络根因定位方法，其中，所述第一KPI预测值为大于预设的KQI业务门限的所述KQI预测值所对应的KPI值；
    所述根据所述第二预测KPI值确定根因KPI值，包括：
    获取每个所述第二预测KPI值的异常影响度；
    根据所述异常影响度确定根因KPI值。
  4. 根据权利要求3所述的无线侧网络根因定位方法，其中，所述获取每个所述第二预测KPI值的异常影响度，包括：
    计算所述KQI预测值和所述KQI业务门限的偏离差值；
    根据所述偏离差值计算所述第二预测KPI值的异常影响度。
  5. 根据权利要求2所述的无线侧网络根因定位方法，其中，在根据所述KQI预测值和所述KPI劣化拐点得到预测KPI集合之前，将多个所述KQI预测值按照从大到小排序。
  6. 根据权利要求1所述的无线侧网络根因定位方法，其中，所述映射函数模型包括多个映射函数；
    所述根据所述映射函数模型，得到多个基于分布的KPI劣化拐点，包括：
    将预设的KQI业务门限分别输入至多个所述映射函数中，通过函数反解得到多个基于分布的KPI劣化拐点。
  7. 根据权利要求1所述的无线侧网络根因定位方法，其中，所述获取预设的映射函数模型，包括：
    获取无线侧数据和核心网侧数据，其中，所述无线侧数据包括基础KPI数据，所述核心网侧数据包括基础KQI数据；
    对所述无线侧数据和所述核心网侧数据进行数据清洗和数据关联，得到全量数据；
    根据所述全量数据，计算所述基础KPI数据和所述基础KQI数据的相关性系数；
    根据所述相关性系数，确定影响目标KQI值的目标KPI集合，其中，所述目标KPI集合包括多个目标KPI值；
    根据所述目标KQI值和所述目标KPI值，通过逻辑回归算法构建映射函数模型，其中，所述映射函数模型包括多个用于建立KPI值和KQI值之间的映射关系的函数。
  8. 根据权利要求7所述的无线侧网络根因定位方法，其中，所述根据所述相关性系数，确定影响目标KQI值的目标KPI集合，包括：
    当所述相关性系数大于预设系数，确定影响目标KQI值的目标KPI集合。
  9. 一种运行控制装置，包括至少一个控制处理器和用于与所述至少一个控制处理器通信连接的存储器；所述存储器存储有可被所述至少一个控制处理器执行的指令，所述指令被所述至少一个控制处理器执行，以使所述至少一个控制处理器能够执行如权利要求1至8任一项所述的无线侧网络根因定位方法。
  10. 一种计算机可读存储介质，所述计算机可读存储介质存储有计算机可执行指令，所述计算机可执行指令用于使计算机执行如权利要求1至8任一项所述的无线侧网络根因定位方法。
PCT/CN2022/127442 2021-12-21 2022-10-25 无线侧网络根因定位方法、运行控制装置及存储介质 WO2023116178A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111571366.0A CN116319251A (zh) 2021-12-21 2021-12-21 无线侧网络根因定位方法、运行控制装置及存储介质
CN202111571366.0 2021-12-21

Publications (1)

Publication Number Publication Date
WO2023116178A1 true WO2023116178A1 (zh) 2023-06-29

Family

ID=86782027

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/127442 WO2023116178A1 (zh) 2021-12-21 2022-10-25 无线侧网络根因定位方法、运行控制装置及存储介质

Country Status (2)

Country Link
CN (1) CN116319251A (zh)
WO (1) WO2023116178A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117560706B (zh) * 2024-01-12 2024-03-22 亚信科技（中国）有限公司 根因分析方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106604290A (zh) * 2016-12-19 2017-04-26 南京华苏科技有限公司 基于网页浏览的用户感知测评无线网络性能方法
US20170262781A1 (en) * 2016-03-14 2017-09-14 Futurewei Technologies, Inc. Features selection and pattern mining for kqi prediction and cause analysis
US20170272319A1 (en) * 2016-03-16 2017-09-21 Futurewei Technologies, Inc. Systems and Methods for Identifying Causes of Quality Degradation in Wireless Networks
CN109947760A (zh) * 2017-07-26 2019-06-28 华为技术有限公司 一种挖掘kpi根因的方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170262781A1 (en) * 2016-03-14 2017-09-14 Futurewei Technologies, Inc. Features selection and pattern mining for kqi prediction and cause analysis
US20170272319A1 (en) * 2016-03-16 2017-09-21 Futurewei Technologies, Inc. Systems and Methods for Identifying Causes of Quality Degradation in Wireless Networks
CN106604290A (zh) * 2016-12-19 2017-04-26 南京华苏科技有限公司 基于网页浏览的用户感知测评无线网络性能方法
CN109947760A (zh) * 2017-07-26 2019-06-28 华为技术有限公司 一种挖掘kpi根因的方法及装置

Also Published As

Publication number Publication date
CN116319251A (zh) 2023-06-23

Similar Documents

Publication Publication Date Title
CN111901814B (zh) 无线网络健康评估方法、装置、电子设备及存储介质
CN101883380B (zh) 一种拥塞处理时选择终端的方法及装置
CN102077530B (zh) 检查无线业务及减缓分组删除以避免无线饱和的方法和设备
CN102104900A (zh) 用户感知的分析方法及设备
US10194338B2 (en) Network optimization method and apparatus, and base station
WO2006129792A1 (ja) 異常検出方法およびシステム、並びに、保全方法およびシステム
CN109996277B (zh) 一种判断质差小区的方法及装置
CN101142789A (zh) 区分服务的无线网络中的QoS测量和监视
EP2489156A1 (en) Network management system and method for identifying and accessing quality of service issues within a communications network
US20210051677A1 (en) Radio frequency resource allocation method, apparatus, device and system, and storage medium
US20220247634A1 (en) Information processing method, apparatus, device and computer readable storage medium
WO2023116178A1 (zh) 无线侧网络根因定位方法、运行控制装置及存储介质
CN115086986A (zh) 策略调整方法及相关设备
CN107371179B (zh) 测量结果上报方法、测量结果接收方法、相关设备和系统
CN113556761A (zh) 切片资源调整方法、系统、终端设备及存储介质
CN107155192B (zh) 一种用户体验质量评估方法及装置
CN113891336A (zh) 通信网络减频退网方法、装置、计算机设备和存储介质
CN107347196B (zh) 一种确定小区拥塞的方法及装置
CN101212771A (zh) 一种负载控制的方法、装置及系统
CN103249068A (zh) 基于群组用户终端的下行链路质量检测方法、设备和系统
CN106961694B (zh) 一种评估ptn接入环传输带宽的方法及装置
CN113923708B (zh) 辅节点的接入和释放方法、装置、电子设备及存储介质
CN107517474A (zh) 一种网络分析优化方法及装置
JP7083130B2 (ja) 通信管理システム及びプログラム
WO2022121513A1 (zh) 性能指标至差值的生成方法、装置、电子设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22909496

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE