Network fault locating method and device based on XDR data

Abstract

The application provides a network fault positioning method and device based on XDR data, wherein the method comprises the following steps: obtaining an XDR data set of a service to be positioned; extracting each XDR data vector in the XDR data set; comparing each XDR data vector with a case XDR data vector of each fault case in a preset fault service library to obtain a fault similarity between the service to be positioned and each fault case; and positioning the fault of the service to be positioned based on the fault similarity to determine a fault positioning result. The method provided by the application calculates the similarity between each XDR data vector and the case XDR data vector of each fault case in the preset fault service library to obtain the fault similarity between the service to be positioned and each fault case, and then determines the fault positioning result of the service to be positioned, without strict synchronization of the time of the XDR collection device of the core network element through which the service passes, fully utilizes the existing expert experience, and realizes accurate, convenient and efficient network fault positioning.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of network communication, in particular to a network fault positioning method and device based on XDR data. BACKGROUND

[0002] At present, in a communication network, after a service fault occurs, a commonly used fault positioning method is to form a full-link XDR of the fault service by synthesizing XDR data output by each interface DPI of a core network, arrange signaling messages with the same error code in time sequence, find a first initiating network element of the error code, and thus locate the fault first breaking point.

[0003] However, the above method requires that the XDR collection devices of each core network element are strictly time-aligned, and requires that the synthesis precision of the XDR is in the order of milliseconds. Thus, once clock jitter occurs between the XDR collection devices of each core network element, causing time inconsistency, there is a risk of failure in fault positioning, thereby increasing the complexity of troubleshooting, prolonging the troubleshooting time, affecting customer perception, and further increasing the operation and maintenance cost. SUMMARY

[0004] The present application provides a network fault positioning method and device based on XDR data to solve the defect that once the XDR collection devices of each network element are out of synchronization, the risk of troubleshooting failure and the difficulty of troubleshooting are increased.

[0005] The present application provides a network fault positioning method based on XDR data, comprising:

[0006] Obtaining an XDR data set of a service to be positioned;

[0007] Extracting each XDR data vector in the XDR data set;

[0008] Comparing each XDR data vector with a case XDR data vector of each fault case in a preset fault service library to obtain a fault similarity between the service to be positioned and each fault case;

[0009] Based on the fault similarity, positioning the fault of the service to be positioned to determine a fault positioning result of the service to be positioned.

[0010] According to the network fault positioning method based on XDR data provided by the present application, the comparison between each XDR data vector and a case XDR data vector of each fault case in a preset fault service library to obtain a fault similarity between the service to be positioned and each fault case comprises:

[0011] perform similarity calculation on each XDR data vector and corresponding case XDR data vector in the case atlas of each fault case, to obtain sub-data similarity between each XDR data vector and corresponding case XDR data vector in the case atlas;

[0012] determine the fault similarity based on the sub-data similarity.

[0013] According to the XDR data-based network fault positioning method provided by the application, the fault similarity is determined based on the sub-data similarity, and the fault positioning of the to-be-positioned service is performed based on the fault similarity, to determine the fault positioning result of the to-be-positioned service, which comprises:

[0014] fuse the sub-data similarity between each XDR data vector and corresponding case XDR data vector in the case atlas of each fault case, to perform average value calculation on the sub-data similarity, and obtain the fault similarity.

[0015] According to the XDR data-based network fault positioning method provided by the application, the fault positioning of the to-be-positioned service is performed based on the fault similarity, to determine the fault positioning result of the to-be-positioned service, which comprises:

[0016] select the maximum fault similarity from the fault similarity, compare the maximum fault similarity with a preset similarity threshold, and in the case that the maximum fault similarity is higher than the preset threshold, take the fault label of the fault case corresponding to the maximum fault similarity as the fault positioning result of the to-be-positioned service.

[0017] According to the XDR data-based network fault positioning method provided by the application, the fault positioning of the to-be-positioned service is performed based on the fault similarity, to determine the fault positioning result of the to-be-positioned service, which further comprises:

[0018] select the maximum fault similarity from the fault similarity, and in the case that the maximum fault similarity is not higher than the preset threshold, take the fault label of the fault case corresponding to the maximum fault similarity as the reference positioning result of the to-be-positioned service, to determine the fault positioning result based on the reference positioning result.

[0019] According to the XDR data-based network fault positioning method provided by the application, the obtaining step of the preset fault service library comprises:

[0020] obtain case XDR data of each fault case;

[0021] construct a case atlas based on the case XDR data corresponding to each interface in each fault case;

[0022] constructing a preset fault service library based on the case XDR data vectors of the case graphs of the respective fault cases and the fault labels of the respective fault cases;

[0023] The fault label comprises a fault type, a fault device type and a fault cause.

[0024] The application further provides a network fault positioning device based on XDR data, comprising:

[0025] An acquisition unit is configured to acquire an XDR data set of a service to be positioned;

[0026] An extraction unit is configured to extract respective XDR data vectors in the XDR data set;

[0027] A comparison unit is configured to compare the respective XDR data vectors with case XDR data vectors of respective fault cases in a preset fault service library to obtain fault similarities between the service to be positioned and the respective fault cases;

[0028] A fault positioning unit is configured to position a fault of the service to be positioned based on the fault similarities to determine a fault positioning result of the service to be positioned.

[0029] The application further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the network fault positioning method based on XDR data according to any one of the above-described embodiments when executing the program.

[0030] The application further provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program is executable on a processor to implement the network fault positioning method based on XDR data according to any one of the above-described embodiments.

[0031] The application further provides a computer program product comprising a computer program, wherein the computer program is executable on a processor to implement the network fault positioning method based on XDR data according to any one of the above-described embodiments.

[0032] The network fault positioning method and device based on XDR data provided by the application can calculate the similarities between respective XDR data vectors and case XDR data vectors of respective fault cases in a preset fault service library to obtain fault similarities between a service to be positioned and the respective fault cases, and then determine a fault positioning result of the service to be positioned, without the strict synchronization of the time of XDR collection devices of core network elements through which the service passes, and fully utilizing the existing expert experience, thus realizing accurate, convenient and efficient network fault positioning. BRIEF DESCRIPTION OF DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0034] Figure 1 is one of the flow schematic diagrams of the network fault locating method based on XDR data provided by the present application;

[0035] Figure 2 is a schematic diagram of the fault case provided by the present application;

[0036] Figure 3 is the second flow schematic diagram of the network fault locating method based on XDR data provided by the present application;

[0037] Figure 4 is a schematic diagram of the network element interface provided by the present application;

[0038] Figure 5 is a structural schematic diagram of the network fault locating device based on XDR data provided by the present application;

[0039] Figure 6 is a structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0040] In order to make the objects, technical solutions and advantages of the present application clearer, the following will combine the drawings in the present application to clearly and completely describe the technical solutions in the present application. Obviously, the described embodiments are some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without any creative effort belong to the protection scope of the present application.

[0041] Currently in a communication network, after a service fails, a method of synthesizing a full-link XDR of the failed service, and locating a fault initial breaking point according to a time sequence of signaling messages with the same error code is generally used. However, this requires that the time of each core network XDR collection device be strictly aligned, and the synthesis accuracy of the XDR be in the millisecond level. This not only puts strict requirements on the time synchronization of the core network XDR collection devices in the whole network, and on the synthesis accuracy of the XDR, but also makes the service fault location strongly dependent on the sending time of the signaling messages containing the error code. Once the XDR collection devices between the core network elements are not time-consistent, the fault location has a risk of failure, thereby increasing the troubleshooting complexity, prolonging the troubleshooting time, affecting the customer perception, and further causing the operation and maintenance cost to rise. In addition, the XDR can also be used to generate index parameters by statistics, and the location can be performed according to a set threshold. However, this method is difficult to accurately represent the individual service problem scenario.

[0042] To solve the above problems, the present application provides a network fault location method based on XDR data, to realize efficient, reliable and universal network fault location without strict synchronization of the time of the core network XDR collection devices through which the service passes. Figure 1 is one of the flowcharts of the network fault location method based on XDR data provided by the present application, as shown in Figure 1 The method comprises the following steps.

[0043] In step 110, the XDR data set of a service to be located is obtained.

[0044] Here, the service to be located refers to a communication service that fails in the communication network, i.e., a service to be located for fault location. In addition, the XDR (x Detail Record, detailed record of signaling and service generated for signaling monitoring platform and signaling application after processing based on full data) data set here refers to a collection of XDR data output by each interface in the core network corresponding to the service to be located, and a single interface outputs multiple XDR data items. The XDR data refers to the detailed record of signaling and service generated by the DPI (Deep Packet Inspection, deep packet inspection) collection for each signaling interface protocol standard in the communication network, which can include interface, process, process result, error code, additional error code, source device type, sink device type, and user number, IP, etc.

[0045] Thus, the XDR data can reflect the single-network element signaling interaction information under each interface in the communication network. When there is abnormal data in the XDR data, the fault type, fault network element type and fault reason can be located through the abnormal data, thereby facilitating the troubleshooting of the failed service to restore the normal communication of the service as soon as possible.

[0046] Specifically, the IMSI (International Mobile Subscriber Identity) or MSISDN (Mobile Station International ISDN Number) corresponding to the to-be-located service can be obtained by acquiring user complaints, complaint work order information, and failed service call slips in a test system. Then, the IMSI or MSISDN can be used as a keyword to collect single-user XDR data output by each interface DPI in the core network. Each interface single-user XDR data can be used as a data subset, and then all interface data subsets under the to-be-located service are collected to obtain the XDR data set of the full link of the to-be-located service, so as to comprehensively locate the fault of the to-be-located service.

[0047] In step 120, each XDR data vector in the XDR data set is extracted.

[0048] Specifically, the XDR data set can be composed of multiple XDR data subsets, and each single-interface XDR data can be used as a node to construct a single XDR data subset into a data graph. Thus, the XDR data vector of each XDR data in the XDR data set can be extracted by using a graph neural network. For example, the data graph constructed by each XDR data subset in the XDR data set can be input into the graph neural network, and the XDR data of each node in the data graph can be extracted by the graph neural network to obtain each XDR data vector.

[0049] It should be noted that, considering that the amount of data contained in the XDR data is large, and not all the contained data is helpful for fault location, for example, the user number and IP information in the XDR data are irrelevant to fault location, when the features of each XDR data in the XDR data set are encoded, the encoding vectors of invalid data can be removed, and the encoding vectors of data related to fault location can be strengthened. For example, the encoding vectors of interface, process, and error code data are strengthened. Thus, the influence of invalid vectors on fault location can be excluded, the accuracy of subsequent confidence calculation can be improved, and the accuracy and positioning efficiency of fault location can be further improved. In an embodiment, the graph neural network can be trained by machine learning combined with expert manual annotation in the training stage of the graph neural network, so that the graph neural network can remove invalid vectors and strengthen valid vectors.

[0050] In step 130, the fault similarity between the to-be-located service and each fault case is obtained by comparing each XDR data vector with the case XDR data vector of each fault case in the preset fault service library.

[0051] Here, the preset fault service library refers to a pre-constructed knowledge base containing multiple fault cases. The preset fault service library can contain fault cases of multiple fault types, multiple fault device types, and multiple fault causes. It should be noted that for a single fault case, only one case graph is annotated for fault positioning, i.e., there is only one service fault for a single fault case. It can be understood that the fault cases in the preset fault service library can be added and modified to make the fault cases in the preset fault service library more rich and comprehensive, thereby realizing rapid conversion of expert experience.

[0052] In addition, the fault similarity can reflect the similarity degree between the full-link XDR data of the to-be-positioned service and the full-link XDR data of the fault case, and can further reflect the similarity degree between the fault positioning of the to-be-positioned service and the fault case. It can be understood that the higher the fault similarity between the to-be-positioned service and the fault case, the higher the similarity between the fault of the to-be-positioned service and the fault of the fault case; otherwise, the lower the similarity between the fault of the to-be-positioned service and the fault of the fault case.

[0053] Specifically, the cosine similarity between each XDR data vector of the to-be-positioned service and the corresponding case XDR data vector in each fault case in the preset fault service library can be calculated to obtain the fault similarity between the to-be-positioned service and each fault case in the preset fault service library.

[0054] It should be noted that the method provided by the embodiment of the present application extracts the vector of the full-link XDR data of the to-be-positioned service, calculates the similarity with the case XDR data vector of each fault case in the pre-constructed preset fault service library, compares the to-be-positioned service with the fault cases in the case knowledge base, and thereby locates the fault first break point / abnormal point. Thus, compared with the prior art, the method for network fault positioning based on XDR data provided by the embodiment of the present application does not require that the signaling process be completely aligned in time, i.e., it does not require strict synchronization of the time of the XDR collection device of the core network element through which the service passes. Moreover, it does not require manual preparation of fault identification rules, greatly saving manpower and resources required for troubleshooting. In addition, the method can make full use of existing expert experience, reserve known fault cases in the case knowledge base, thereby realizing rapid conversion of expert experience, and thereby realizing accurate, convenient and efficient network fault positioning, greatly improving the user experience.

[0055] Step 140, based on the fault similarity, positioning the fault of the to-be-positioned service to determine the fault positioning result of the to-be-positioned service.

[0056] Specifically, the fault positioning of the to-be-positioned service can be performed according to the fault similarity, and the fault positioning result of the to-be-positioned service is determined. For example, the fault case corresponding to the lowest fault similarity can be excluded, and the fault case corresponding to the highest fault similarity can be taken as the fault positioning result of the to-be-positioned service, so that the analysis process of the fault positioning result is saved, and the fault positioning of the to-be-positioned service is quickly and accurately performed.

[0057] The method provided by the embodiment of the application can calculate the similarity between each XDR data vector and the case XDR data vector of each fault case in the preset fault service library, obtain the fault similarity between the to-be-positioned service and each fault case, and then determine the fault positioning result of the to-be-positioned service, without strict synchronization of the time of the XDR collection device of the core network element passed by the service, fully utilizes the existing expert experience, and accurately, conveniently and efficiently realizes the network fault positioning.

[0058] Based on any of the above embodiments, step 130 comprises:

[0059] Based on the similarity calculation of each XDR data vector and the corresponding case XDR data vector in the case graph of each fault case, the sub-data similarity between each XDR data vector and the corresponding case XDR data vector in the case graph is obtained.

[0060] Based on the sub-data similarity, the fault similarity is determined.

[0061] Here, the XDR data set can be composed of a plurality of XDR data subsets, and a single XDR data subset corresponds to an interface in the full link of the to-be-positioned service. In addition, the case graph here can refer to a data graph constructed by taking each single-interface XDR data in the XDR data set as a node.

[0062] Specifically, in order to further improve the positioning efficiency and positioning accuracy of fault positioning, when performing similarity calculation, the XDR data vector in each XDR data subset of each interface in the to-be-positioned service can be calculated with the corresponding case XDR data vector in the case graph of each fault case to obtain the sub-data similarity between each XDR data vector and the corresponding case XDR data vector in the case graph. For example, the to-be-positioned service includes interfaces A, B and C. The A interface includes XDR data vectors 、 、 , the B interface includes XDR data vectors 、 、 , and the C interface includes XDR data vectors 、 、 The default fault service database contains fault cases 1, 2, and 3. Assume each fault case, like the service to be located, includes interfaces A, B, and C, and each interface also contains a corresponding number of XDR data vectors. Therefore, when comparing the service to be located with fault case 1, the XDR data vectors contained in interface A of the service to be located can be used to determine the fault cases. , , The XDR data vectors contained in interface A of fault case 1 are respectively related to the XDR data vectors contained in the fault case 1. Calculate the similarity, that is and Perform similarity calculation. and Similarity calculations are performed to obtain the sub-data similarity between each XDR data vector under interface A and the corresponding case XDR data vector in the case graph of the corresponding interface.

[0063] Similarly, the XDR data vector contained in interface B of the service to be located is compared with the XDR data vector contained in interface B of fault case 1. Similarity calculations are performed until the similarity of each XDR data vector in the XDR data subset of all interfaces in the service to be located is calculated with the corresponding case XDR data vector in the case map of the fault case. Then, the service to be located is compared with each fault case using the comparison method between the service to be located and a single fault case.

[0064] Finally, the fault similarity between the service to be located and the fault case can be obtained by fusing the sub-data similarity between the XDR data of the service to be located and the XDR data vector of the corresponding case.

[0065] Understandably, similarity calculations are performed on the XDR data vectors of the corresponding interfaces in the fault cases to obtain the similarity between the XDR data of the service to be located and the XDR data of the corresponding interfaces in each fault case, thereby achieving efficient and accurate comparison between the service to be located and each fault case.

[0066] Based on any of the above embodiments, determining the fault similarity based on the sub-data similarity includes:

[0067] The fault similarity is obtained by merging the sub-data similarity between each XDR data vector and the corresponding case XDR data vector in the case map of each fault case, and then calculating the average value of the sub-data similarity.

[0068] Specifically, Table 1 shows the sub-data similarity between each XDR data vector and the corresponding case XDR data vector in the case atlas, as shown in Table 1 below:

[0069] Table 1

[0070]

[0071] As can be seen from Table 1, only the similarity of XDR sub-vector 1 and case 1-XDR graph sub-vector 1 is taken as the sub-data similarity of XDR sub-vector 1 and case 1. Thus, the confidence of XDR sub-vector 1 in the to-be-located service and case 1 is 0.95. In the same way, the sub-data similarity of each XDR sub-vector in the to-be-located service and the corresponding case sub-vector in case 1 is obtained, and then the overall similarity of the to-be-located service and case 1 is obtained. Further, in the same way, the fault similarity between the to-be-located service and each fault case is calculated, and accurate comparison between the to-be-located service and the fault case is realized, so as to improve the accuracy of network fault positioning.

[0072] Based on any of the above embodiments, step 140 comprises:

[0073] The maximum fault similarity is selected from the fault similarities, and the maximum fault similarity is compared with a preset similarity threshold. In the case where the maximum fault similarity is higher than the preset threshold, the fault label of the fault case corresponding to the maximum fault similarity is taken as the fault positioning result of the to-be-located service.

[0074] Here, the preset similarity threshold can reflect the basis for determining whether the fault causes of the to-be-located service and the fault case are the same. When the fault similarity is higher than the similarity threshold, it indicates that the fault causes of the to-be-located service and the fault case are highly the same. When the fault similarity is lower than the similarity threshold, it indicates that there is a difference between the fault causes of the to-be-located service and the fault case.

[0075] Specifically, the maximum fault similarity can be selected from the fault similarities of the to-be-located service and each fault case, and the maximum fault similarity is compared with the preset similarity threshold. For example, the maximum fault similarity is 90%, and the preset similarity threshold is 85%. Then, the fault label of fault case 1 corresponding to the maximum fault similarity can be taken as the fault positioning result of the to-be-located service, that is, it can be considered that the fault network element type, fault type, and fault cause of the XDR data of the to-be-located service are consistent with those of fault case 1, so as to realize network fault positioning of the to-be-located service, and obtain the fault network element type, fault type, fault cause, and identifier of the fault network element.

[0076] The method provided by the embodiment of the application obtains the fault similarity between the to-be-positioned service and the fault case by calculating the similarity between XDR data vectors, without implementing complete alignment of signaling processes in time sequence, avoiding the risk of inconsistent time of XDR collection devices between core network elements, positioning failure of faults, improving the reliability of network fault positioning based on XDR data, without spending a large amount of manpower and time cost for fault analysis, improving the accuracy and efficiency of network fault positioning, and further improving the user experience.

[0077] It should be noted that the higher the fault similarity is, the closer the fault causes of the to-be-positioned service and the fault case are, and step 140 further includes:

[0078] Selecting a maximum fault similarity from the fault similarities, and in a case where the maximum fault similarity is not higher than a preset threshold, taking the fault label of the fault case corresponding to the maximum fault similarity as a reference positioning result of the to-be-positioned service, to determine the fault positioning result based on the reference positioning result.

[0079] Specifically, the maximum fault similarity is selected from the fault similarities of the to-be-positioned service and each fault case, and in a case where the maximum fault similarity is not higher than a preset threshold, it is indicated that the fault of the to-be-positioned service is not completely the same as the fault cause of the fault case corresponding to the maximum fault similarity. However, if the maximum fault similarity is lower than the preset threshold but higher than a minimum threshold, the fault label of the corresponding fault case can be taken as the reference positioning result of the to-be-positioned service. For example, when the maximum fault similarity is lower than 85% but higher than 75%, the fault label of the corresponding fault case can be taken as the reference positioning result of the to-be-positioned service. Thus, the fault positioning of the to-be-positioned service can be performed based on the reference positioning result, so as to greatly save the time required for fault positioning and improve the efficiency of fault positioning.

[0080] It should be noted that when the maximum fault similarity is lower than the minimum threshold, it can be indicated that there is no case matching the to-be-positioned service in the preset fault knowledge base. Thus, after obtaining the to-be-positioned service, the to-be-positioned service can be updated to the preset fault knowledge base, so as to enrich the preset fault knowledge base and accumulate experience for subsequent fault positioning.

[0081] Based on any of the above embodiments, the obtaining step of the preset fault service library includes:

[0082] Obtaining case XDR data of each fault case;

[0083] Based on the case XDR data corresponding to each interface in each fault case, a case graph is constructed.

[0084] a case XDR data vector based on a case graph of each fault case, and a fault label of each fault case, to construct a preset fault service library;

[0085] The fault label includes a fault type, a fault device type, and a fault cause.

[0086] The case XDR data here can include interface, process, process result, error code, additional error code, source device type, and sink device type information. Specifically, the case XDR data output by each interface of the core network can be collected by taking the IMSI or MSISDN corresponding to the fault case as a key. For a single fault case, the case XDR data corresponding to each interface in the fault case can be respectively constructed to obtain a case graph corresponding to each interface. The case graphs of the remaining fault cases are constructed in this way. For example, if fault case 1 includes n interfaces, the number of case graphs of fault case 1 is n. Table 2 shows the case graphs of two fault cases, as shown in the following table 2:

[0087] Table 2

[0088]

[0089] As shown in Table 2, XDR graph 1 represents all case XDR data output by interfaces 1 to n in fault case 1, i.e., sub-vector 1, …, sub-vector n. XDR graph 2 represents all case XDR data output by interfaces 1 to m in fault case 2, i.e., sub-vector 1, …, sub-vector m.

[0090] Further, the case XDR data vector in the case graph in which a service fault exists in each fault case can be fault-labeled to obtain a preset fault service library containing fault labels. It should be noted that a single service flow can include multiple case graphs, and in the multiple case graphs of a single service flow, there are successful case graphs and fault case graphs. The successful case graphs can be obtained by training or fixed by a model, and the fault case graphs can be obtained by labeling or comparison. That is, a fault case is a fault positioning label for a case graph. Figure 2 is a schematic diagram of a fault case provided by the present application, as Figure 2 shown, the fault case includes an XDR graph and a fault label, and the fault label includes a fault type, a fault device type, and a fault cause.

[0091] The method provided by the embodiments of the present application can collect and arrange existing expert knowledge by constructing a preset case knowledge base, so that network fault positioning can be quickly and accurately realized when fault positioning is performed.

[0092] Based on any of the above embodiments,Figure 3 This is the second flowchart of the network fault location method based on XDR data provided by the present invention, as shown below. Figure 3 As shown, the method includes:

[0093] S0, Training the XDR vector for the business;

[0094] Specifically, during the training phase, a neural network that performs vector encoding on the XDR data can be trained to eliminate invalid vectors, i.e., those that do not affect the fault location XDR fields, such as user numbers and IP addresses, while valid vectors, such as interfaces, processes, and error codes, can be strengthened. A successful business process model or graph can then be established; this business process is obtained by combining multiple cases according to the established business process.

[0095] S1. Establish a case knowledge base;

[0096] Specifically, a case knowledge base is constructed by building XDR data and fault annotations for known business cases. For a single business case, an XDR map is constructed from the XDR data output by each network element interface. Figure 4 This is a schematic diagram of the network element interface provided by the present invention, as shown below. Figure 4 As shown in the diagram, the components include: GPRS control plane, BTS base station transceiver station, BSC / PCU (Base Station Controller / Packet Control Unit), HSS (Home Subscriber Server), PCRF (Policy and Charging Rule Function), UMTS (Universal Mobile Telecommunications System), NodeB (UTRAN base station node), MME (Mobility Management Entity), Operator Service, RNC (Radio Network Controller), E-UTRAN (Evolved UMTS Terrestrial Radio Access Network), PDN GW (Packet Data Network Gateway), and Serving GW (Serving Gateway). These gateways are interconnected via network element interfaces, such as...Figure 4 S1-MME, S2b, S1-U, etc. in FIG. 1. Wherein, the fields contained in the interface corresponding to each type are also different.

[0097] S2, generating a fault service XDR vector;

[0098] Specifically, by obtaining the XDR data of the full link of the fault service, and then encoding the XDR data, the fault service XDR vector is obtained.

[0099] S3, calculating a confidence level;

[0100] Specifically, based on the comparison between each fault service XDR data vector and the corresponding case XDR data vector in the preset fault service library, the fault similarity between the fault service and each fault case is obtained. In detail, for example, each fault service XDR data vector can be divided according to the network element interface outputted by it. The XDR data subset corresponding to each interface is obtained. Based on the similarity calculation between each fault service XDR data vector and the corresponding case XDR data vector in the case graph of each fault case, the sub-data similarity between each fault service XDR data vector and the corresponding case XDR data vector in the case graph is obtained. The arithmetic mean of the sub-data similarity of each fault case is calculated to obtain the fault similarity between the fault service and each fault case.

[0101] S4, locating a fault;

[0102] Specifically, the maximum fault similarity is selected from each fault similarity, and the maximum fault similarity is compared with the preset similarity threshold. In the case where the maximum fault similarity is higher than the preset threshold, the fault label of the fault case corresponding to the maximum fault similarity is taken as the fault positioning result of the to-be-positioned service. In the case where the maximum fault similarity is not higher than the preset threshold, the fault label of the fault case corresponding to the maximum fault similarity is taken as the reference positioning result of the to-be-positioned service, so as to determine the fault positioning result based on the reference positioning result.

[0103] Based on any one of the above embodiments, Figure 5 is a structural schematic diagram of a network fault positioning device based on XDR data provided by the present application, as Figure 5 shown, the device comprises:

[0104] An acquisition unit 510 acquires an XDR data set of a to-be-positioned service;

[0105] An extraction unit 520 extracts each XDR data vector in the XDR data set.

[0106] The comparison unit 530 compares each XDR data vector with a case XDR data vector of each fault case in the preset fault service library to obtain a fault similarity between the to-be-located service and each fault case;

[0107] The fault positioning unit 540 performs fault positioning on the to-be-located service based on the fault similarity to determine a fault positioning result of the to-be-located service.

[0108] The device provided in the embodiment of the application performs similarity calculation on each XDR data vector and a case XDR data vector of each fault case in the preset fault service library to obtain a fault similarity between the to-be-located service and each fault case, and then determines a fault positioning result of the to-be-located service, without strict synchronization of the time of the XDR collection device of the core network element passed by the service, fully utilizes existing expert experience, and realizes accurate, convenient and efficient network fault positioning.

[0109] Based on any of the above embodiments, the comparison unit is specifically configured to:

[0110] The comparison unit is specifically configured to:

[0111] The comparison unit is specifically configured to:

[0112] Based on any of the above embodiments, the comparison unit is specifically configured to:

[0113] The comparison unit is specifically configured to:

[0114] Based on any of the above embodiments, the fault positioning unit is specifically configured to:

[0115] The fault positioning unit is specifically configured to:

[0116] Based on any of the above embodiments, the fault positioning unit is specifically configured to:

[0117] The maximum fault similarity is selected from the fault similarities. If the maximum fault similarity is not higher than a preset threshold, the fault label of the fault case corresponding to the maximum fault similarity is used as the reference positioning result of the service to be located, so as to determine the fault positioning result based on the reference positioning result.

[0118] Based on any of the above embodiments, the step of obtaining the preset fault service database includes:

[0119] Obtain the XDR data for each failure case;

[0120] Based on the case XDR data corresponding to each interface in each of the aforementioned fault cases, a case map is constructed.

[0121] Based on the case XDR data vector of the case map of each fault case and the fault label of each fault case, a preset fault service library is constructed.

[0122] The fault label includes the fault type, the type of faulty equipment, and the cause of the fault.

[0123] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 610, a communications interface 620, a memory 630, and a communication bus 640. The processor 610, communications interface 620, and memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute a network fault location method based on XDR data. This method includes: acquiring an XDR dataset of the service to be located; extracting each XDR data vector from the XDR dataset; comparing each XDR data vector with case XDR data vectors of each fault case in a preset fault service database to obtain the fault similarity between the service to be located and each fault case; and based on the fault similarity, performing fault location on the service to be located to determine the fault location result of the service to be located.

[0124] In addition, the logic instructions in the memory 630 described above can be implemented in the form of software functional units and sold or used as independent products, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, etc.

[0125] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program can be executed by a processor to enable a computer to execute the network fault locating method based on XDR data provided by the above-mentioned methods, the method comprising: obtaining an XDR data set of a service to be located; extracting each XDR data vector in the XDR data set; comparing each XDR data vector with a case XDR data vector of each fault case in a preset fault service library to obtain a fault similarity between the service to be located and each fault case; and based on the fault similarity, locating a fault of the service to be located to determine a fault locating result of the service to be located.

[0126] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program can be executed by a processor to enable a computer to execute the network fault locating method based on XDR data provided by the above-mentioned methods, the method comprising: obtaining an XDR data set of a service to be located; extracting each XDR data vector in the XDR data set; comparing each XDR data vector with a case XDR data vector of each fault case in a preset fault service library to obtain a fault similarity between the service to be located and each fault case; and based on the fault similarity, locating a fault of the service to be located to determine a fault locating result of the service to be located.

[0127] The device embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purposes of the embodiments according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0128] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0129] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A network fault locating method based on XDR data, characterized in that, The method comprises the following steps: An XDR data set of a service to be positioned is acquired, wherein the XDR data set is obtained by collecting data subsets of all interfaces under the service to be positioned based on single-user XDR data of each interface as a data subset; Each XDR data vector in the XDR data set is extracted, comprising: taking each single-interface XDR data as a node, constructing the data subsets of the interfaces into a data graph, inputting the data graph constructed from each data subset in the XDR data set into a graph neural network for feature extraction to obtain the XDR data vector; The fault similarity between the service to be positioned and each fault case is obtained by comparing the XDR data vector with the case XDR data vector of each fault case in a preset fault service library; The fault of the service to be positioned is located based on the fault similarity, and the fault positioning result of the service to be positioned is determined; The fault similarity between the service to be positioned and each fault case is obtained by comparing the XDR data vector with the case XDR data vector of each fault case in a preset fault service library, comprising: The sub-data similarity between the XDR data vector and the corresponding case XDR data vector in the case graph of each fault case is obtained by performing similarity calculation based on the XDR data vector and the corresponding case XDR data vector in the case graph of each fault case; The fault similarity is determined based on the sub-data similarity.

2. The method for network fault locating based on XDR data according to claim 1, characterized in that, The fault similarity is determined based on the sub-data similarity, comprising: The sub-data similarity between the XDR data vector and the corresponding case XDR data vector in the case graph of each fault case is fused to perform average value calculation of the sub-data similarity, and the fault similarity is obtained.

3. The method for network fault localization based on XDR data according to claim 1, wherein, The fault of the service to be positioned is located based on the fault similarity, and the fault positioning result of the service to be positioned is determined, comprising: The maximum fault similarity is selected from the fault similarity, and the maximum fault similarity is compared with a preset similarity threshold; if the maximum fault similarity is higher than the preset threshold, the fault label of the fault case corresponding to the maximum fault similarity is taken as the fault positioning result of the service to be positioned.

4. The method for network fault locating based on XDR data according to claim 3, characterized in that, The fault of the service to be positioned is located based on the fault similarity, and the fault positioning result of the service to be positioned is determined, further comprising: The maximum fault similarity is selected from the fault similarity, and if the maximum fault similarity is not higher than the preset threshold, the fault label of the fault case corresponding to the maximum fault similarity is taken as a reference positioning result of the service to be positioned, and the fault positioning result is determined based on the reference positioning result.

5. The method for network fault locating based on XDR data according to any one of claims 1 to 4, characterized in that, The acquisition step of the preset fault service library comprises: Case XDR data of each fault case is acquired; A case graph is constructed based on the case XDR data corresponding to each interface in each fault case; A preset fault service library is constructed based on the case XDR data vector of the case graph of each fault case and the fault label of each fault case. The fault label includes a fault type, a fault device type, and a fault cause.

6. A network fault locating apparatus based on XDR data, characterized by, Comprise: An acquisition unit acquires an XDR dataset of a service to be positioned, the XDR dataset being obtained by collecting data subsets of all interfaces under the service to be positioned based on single-user XDR data of each interface as a data subset; An extraction unit extracts each XDR data vector in the XDR dataset, specifically for taking each single-interface XDR data as a node, constructing the data subsets of the interfaces into a data graph, inputting the data graph constructed by each data subset in the XDR dataset into a graph neural network for feature extraction, and obtaining the XDR data vectors; A comparison unit compares each XDR data vector with a case XDR data vector of each fault case in a preset fault service library to obtain a fault similarity between the service to be positioned and the fault cases; A fault positioning unit positions a fault of the service to be positioned based on the fault similarity and determines a fault positioning result of the service to be positioned. The comparison unit is specifically configured to: Calculate a similarity between each XDR data vector and a corresponding case XDR data vector in a case graph of each fault case based on the XDR data vectors and the corresponding case XDR data vectors in the case graph of each fault case to obtain a sub-data similarity between each XDR data vector and the corresponding case XDR data vector in the case graph of each fault case; Determine the fault similarity based on the sub-data similarity.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the network fault positioning method based on XDR data according to any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the network fault positioning method based on XDR data according to any one of claims 1 to 5.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the network fault positioning method based on XDR data according to any one of claims 1 to 5.

Images