WO2020119711A1 - 一种根因定位方法、服务器和存储介质 - Google Patents

一种根因定位方法、服务器和存储介质 Download PDF

Info

Publication number
WO2020119711A1
WO2020119711A1 PCT/CN2019/124497 CN2019124497W WO2020119711A1 WO 2020119711 A1 WO2020119711 A1 WO 2020119711A1 CN 2019124497 W CN2019124497 W CN 2019124497W WO 2020119711 A1 WO2020119711 A1 WO 2020119711A1
Authority
WO
WIPO (PCT)
Prior art keywords
root cause
key performance
data set
performance indicator
correlation coefficient
Prior art date
Application number
PCT/CN2019/124497
Other languages
English (en)
French (fr)
Inventor
熊科
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 WO2020119711A1 publication Critical patent/WO2020119711A1/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
    • H04L41/0677—Localisation of faults
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/04—Arrangements for maintaining operational condition
    • 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
    • H04L41/065—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 involving logical or physical relationship, e.g. grouping and hierarchies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/50—Testing arrangements

Definitions

  • the embodiments of the present disclosure relate to the technical field of communications, and in particular, to a root cause location method, a server, and a storage medium.
  • the main method for analyzing wireless network problems is to use the thresholds given by expert experience to judge wireless network indicators. And through human analysis data to achieve the root cause of wireless network problems.
  • the simple method of locating the root cause of the problem through human analysis data not only requires a lot of labor costs, but as the current wireless network becomes more and more complex, it is difficult to rely solely on expert thresholds Accurate and effective root cause positioning, so the root cause positioning method in some cases cannot meet the actual needs of users.
  • the purpose of the embodiments of the present disclosure is to provide a root cause location method, a server, and a storage medium, so that when the key performance indicators in the data set to be tested are determined to be abnormal, accurate location of the root cause can be automatically achieved.
  • the embodiments of the present disclosure provide a root cause location method, which includes the following steps: detecting key performance indicators in the data set to be tested, and extracting data of abnormal key performance indicators in the data set to be detected to obtain the first A data set; calculate the correlation coefficient matrix between the key performance indicators and multiple non-key performance indicators in the first data set; determine the root cause of the abnormal key performance indicators according to the correlation coefficient matrix and the known root cause positioning tree, where, The root cause location tree stores the correspondence between key performance indicators, multiple non-key performance indicators, and root causes.
  • An embodiment of the present disclosure also provides a server including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are at least one The processor executes to enable at least one processor to execute the root cause locating method as described above.
  • Embodiments of the present disclosure also provide a computer-readable storage medium that stores a computer program, and when the computer program is executed by a processor, the root cause location method described above is implemented.
  • FIG. 1 is a flowchart of a root cause location method in the first embodiment of the present disclosure
  • FIG. 2 is a schematic structural diagram of a root cause location tree in the first embodiment of the present disclosure
  • FIG. 3 is a schematic diagram of correlation coefficient labeling between neighboring nodes in a root cause location tree in the first embodiment of the present disclosure
  • FIG. 5 is a schematic structural diagram of a server in a third embodiment of the present disclosure.
  • the first embodiment of the present disclosure relates to a root cause location method applied to a server.
  • the process is shown in Figure 1 and includes the following steps:
  • Step 101 Detect key performance indicators in the data set to be tested, and extract data with abnormal key performance indicators in the data set to be detected to obtain the first data set.
  • the data set to be tested in this embodiment may be base station performance statistical data in a designated area of the entire network, or measurement report data reported by the terminal within a period of time.
  • it is generally necessary to pre-process the data fill in the missing data, or remove the obvious abnormal data.
  • One of the performance indicators is selected as a key performance indicator, and the remaining performance indicators can be used as non-key performance indicators, and the key performance indicators in this embodiment refer to performance indicators that the service provider pays attention to.
  • the performance indicators of the data set to be tested include: call accessibility-ACCESSIBILITY, reference signal received power-RSRP, downlink signal to interference plus noise ratio-DLSINR, average distance from base station to user-DISTANCE and LTE reference
  • the signal reception quality-RSRQ uses ACCESSIBILITY as a key performance indicator, and the remaining performance indicators as non-key performance indicators as an example.
  • the following can be used: calculating the standard deviation ⁇ and the mean ⁇ of the key performance indicators ACCESSIBILITY in the data set to be tested, and according to the standard Difference and mean construct confidence interval [ ⁇ -nu, ⁇ +nu], using the minimum value in the confidence interval as the lower confidence limit and the maximum value in the confidence interval as the upper confidence limit, so that the key performance indicators in the data set to be tested are located in the confidence interval Other data is regarded as abnormal data.
  • the key performance indicator ACCESSIBILITY of the data set to be tested is considered abnormal, and the key performance indicator of the data set to be tested is located in the confidence interval
  • Step 102 Calculate the correlation coefficient matrix between the key performance indicators and multiple non-key performance indicators in the first data set.
  • the first data set is shown in Table 1 below.
  • Euclidean distance Euclidean distance
  • Manhattan distance Chebyshev distance
  • Jaccard coefficient Jaccard coefficient
  • cosine similarity Algorithms such as degree and Pearson correlation coefficient calculate the correlation coefficient matrix between key performance indicators and multiple non-key performance indicators.
  • Pearson correlation coefficient algorithm is used as an example for description.
  • the correlation coefficient between the first column of key performance indicators ACCESSIBILITY and the second column of non-key performance indicators RSRP in the first data set is calculated using the following formula (1):
  • r 12 represents the correlation coefficient between the key performance indicator ACCESSIBILITY and the non-key performance indicator RSRP
  • m represents the total amount of data in the first data set
  • x i1 represents the key performance indicator ACCESSIBILITY in the sample with the serial number i in the first data set Value
  • x i2 The value of the non-key performance indicator RSRP in the sample with sequence number i in the first data set.
  • r 12 represents the correlation coefficient between the key performance indicator ACCESSIBILITY and the non-key performance indicator RSRP
  • m represents the total amount of data in the first data set
  • x i1 represents the key performance indicator ACCESSIBILITY in the sample with the serial number i in the first data set Value
  • x i2 The value of the non-key performance indicator RSRP in the sample with sequence number i in the first data set.
  • Table 2 As shown, it is a correlation coefficient matrix finally obtained according to the first data set D in the embodiment of the present disclosure
  • Step 103 Determine the root cause when the key performance index is abnormal according to the correlation coefficient matrix and the known root cause locating tree.
  • the root cause location tree in this embodiment stores the correspondence between key performance indicators, multiple non-key performance indicators, and root causes.
  • the root cause location tree whose key performance index is ACCESSIBILITY Among them, each branch of the root cause location tree represents the possible different root causes corresponding to the failure of ACCESSIBILITY, that is, the poor accessibility of the communication call.
  • the root cause location tree includes: root nodes, intermediate nodes, and leaf nodes, where the root node represents a key performance indicator, the leaf node represents a non-key performance indicator directly related to the root cause, and the intermediate node represents a non-indirect related non-related root cause Key performance indicators.
  • the root cause of the abnormal performance of the key performance indicator ACCESSIBILITY is determined by adopting the following method: determining a branch with the largest sum of correlation coefficients in the root cause location tree according to the correlation coefficient matrix; and maximizing the sum of correlation coefficients
  • the root cause corresponding to the leaf node in a branch is used as the root cause when the key performance index is abnormal.
  • the root cause of the ACCESSIBILITY key tree includes three branches, the first branch is ACCESSIBILITY->RSRP, the second branch is ACCESSIBILITY->RSRQ->DISTANCE, and the third branch is ACCESSIBILITY ->RSRQ->DLSINR, and combine the correlation coefficient matrix in Table 2 to obtain the correlation coefficient between adjacent nodes of each branch, as shown in FIG.
  • the correlation coefficient product between adjacent nodes of the first branch The result is 0.34
  • the value of 0.48 is the largest, so the third branch is taken as the branch with the largest correlation coefficient product. Because the root corresponding to the leaf node in the third branch is "downlink interference", it can be directly determined that the cause of the poor accessibility of the current communication call is the presence of downlink interference. After finding the root cause, the operation and maintenance personnel can adjust and maintain the system in time according to the results obtained.
  • the root cause location method calculates the difference between the key performance indicators and non-key performance indicators in the abnormal part of the data to be detected when it is determined that the key performance indicators in the data set to be tested are abnormal Correlation coefficient matrix, by searching the correlation coefficient matrix on the basis of the known root cause locating tree, the root cause corresponding to the critical performance abnormality can be automatically obtained, and the obtained root cause is also more accurate.
  • the second embodiment of the present disclosure relates to a root cause location method.
  • This embodiment has been further improved on the basis of the first embodiment.
  • the improvement is that the process of constructing the root cause positioning tree is described.
  • the flow of the root cause location method in this embodiment is shown in FIG. 4.
  • step 201 to step 205 are included, where step 203 to step 205 are substantially the same as step 101 to step 103 in the first embodiment, and will not be repeated here.
  • the following mainly introduces the differences, which are not described in this
  • Step 201 Obtain a historical sample data set.
  • the historical sample data set in this embodiment is a data set formed after data cleaning.
  • L ⁇ y 1 ,y 2 ...y m ⁇ , indicating that the historical sample data set L includes m data, and each data in the historical sample data set includes key performance indicators and multiple non-key performance indicators.
  • Step 202 construct a root cause localization tree according to the historical sample data set.
  • the first performance index corresponding to each root cause is determined, and the first performance index is a non-key performance index directly related to the root cause; the second performance index corresponding to each root cause is determined, and The second performance index is a non-key performance index that is indirectly related to the root cause; according to the key performance index, the first performance index, and the second performance index, the training set corresponding to each root cause is obtained from the historical sample data set; according to each The training set corresponding to the root cause constructs the root cause localization tree.
  • the non-key performance indicator RSRP directly corresponds to the root cause: weak downlink coverage
  • the non-key performance indicator DISTANCE directly corresponds to the root cause: the antenna downtilt angle is too large
  • the non-key performance indicator DLSINR directly corresponds to the root cause: downlink interference. Therefore, for the root cause of "too large antenna downtilt angle", DISTANCE can be used as the first performance index
  • the non-key performance indicators that may be related to the root cause of "root antenna downtilt angle too large” also include RSRQ based on the historical experience of experts. RSRQ can be used as the second performance indicator.
  • the method used is to calculate the key performance indicators and non-critical performance of the training set corresponding to each root cause. Correlation coefficient between indicators; and determine the optimal path of key performance indicators, non-key performance indicators and root causes according to the correlation coefficient, and use the optimal path as a branch of the root cause positioning tree; determine each of the root cause positioning trees The branch corresponding to the root cause; merge the branches corresponding to each root cause to construct the root cause positioning tree.
  • the method of calculating the correlation coefficient between the key performance indicators and non-key performance indicators in the third training data set may also use Peel Inferior correlation coefficient algorithm, the calculation process is almost the same as that in the first embodiment, so no more details will be given in this embodiment.
  • the correlation coefficient between ACCESSIBILITY and RSRP is 0.8, the correlation coefficient between RSRP and RSRQ is 0.6, and the correlation coefficient between ACCESSIBILITY and RSRQ is 0.3, indicating that For the root cause "downlink weak coverage", the non-key performance indicator RSRQ determined by expert experience is invalid, so the branch corresponding to the root cause "downlink weak coverage” is: ACCESSIBILITY->RSRP, and the root cause can be obtained by the same reason
  • the branch corresponding to "antenna downtilt angle is too large” is "ACCESSIBILITY->RSRQ->DISTANCE", and the branch corresponding to "downlink interference” is "ACCESSIBILITY->RSRQ->DLSINR".
  • the structure of the root cause localization tree shown in Fig. 2 is merged.
  • the performance index ACCESSIBILITY is taken as an example of key performance indicators.
  • the method of constructing a root cause positioning tree using other performance indicators as key performance indicators is almost the same as that of this embodiment, so this embodiment does not Repeat them again.
  • steps 203 to 205 are performed.
  • the root cause locating method when it is determined that the key performance index in the data set to be tested is abnormal, calculates the key performance index and non-key in the abnormal part of the data to be detected
  • the correlation coefficient matrix between the performance indicators by searching the correlation coefficient matrix on the basis of the known root cause locating tree, can automatically obtain the root cause corresponding to the critical performance abnormality, and the obtained root cause is also more accurate.
  • the root cause localization tree is constructed from the acquired historical sample data sets, and no specific experts are involved.
  • a third embodiment of the present disclosure relates to a server, as shown in FIG. 5, including at least one processor 501; and a memory 502 communicatively connected to the at least one processor 501; wherein the memory 502 stores at least one processor
  • the instruction executed by 501 is executed by at least one processor 501, so that the at least one processor 501 can execute the root cause locating method in the foregoing embodiment.
  • the processor 501 uses a Central Processing Unit (CPU) as an example, and the memory 502 uses a Random Access Memory (RAM) as an example.
  • the processor 501 and the memory 502 may be connected through a bus or other means. In FIG. 5, the connection through a bus is used as an example.
  • the memory 502 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs, and modules.
  • the program for implementing the root cause location method in the embodiments of the present disclosure is Stored in the memory 502.
  • the processor 501 runs non-volatile software programs, instructions, and modules stored in the memory 502 to execute various functional applications and data processing of the device, that is, to implement the above-mentioned root cause location method.
  • the memory 502 may include a storage program area and a storage data area, wherein the storage program area may store an operating system and application programs required by at least one function; the storage data area may store a list of options, and the like.
  • the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices.
  • the memory 502 may optionally include memories remotely set with respect to the processor 501, and these remote memories may be connected to an external device through a network. Examples of the above network include but are not limited to the Internet, intranet, local area network, mobile communication network, and combinations thereof.
  • One or more program modules are stored in the memory 502, and when executed by one or more processors 501, execute the root cause location method in any of the above method embodiments.
  • a sixth embodiment of the present disclosure relates to a computer-readable storage medium that stores a computer program, which when executed by a processor can implement root cause location involved in any method embodiment of the present disclosure method.
  • the embodiments of the present disclosure calculate the correlation coefficient matrix between the key performance indicators and non-key performance indicators in the abnormal part of the data to be detected when it is determined that the key performance indicators in the data set to be tested are abnormal.
  • the root cause corresponding to the abnormal key performance can be automatically obtained, and the obtained root cause is also more accurate. This makes it possible to automatically locate the root cause automatically when the key performance indicators in the data set to be tested are abnormal.
  • a storage medium includes several instructions to make a device (may be A single chip microcomputer, a chip, etc.) or a processor executes all or part of the steps of the methods described in various embodiments of the present disclosure.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code .

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Position Fixing By Use Of Radio Waves (AREA)

Abstract

本公开实施例涉及通信技术领域，公开了一种根因定位方法、服务器和存储介质。本公开中，检测待测试数据集中的关键性能指标，并将待检测数据集中关键性能指标异常的数据提取出来获得第一数据集；计算第一数据集中关键性能指标和多个非关键性能指标之间的相关系数矩阵；根据相关系数矩阵和已知的根因定位树确定关键性能指标异常时的根因，其中，根因定位树中保存了关键性能指标、多个非关键性能指标和根因的对应关系。

Description

一种根因定位方法、服务器和存储介质
本公开要求享有2018年12月13日提交的名称为“一种根因定位方法、服务器和存储介质”的中国专利申请CN201811525550.X的优先权，其全部内容通过引用并入本文中。
技术领域
本公开实施例涉及通信技术领域，特别涉及一种根因定位方法、服务器和存储介质。
背景技术
随着无线通信技术的更新换代，无线网络系统结构日趋复杂，无线网络出现的问题越来越多，而目前分析无线网络问题的主要方法是使用专家经验给出的阈值对无线网络指标进行判断，并通过人力分析数据实现对无线网络问题的根因定位。
发明人发现一些情况中至少存在如下问题：单纯的通过人力分析数据定位问题根因的方法，不仅需要投入大量的人力成本，而且随着当前无线网络越来越复杂，单纯的依靠专家阈值很难准确有效的实现根因定位，因此一些情况中的根因定位方法无法满足用户的实际需求。
发明内容
本公开实施方式的目的在于提供一种根因定位方法、服务器和存储介质，使得能够在确定待测试数据集中的关键性能指标异常的情况下，自动实现对根因的准确定位。
为解决上述技术问题，本公开的实施方式提供了一种根因定位方法，包括以下步骤：检测待测试数据集中的关键性能指标，并将待检测数据集中关键性能指标异常的数据提取出来获得第一数据集；计算第一数据集中关键性能指标和多个非关键性能指标之间的相关系数矩阵；根据相关系数矩阵和已知的根因定位树确定关键性能指标异常时的根因，其中，根因定位树中保存了关键性能指标、多个非关键性能指标和根因的对应关系。
本公开的实施方式还提供了一种服务器，包括：至少一个处理器；以及，与至少一个处理器通信连接的存储器；其中，存储器存储有可被至少一个处理器执行的指令，指令被至少一个处理器执行，以使至少一个处理器能够执行如上所述的根因定位方法。
本公开的实施方式还提供了一种计算机可读存储介质，存储有计算机程序，计算机程序被处理器执行时实现如上所述的根因定位方法。
附图说明
一个或多个实施例通过与之对应的附图中的图片进行示例性说明，这些示例性说明并不 构成对实施例的限定，附图中具有相同参考数字标号的元件表示为类似的元件，除非有特别申明，附图中的图不构成比例限制。
图1是本公开第一实施例中根因定位方法的流程图；
图2是本公开第一实施例中根因定位树的结构示意图；
图3是本公开第一实施例中根因定位树中相邻节点之间的相关系数标注示意图；
图4是本公开第二实施例中根因方法的流程图；
图5是本公开第三实施例中服务器的结构示意图。
具体实施方式
为使本公开实施例的目的、技术方案和优点更加清楚，下面将结合附图对本公开的各实施方式进行详细的阐述。然而，本领域的普通技术人员可以理解，在本公开各实施方式中，为了使读者更好地理解本公开而提出了许多技术细节。但是，即使没有这些技术细节和基于以下各实施方式的种种变化和修改，也可以实现本公开所要求保护的技术方案。
本公开的第一实施方式涉及一种根因定位方法，应用于服务器。流程如图1所示，包括以下步骤：
步骤101，检测待测试数据集中的关键性能指标，并将待检测数据集中关键性能指标异常的数据提取出来获得第一数据集。
在一个实施例中，本实施例中的待测试数据集可以是全网指定区域内的基站性能统计数据，或者是一段时间内终端上报的测量报告数据。并且在获得待测试数据集后一般还需要对数据进行预处理，对缺失的数据进行填充或者对明显异常的数据进行剔除。例如，经过预处理后的待测试数据集D＝{x 1,x 2...x n}，表示待测试数据集D中包括n个数据，并且每一个数据包括多个性能指标，从多个性能指标中选择出一个作为关键性能指标，其余性能指标就可以作为非关键性能指标，并且本实施方式中的关键性能指标指的是服务商所关注的性能指标。
在一个实施例中，待测试数据集的性能指标包括：呼叫接入性-ACCESSIBILITY、参考信号接收功率-RSRP、下行信号与干扰加噪声比-DLSINR、基站到用户的平均距离-DISTANCE和LTE参考信号接收质量-RSRQ，在本实施方式中是将ACCESSIBILITY作为关键性能指标，其余性能指标作为非关键性能指标为例进行的说明。
在一个实施例中，本实施方式中在判断待测试数据集中的关键性能指标，例如ACCESSIBILITY是否异常时，可以采用：计算待检测数据集中关键性能指标ACCESSIBILITY的标准差σ和均值μ，并根据标准差和均值构建置信区间[σ-nu,σ+nu]，将置信区间中的最小值作为置信下限，将置信区间中的最大值作为置信上限，从而将待测试数据集中关键性能指标位于置信区间之外的数据作为异常数据，当异常数据与待测试数据集总数据量的比值超过10％时，则认为待测试数据集关键性能指标ACCESSIBILITY异常，并将待测试数据集中 关键性能指标位于置信区间之外的数据提取出来作为第一数据集D1＝{x 1,x 2...x m}，并且m≤n。
步骤102，计算第一数据集中关键性能指标和多个非关键性能指标之间的相关系数矩阵。
在一个实施例中，在本实施方式中第一数据集如下表1所示，根据第一样本数据集D1可以采用欧氏距离、曼哈顿距离、切比雪夫距离、杰卡德系数、余弦相似度、皮尔逊相关系数等算法计算关键性能指标和多个非关键性能指标之间的相关系数矩阵，本实施方式中以皮尔逊相关系数算法为例进行说明。
例如，计算第一数据集中第一列关键性能指标ACCESSIBILITY和第二列非关键性能指标RSRP之间的相关系数利用如下公式(1)计算获得：
Figure PCTCN2019124497-appb-000001
其中，r 12表示关键性能指标ACCESSIBILITY和非关键性能指标RSRP之间的相关系数，m表示第一数据集中的数据总量，x i1表示第一数据集中序号为i的样本中关键性能指标ACCESSIBILITY的数值，x i2第一数据集中序号为i的样本中非关键性能指标RSRP的数值。当然本实施方式中仅是以ACCESSIBILITY和RSRP之间的相关系数为例进行的举例说明，对于其它性能指标之间相关系数的计算方式与此相同，本实施方式中不再进行赘述，如表2所示，为本公开实施方式中根据第一数据集D所最终获得的相关系数矩阵。
表1
Figure PCTCN2019124497-appb-000002
表2
Figure PCTCN2019124497-appb-000003
Figure PCTCN2019124497-appb-000004
步骤103，根据相关系数矩阵和已知的根因定位树确定关键性能指标异常时的根因。
在一个实施例中，在本实施方式中根因定位树中保存了关键性能指标、多个非关键性能指标和根因的对应关系，如图2所示为关键性能指标为ACCESSIBILITY的根因定位树，其中，根因定位树的每一个分支分代表了当ACCESSIBILITY发生故障，即通信呼叫接入性差时所对应的可能的不同根因。并且，根因定位树包括：根节点、中间节点和叶子节点，其中，根节点表示关键性能指标，叶子节点表示与根因直接相关的非关键性能指标，中间节点表示与根因间接相关的非关键性能指标。
其中，在本实施方式中，确定关键性能指标ACCESSIBILITY异常时的根因，所采用的方式为：根据相关系数矩阵确定根因定位树中相关系数之和最大的一个分支；将相关系数之和最大的一个分支中的叶子节点所对应的根因作为关键性能指标异常时的根因。
需要说明的是，在确定根因定位树中相关系数之和最大的一个分支时，通过确定根因定位树中每一个分支所包含的根节点、中间节点和叶子节点；根据相关系数矩阵查找每一个分支中相邻节点之间的相关系数，并将每一个分支中所有相邻节点之间的相关系相乘，获得每一个分支的相关系数乘积结果；将每一个分支中的相关系数乘积结果进行对比，确定相关系数乘积结果最大的一个分支。
例如，如图2所示关键性能指标为ACCESSIBILITY的根因定位树包括三个分支，第一分支为ACCESSIBILITY->RSRP，第二个分支为ACCESSIBILITY->RSRQ->DISTANCE，第三个分支为ACCESSIBILITY->RSRQ->DLSINR，并且结合表2中的相关系数矩阵获得每一个分支的相邻节点之间的相关系数，如图3所示，第一个分支的相邻节点之间的相关系数乘积结果为0.34，第二个分支的相邻节点之间的相关系数乘积结果为0.6*0.32＝0.192，第三个分支的相邻节点之间的相关系数乘积结果为0.6*0.8＝0.48，经过对比0.48的数值最大，所以将第三个分支作为相关系数乘积结果最大的一个分支。因为第三个分支中的叶子节点所对应的根因为“下行干扰”，因此可以直接确定造成当前通信呼叫接入性差时的原因是存在下行干扰。在得出根因后，运维人员可以根据所得出的结果及时进行系统的调整和维修。
与一些情况相比，本实施方式提供的根因定位方法，当确定待测试数据集中的关键性能指标异常的情况下，计算待检测数据异常部分中的关键性能指标和非关键性能指标之间的相关系数矩阵，在已知的根因定位树的基础上通过查找相关系数矩阵，可以自动获取到关键性能异常时所对应的根因，并且所获得的根因也更加准确。
本公开的第二实施方式涉及一种根因定位方法。本实施例在第一实施例的基础上做了进一步改进，改进之处为：对构造根因定位树的过程进行了说明。本实施例中的根因定位方法的流程如图4所示。在本实施例中，包括步骤201至步骤205，其中步骤203至步骤205与第一实施方式中的步骤101至步骤103大致相同，此处不再赘述，下面主要介绍不同之处，未在本实施方式中详尽描述的技术细节，可参见第一实施例所提供的根因定位方法，此处不再赘述。
步骤201，获取历史样本数据集。
在一个实施例中，本实施方式中的历史样本数据集是通过数据清洗之后所形成的数据集。例如，L＝{y 1,y 2...y m}，表示历史样本数据集L中包括m个数据，并且历史样本数据集中的每一个数据包括关键性能指标和多个非关键性能指标。
步骤202，根据历史样本数据集构造根因定位树。
在本实施方式中，确定每一个根因所对应的第一性能指标，而第一性能指标为与根因直接相关的非关键性能指标；确定每一个根因所对应的第二性能指标，而第二性能指标为与根因间接相关的非关键性能指标；根据关键性能指标、第一性能指标和第二性能指标，从历史样本数据集中获取每一个根因所对应的训练集；根据每一个根因所对应的训练集构建根因定位树。
例如，非关键性能指标RSRP直接对应根因：下行弱覆盖；非关键性能指标DISTANCE直接对应根因：天线下倾角过大；非关键性能指标DLSINR直接对应根因：下行干扰。因此针对根因“天线下倾角过大”，可以将DISTANCE作为第一性能指标，而根据专家历史经验确定与根因“天线下倾角过大”可能存在关系的非关键性能指标还包括RSRQ，所以可以将RSRQ作为第二性能指标。因此，可以从历史样本数据获得仅包含{ACCESSIBILITY，RSRQ，DISTANCE}性能指标的数据，获得根因“天线下倾角过大”所对应的第一训练数据集，同理，可以从历史样本数据获得仅包含{ACCESSIBILITY，RSRQ，DLSINR}性能指标的数据，获得根因“下行干扰”所对应的第二训练数据集；从历史样本数据获得仅包含{ACCESSIBILITY，RSRP，RSRQ}性能指标的数据，获得根因“下行弱覆盖”所对应的第三训练数据集。
需要说明的是，本实施方式中在根据每一个根因所对应的训练集构造根因定位树时，采用的方式是，计算每一个根因所对应的训练集中的关键性能指标和非关键性能指标之间的相关系数；并根据相关系数确定关键性能指标、非关键性能指标和根因的最优路径，并将最优路径作为根因定位树的一个分支；确定根因定位树中每一个根因所对应的分支；将每一个根 因所对应的分支进行合并，构建根因定位树。
例如，以第三训练数据集为例说明根因“下行弱覆盖”所对应的分支，在计算第三训练数据集中的关键性能指标和非关键性能指标之间的相关系数的方式也可以采用皮尔逊相关系数算法，计算过程与第一实施方式大致相同，因此本实施方式中不再进行赘述。如果经过计算确定对于根因“下行弱覆盖”来说，ACCESSIBILITY与RSRP之间的相关系数为0.8，RSRP与RSRQ之间的相关系数为0.6，ACCESSIBILITY与RSRQ之间的相关系数为0.3，说明对于根因“下行弱覆盖”来说通过专家经验所确定的非关键性能指标RSRQ是无效的，因此对于根因“下行弱覆盖”所对应的分支为：ACCESSIBILITY->RSRP，同理可以获得根因“天线下倾角过大”所对应的分支为“ACCESSIBILITY->RSRQ->DISTANCE”，根因“下行干扰”所对应的分支为“ACCESSIBILITY->RSRQ->DLSINR”，将所获得的各个分支进行合并获得如图2所示的根因定位树的结构。当然，本实施方式中是以性能指标ACCESSIBILITY作为关键性能指标为例进行的说明，对于将其它性能指标作为关键性能指标构建根因定位树的方式与本实施方式大致相同，因此本实施方式中不再进行赘述。
在步骤202后，执行步骤203至205。
与一些情况相比，本实施方式提供的本实施方式提供的根因定位方法，当确定待测试数据集中的关键性能指标异常的情况下，计算待检测数据异常部分中的关键性能指标和非关键性能指标之间的相关系数矩阵，在已知的根因定位树的基础上通过查找相关系数矩阵，可以自动获取到关键性能异常时所对应的根因，并且所获得的根因也更加准确。并且在进行根因定位的过程中通过获取的历史样本数据集构造根因定位树，不需要涉及具体的专家参与其中。
上面各种方法的步骤划分，只是为了描述清楚，实现时可以合并为一个步骤或者对某些步骤进行拆分，分解为多个步骤，只要包括相同的逻辑关系，都在本专利的保护范围内；对算法中或者流程中添加无关紧要的修改或者引入无关紧要的设计，但不改变其算法和流程的核心设计都在该专利的保护范围内。
本公开第三实施方式涉及一种服务器，如图5所示，包括至少一个处理器501；以及，与至少一个处理器501通信连接的存储器502；其中，存储器502存储有可被至少一个处理器501执行的指令，指令被至少一个处理器501执行，以使至少一个处理器501能够执行上述实施例中的根因定位方法。
本实施例中，处理器501以中央处理器(Central Processing Unit，CPU)为例，存储器502以可读写存储器(Random Access Memory，RAM)为例。处理器501、存储器502可以通过总线或者其他方式连接，图5中以通过总线连接为例。存储器502作为一种非易失性计算机可读存储介质，可用于存储非易失性软件程序、非易失性计算机可执行程序以及模块，如本公开实施例中实现根因定位方法的程序就存储于存储器502中。处理器501通过运行存储在存储器502中的非易失性软件程序、指令以及模块，从而执行设备的各种功能应用以及 数据处理，即实现上述根因定位方法。
存储器502可以包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需要的应用程序；存储数据区可存储选项列表等。此外，存储器可以包括高速随机存取存储器，还可以包括非易失性存储器，例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实施例中，存储器502可选包括相对于处理器501远程设置的存储器，这些远程存储器可以通过网络连接至外接设备。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
一个或者多个程序模块存储在存储器502中，当被一个或者多个处理器501执行时，执行上述任意方法实施例中的根因定位方法。
上述产品可执行本公开实施例所提供的方法，具备执行方法相应的功能模块和有益效果，未在本实施例中详尽描述的技术细节，可参见本公开实施例所提供的方法。
本公开的第六实施方式涉及一种计算机可读存储介质，该计算机可读存储介质中存储有计算机程序，该计算机程序被处理器执行时能够实现本公开任意方法实施例中涉及的根因定位方法。
本公开实施方式相对于一些情况而言，当确定待测试数据集中的关键性能指标异常的情况下，计算待检测数据异常部分中的关键性能指标和非关键性能指标之间的相关系数矩阵，在已知的根因定位树的基础上通过查找相关系数矩阵，可以自动获取到关键性能异常时所对应的根因，并且所获得的根因也更加准确。使得能够在确定待测试数据集中的关键性能指标异常的情况下，自动实现对根因的准确定位。
本领域技术人员可以理解，实现上述实施例方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序存储在一个存储介质中，包括若干指令用以使得一个设备(可以是单片机，芯片等)或处理器(processor)执行本公开各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
本领域的普通技术人员可以理解，上述各实施方式是实现本公开的实施例，而在实际应用中，可以在形式上和细节上对其作各种改变，而不偏离本公开的精神和范围。

Claims (10)

  1. 一种根因定位方法，其中，应用于服务器，包括：
    检测待测试数据集中的关键性能指标，并将所述待检测数据集中关键性能指标异常的数据提取出来获得第一数据集；
    计算所述第一数据集中所述关键性能指标和多个非关键性能指标之间的相关系数矩阵；
    根据所述相关系数矩阵和已知的根因定位树确定所述关键性能指标异常时的根因，其中，所述根因定位树中保存了关键性能指标、所述多个非关键性能指标和根因的对应关系。
  2. 根据权利要求1所述的根因定位方法，其中，所述根据所述相关系数矩阵和已知的根因定位树确定所述关键性能指标异常时的根因之前，还包括：
    获取历史样本数据集，其中，所述历史样本数据集中的每一个数据包括所述关键性能指标和所述多个非关键性能指标；
    根据所述历史样本数据集构造所述根因定位树。
  3. 根据权利要求2所述的根因定位方法，其中，所述根据所述历史样本数据集构造所述根因定位树，包括：
    确定每一个根因所对应的所述第一性能指标，其中，所述第一性能指标为与所述根因直接相关的所述非关键性能指标；
    确定每一个所述根因所对应的第二性能指标，其中，所述第二性能指标为与所述根因间接相关的所述非关键性能指标；
    根据所述关键性能指标、所述第一性能指标和所述第二性能指标，从所述历史样本数据集中获取每一个所述根因所对应的训练集；
    根据每一个所述根因所对应的训练集构建所述根因定位树。
  4. 根据权利要求3所述的根因定位方法，其中，所述根据每一个所述根因所对应的训练集构建所述根因定位树，包括：
    计算每一个所述根因所对应的训练集中的所述关键性能指标和所述非关键性能指标之间的相关系数；
    并根据所述相关系数确定所述关键性能指标、所述非关键性能指标和所述根因的最优路径，并将所述最优路径作为所述根因定位树的一个分支；
    确定所述根因定位树中每一个根因所对应的分支；
    将每一个根因所对应的分支进行合并，构建所述根因定位树。
  5. 根据权利要求4所述的根因定位方法，其中，所述根因定位树包括：根节点、中间节点和叶子节点，其中，所述根节点表示所述关键性能指标，所述叶子节点表示与所述根因直 接相关的所述非关键性能指标，所述中间节点表示与所述根因间接相关的所述非关键性能指标。
  6. 根据权利要求1所述的根因定位方法，其中，所述检测待测试数据集中的关键性能指标，并将所述待检测数据集中关键性能指标异常的数据提取出来获得第一数据集，包括：
    计算所述待检测数据集中所述关键性能指标的标准差和均值；
    根据所述标准差和所述均值构建置信区间；
    将所述待检测数据集中关键性能指标位于所述置信区间之外的数据作为所述关键性能指标异常的数据；
    将所述待检测数据集中关键性能指标位于所述置信区间之外的数据提取出来作为所述第一数据集。
  7. 根据权利要求5所述的根因定位方法，其中，所述根据所述相关系数矩阵和已知的根因定位树确定所述关键性能指标异常时的根因，包括：
    根据所述相关系数矩阵确定所述根因定位树中相关系数乘积结果最大的一个分支；
    将所述相关系数乘积结果最大的一个分支中的叶子节点所对应的根因作为所述关键性能指标异常时的根因。
  8. 根据权利要求7所述的根因定位方法，其中，所述根据所述相关系数矩阵确定所述根因定位树中相关系数乘积结果最大的一个分支，包括：
    确定所述根因定位树中每一个分支所包含的所述根节点、所述中间节点和所述叶子节点；
    根据所述相关系数矩阵查找所述每一个分支中相邻节点之间的相关系数，并将所述每一个分支中所有相邻节点之间的相关系相乘，获得所述每一个分支的相关系数乘积结果；
    将所述每一个分支中的相关系数乘积结果进行对比，确定所述相关系数乘积结果最大的一个分支。
  9. 一种服务器，其中，包括：
    至少一个处理器；以及，
    与所述至少一个处理器通信连接的存储器；其中，
    所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行如权利要求1至8任一项所述的根因定位方法。
  10. 一种计算机可读存储介质，存储有计算机程序，其中，所述计算机程序被处理器执行时实现权利要求1至8任一项所述的根因定位方法。
PCT/CN2019/124497 2018-12-13 2019-12-11 一种根因定位方法、服务器和存储介质 WO2020119711A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811525550.XA CN111327435B (zh) 2018-12-13 2018-12-13 一种根因定位方法、服务器和存储介质
CN201811525550.X 2018-12-13

Publications (1)

Publication Number Publication Date
WO2020119711A1 true WO2020119711A1 (zh) 2020-06-18

Family

ID=71076067

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/124497 WO2020119711A1 (zh) 2018-12-13 2019-12-11 一种根因定位方法、服务器和存储介质

Country Status (2)

Country Link
CN (1) CN111327435B (zh)
WO (1) WO2020119711A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112882796A (zh) * 2021-02-25 2021-06-01 深信服科技股份有限公司 异常根因分析方法和装置，及存储介质
CN114338351A (zh) * 2021-12-31 2022-04-12 天翼物联科技有限公司 网络异常根因确定方法、装置、计算机设备及存储介质
CN114760186A (zh) * 2022-03-23 2022-07-15 深信服科技股份有限公司 告警分析方法、装置、电子设备及存储介质
CN115150250A (zh) * 2021-03-31 2022-10-04 中国电信股份有限公司 基于因果学习的物联网异常根因定位方法以及装置
CN112882796B (zh) * 2021-02-25 2024-05-24 深信服科技股份有限公司 异常根因分析方法和装置，及存储介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112446647A (zh) * 2020-12-14 2021-03-05 上海众源网络有限公司 异常元素的定位方法、装置、电子设备及存储介质
CN113011707B (zh) * 2021-02-19 2024-03-01 阳光保险集团股份有限公司 一种面向流程阶段的指标异常的根因定位方法及系统
CN114760190B (zh) * 2022-04-11 2023-06-20 北京邮电大学 一种面向服务的融合网络性能异常检测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150205856A1 (en) * 2014-01-21 2015-07-23 Decision Makers - Learning & Research Simulations Ltd. Dynamic brownian motion with density superposition for abnormality detection
CN105096041A (zh) * 2015-07-24 2015-11-25 北京中电普华信息技术有限公司 一种指标异动的溯源及预测方法、装置
CN107480849A (zh) * 2017-06-28 2017-12-15 北京邮电大学 一种应用于电网的空间降维方法及装置
CN107579846A (zh) * 2017-08-23 2018-01-12 昆明理工大学 一种云计算故障数据检测方法及系统
CN108173670A (zh) * 2016-12-07 2018-06-15 华为技术有限公司 检测网络的方法和装置

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150333998A1 (en) * 2014-05-15 2015-11-19 Futurewei Technologies, Inc. System and Method for Anomaly Detection
US9807229B1 (en) * 2014-10-10 2017-10-31 Carrier Iq, Inc. Wireless customer experience fault analytics system and method of operation
US9424121B2 (en) * 2014-12-08 2016-08-23 Alcatel Lucent Root cause analysis for service degradation in computer networks
CN107171819B (zh) * 2016-03-07 2020-02-14 北京华为数字技术有限公司 一种网络故障诊断方法及装置
CN105873105B (zh) * 2016-04-22 2018-07-03 中国科学技术大学 一种基于网络体验质量的移动通信网异常检测和定位方法
CN108009040B (zh) * 2017-12-12 2021-05-04 杭州时趣信息技术有限公司 一种确定故障根因的方法、系统和计算机可读存储介质
CN108322917B (zh) * 2018-05-14 2022-02-15 北京星网锐捷网络技术有限公司 无线网络接入故障定位方法、设备、系统及存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150205856A1 (en) * 2014-01-21 2015-07-23 Decision Makers - Learning & Research Simulations Ltd. Dynamic brownian motion with density superposition for abnormality detection
CN105096041A (zh) * 2015-07-24 2015-11-25 北京中电普华信息技术有限公司 一种指标异动的溯源及预测方法、装置
CN108173670A (zh) * 2016-12-07 2018-06-15 华为技术有限公司 检测网络的方法和装置
CN107480849A (zh) * 2017-06-28 2017-12-15 北京邮电大学 一种应用于电网的空间降维方法及装置
CN107579846A (zh) * 2017-08-23 2018-01-12 昆明理工大学 一种云计算故障数据检测方法及系统

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112882796A (zh) * 2021-02-25 2021-06-01 深信服科技股份有限公司 异常根因分析方法和装置，及存储介质
CN112882796B (zh) * 2021-02-25 2024-05-24 深信服科技股份有限公司 异常根因分析方法和装置，及存储介质
CN115150250A (zh) * 2021-03-31 2022-10-04 中国电信股份有限公司 基于因果学习的物联网异常根因定位方法以及装置
CN115150250B (zh) * 2021-03-31 2024-01-12 中国电信股份有限公司 基于因果学习的物联网异常根因定位方法以及装置
CN114338351A (zh) * 2021-12-31 2022-04-12 天翼物联科技有限公司 网络异常根因确定方法、装置、计算机设备及存储介质
CN114338351B (zh) * 2021-12-31 2024-01-12 天翼物联科技有限公司 网络异常根因确定方法、装置、计算机设备及存储介质
CN114760186A (zh) * 2022-03-23 2022-07-15 深信服科技股份有限公司 告警分析方法、装置、电子设备及存储介质
CN114760186B (zh) * 2022-03-23 2024-05-28 深信服科技股份有限公司 告警分析方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN111327435B (zh) 2022-07-05
CN111327435A (zh) 2020-06-23

Similar Documents

Publication Publication Date Title
WO2020119711A1 (zh) 一种根因定位方法、服务器和存储介质
CN111064614B (zh) 一种故障根因定位方法、装置、设备及存储介质
US20180212986A1 (en) Network attack detection method and device
CN110083475B (zh) 一种异常数据的检测方法及装置
CN109981326B (zh) 家庭宽带感知故障定位的方法及装置
US11368901B2 (en) Method for identifying a type of a wireless hotspot and a network device thereof
CN110784336A (zh) 基于物联网的多设备智能定时延时场景设置方法及系统
US11036490B2 (en) Proactive storage system-based software version analysis using machine learning techniques
WO2022057501A1 (zh) 异常终端的识别方法、分析装置及设备、存储介质
CN116756522B (zh) 概率预报方法、装置、存储介质及电子设备
CN110582091B (zh) 定位无线质量问题的方法和装置
CN111315026B (zh) 信道的选择方法、装置、网关及计算机可读存储介质
CN110995587A (zh) 一种路由不稳定事件源定位方法及装置
JP7389860B2 (ja) セキュリティ情報の処理方法、装置、電子機器、記憶媒体およびコンピュータプログラム
CN108133234B (zh) 基于稀疏子集选择算法的社区检测方法、装置及设备
CN113946566B (zh) Web系统指纹库的构建方法、装置和电子设备
EP4117236A1 (en) Anomaly detection across domains
CN112637880B (zh) 问题小区处理方法以及装置
CN111641535B (zh) 网络监控方法、装置、电子设备和介质
CN105307212B (zh) 一种基站流量数据的处理方法及基站
US11175911B2 (en) Reactive storage system-based software version analysis using machine learning techniques
CN112241767A (zh) 管网水压监测节点布置方法、装置、设备及存储介质
BR112021013973A2 (pt) Método e sistema para estimativa de saúde da planta
WO2022228062A1 (zh) 网络故障分析方法、装置、设备及存储介质
CN117150233B (zh) 一种电网异常数据治理方法、系统、设备及介质

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: 19894528

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19894528

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 06/12/2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19894528

Country of ref document: EP

Kind code of ref document: A1