Time synchronization fault delimiting method and device based on machine learning and storage medium

Abstract

The embodiment of the invention provides a time synchronization fault delimiting method based on machine learning, time synchronization fault delimiting equipment based on machine learning and a storage medium, and belongs to the technical field of communication. The time synchronization fault delimiting method based on machine learning comprises the following steps: determining a monitoring time synchronization link; and performing fault analysis on the monitoring time synchronization link through a preset machine learning algorithm, and determining a corresponding fault network element and / or a fault module to realize fault delimitation. According to the invention, for the time synchronization link which needs to have a monitoring requirement, fault analysis is carried out on the time synchronization link through the preset machine learning algorithm, network element level delimitation, module level delimitation and network element-module secondary delimitation are realized, and the reliability of fault delimitation in a communication system is improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a time synchronization fault demarcation method based on machine learning, a time synchronization fault demarcation device based on machine learning, and a storage medium. Background Art

[0002] With the rapid development of mobile communication network technology, the network scale is getting larger and more complex. In the time synchronization network, there are master and slave clock nodes. The master and slave nodes send synchronization data frames, record and add timestamps to the data frames. The nodes calculate the deviation and transmission delay between the master and slave clocks based on the acquired data frames and timestamp information, thereby performing time synchronization. Once a node in the synchronization link has a time error, such as a frequency layer failure or a fault time error generated by the device during time transfer, it will cause the synchronization time error of other nodes after the faulty node on the synchronization link to become larger and larger, ultimately affecting the normal operation of the services in the system.

[0003] In communication systems, time synchronization plays a vital role. Clock synchronization failure may lead to failure of communication connection between communication systems, which will affect the operation and business of the system. Therefore, high-precision time synchronization technology is needed to meet the needs of communication networks.

[0004] In the prior art, faults are generally detected manually, which requires printing out the synchronization timestamps and related configuration information of the network element equipment, and then observing and analyzing the fluctuations and changes of hundreds of timestamps to determine the fault location. Usually, fault detection of synchronous links is manual point-by-point detection, which not only requires manual search for the network element where the fault occurs from the network topology, but also requires each business expert to locate the module to which the fault belongs from the faulty network element and complete the fault repair. However, the number of nodes involved in synchronization in the network topology is huge, and manual positioning faces the problems of complex network topology and huge amount of data to be analyzed. Manual fault location is prone to deviations, so how to improve the reliability of fault demarcation in communication systems has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The embodiments of the present invention provide a time synchronization fault demarcation method based on machine learning, a time synchronization fault demarcation device based on machine learning, and a storage medium, aiming to improve the reliability of fault demarcation in a communication system.

[0006] In a first aspect, an embodiment of the present invention provides a time synchronization fault demarcation method based on machine learning, comprising:

[0007] Determine monitoring time synchronization link;

[0008] The monitoring time synchronization link is subjected to fault analysis by a preset machine learning algorithm, and the corresponding faulty network element and / or faulty module is determined to achieve fault demarcation.

[0009] In a second aspect, an embodiment of the present invention further provides a time synchronization fault demarcation device based on machine learning, wherein the time synchronization fault demarcation device based on machine learning comprises a processor, a memory, a computer program stored on the memory and executable by the processor, and a data bus for realizing connection and communication between the processor and the memory, wherein when the computer program is executed by the processor, any one of the time synchronization fault demarcation methods based on machine learning provided in the specification of the present invention is realized.

[0010] In a third aspect, an embodiment of the present invention further provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement any one of the time synchronization fault delimitation methods based on machine learning provided in the specification of the present invention.

[0011] The embodiments of the present invention provide a method for time synchronization fault delimitation based on machine learning, a device for time synchronization fault delimitation based on machine learning, and a storage medium. The embodiments of the present invention include determining a monitored time synchronization link; performing fault analysis on the monitored time synchronization link through a preset machine learning algorithm, and determining the corresponding faulty network element and / or faulty module to achieve fault delimitation. The present invention performs fault analysis on a time synchronization link that needs to be monitored through a preset machine learning algorithm, achieves network element-level delimitation, module-level delimitation, and network element-module two-level delimitation, and improves the reliability of fault delimitation in the communication system. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0013] Figure 1 A flowchart of a time synchronization fault demarcation method based on machine learning provided in the first embodiment of the present invention;

[0014] Figure 2 A flowchart of a method for time synchronization fault demarcation based on machine learning provided in the second embodiment of the present invention;

[0015] Figure 3A schematic diagram of a faulty network element delimiting method for time synchronization fault delimitation based on machine learning provided in an embodiment of the present invention;

[0016] Figure 4 A flowchart of a method for time synchronization fault demarcation based on machine learning provided in the third embodiment of the present invention;

[0017] Figure 5 A schematic block diagram of the structure of a time synchronization fault demarcation device based on machine learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0020] It should be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0021] The embodiment of the present invention provides a method for delimiting a time synchronization fault based on machine learning, a device for delimiting a time synchronization fault based on machine learning, and a storage medium. The method for delimiting a time synchronization fault based on machine learning can be applied to a mobile terminal, which can be an electronic device such as a mobile phone, a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device.

[0022] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0023] The following will be combined Figure 1 The scenario in the embodiment of the present invention provides a detailed introduction to the method for demarcating time synchronization faults based on machine learning. It should be noted that Figure 1The scenario is only used to explain the time synchronization fault demarcation method based on machine learning provided in an embodiment of the present invention, but does not constitute a limitation on the application scenario of the time synchronization fault demarcation method based on machine learning provided in an embodiment of the present invention.

[0024] Please refer to Figure 1 , Figure 1 A flowchart of a time synchronization fault demarcation method based on machine learning is provided for the first embodiment of the present invention.

[0025] like Figure 1 As shown, the time synchronization fault demarcation method based on machine learning includes steps S10 to S30.

[0026] Step S10, determining a monitoring time synchronization link;

[0027] In one embodiment, different devices correspond to different communication protocols. The current device takes the SPN (Slicing Packet Network) network as an example to briefly explain the clock synchronization principle of the 1588 protocol used by the SPN.

[0028] Step 1: The master clock sends a Sync message to the slave clock, carrying its own clock t1;

[0029] Step 2: Receive the Sync message from the slave clock and mark it with its own clock t2 at that moment;

[0030] Step 3: The slave clock sends a Delay_Req message to the master clock, carrying its own clock t3 at that moment;

[0031] Step 4: The master clock receives the Delay_Req message and marks its own clock t4 at that time;

[0032] Step 5: The master clock sends a Delay_Resp message to the slave clock, carrying the clock t4;

[0033] Assuming that the difference between the slave clock and the master clock is offset, the transmission delay from the master clock to the slave clock is t_ms, and the transmission delay from the slave clock to the master clock is t_sm, then there are the following two equations:

[0034] t2-t1=t_ms+offset;

[0035] t4-t3=t_sm-offset;

[0036] Assuming that the two-way delay is the same, that is, delay = t_ms = t_sm, the following results can be calculated:

[0037] offset=[(t2-t1)-(t4-t3)] / 2, and finally correct the time according to the offset.

[0038] In another embodiment, taking the 802.1AS protocol used in a TSN (Time-Sensitive Network) network as an example, the clock synchronization principle of the 802.1AS protocol is briefly described.

[0039] Step 1: The master clock sends a Sync message to the slave clock through the intermediate node TC, marking the timestamp as t1;

[0040] Step 2: After receiving the message from the master clock, the TC node sends a Pdelay_Req request message to the master clock. Similar to 1588, the transmission delay ΔTA is calculated based on the returned response message.

[0041] Step 3: The TC node adds its own forwarding processing time ΔTms and transmission delay ΔTA to the synchronization data message and sends it to the slave clock node;

[0042] Step 4: The slave clock sends a Pdelay_Req request message to the TC node to calculate the transmission delay ΔTB;

[0043] Step 5: The slave clock receives the message forwarded by the TC node to the master clock and marks the timestamp as t2;

[0044] According to the transmission delay and t1 and t2, it can be concluded that: offset = t2-t1-ΔTA-ΔTms-ΔTB.

[0045] Specifically, a time synchronization link that needs to be analyzed is manually selected on the network topology selection interface to trigger fault delimitation analysis. According to the selected link, the relevant time synchronization configuration instance is indexed, and an analysis request for the configuration instance is sent to the corresponding network element, triggering the network element to perform fault network element-level reasoning.

[0046] Step S20: Perform fault analysis on the monitoring time synchronization link through a preset machine learning algorithm to determine the corresponding faulty network element and / or faulty module to achieve fault demarcation.

[0047] In a specific embodiment, different fault delimitation instructions correspond to different levels of delimitation. For example, network element-level delimitation corresponds to network element-level delimitation instructions, module-level delimitation corresponds to module-level delimitation instructions, and network element-module delimitation corresponds to network element-module delimitation instructions. The above-mentioned fault delimitation instructions can be manually issued by the user, or can be intelligently issued by the target device. For example, when a time synchronization anomaly is detected, a fault delimitation instruction is automatically issued to determine the faulty network element and / or faulty module. The time synchronization fault delimitation method based on machine learning provided in this embodiment can be applied to different scenarios, such as the delimitation of faulty network elements, the delimitation of faulty modules, and the secondary delimitation of network elements and modules.

[0048] The steps for fault delimitation of network elements and modules are similar. Here, we take the faulty network element as an example. The network element collects the message timestamp information from the underlying latch of the instance, and forms important features through several steps of data processing. The extracted important features are sent to the machine learning algorithm to infer whether the timestamp information is abnormal, and the inference results are summarized. After obtaining the inference results of each network element in the selected link, among all the network elements with faulty inference results, the network element at the upstream of the synchronization link is determined as the network element to which the fault belongs.

[0049] For the NE-module secondary delimitation, it is necessary to locate the faulty NE through the above method, and then send the corresponding configuration instance analysis request to the faulty NE to trigger the faulty NE to perform module-level reasoning of the fault. The NE collects and integrates synchronization timestamp information, configuration information, and related module diagnostic information, and extracts important features through data processing. The extracted important features are sent to the machine learning algorithm to infer the faulty module number, and the reasoning results are pushed to the user end.

[0050] The embodiments of the present invention provide a method, device and storage medium for time synchronization fault delimitation based on machine learning, which belongs to the field of communication technology. The method for time synchronization fault delimitation based on machine learning includes: determining a monitored time synchronization link; performing fault analysis on the monitored time synchronization link through a preset machine learning algorithm, and determining the corresponding faulty network element and / or faulty module to achieve fault delimitation. The present invention performs fault analysis on the time synchronization link for which monitoring requirements are required through a preset machine learning algorithm, and achieves network element-level delimitation, module-level delimitation and network element-module two-level delimitation, thereby improving the reliability of fault delimitation in the communication system.

[0051] Please refer to Figure 2 , Figure 2 A flowchart of a time synchronization fault demarcation method based on machine learning is provided for the second embodiment of the present invention.

[0052] based on Figure 1 The embodiment shown, this embodiment Figure 2As shown, step S20 includes step S201 to step S202.

[0053] Step S201: upon receiving a network element level delimitation instruction and / or a module level delimitation instruction, obtaining timestamp information of the monitoring time synchronization link;

[0054] Specifically, according to the selected synchronization link, the 1588 configuration instance of the network element node is indexed, including but not limited to the Master and Slave configuration instances; and an analysis request for the configuration instance is formed and sent to the corresponding network element to trigger network element-level reasoning. The indexing method includes but is not limited to searching the 1588 configuration instance on the network element through the link port.

[0055] After receiving the request, the network element configures the underlying high-speed chip to latch the message timestamp (T1~T4) of the corresponding 1588 instance; the data acquisition module reads the latched timestamp and caches multiple sampling points to form timestamp information, and performs data processing, data cleaning and feature engineering to extract important features that can characterize the characteristics of timestamp information changes. The steps of feature engineering include but are not limited to differential processing, time series statistical feature extraction, time domain / frequency domain feature extraction of time series, feature normalization, etc. The categories of extracted features include but are not limited to the following categories: 1. Time domain features, such as differential mean and differential variance; 2. Statistical features, such as mean, variance, peak, etc.; 3. Frequency domain features, such as Fourier transform or wavelet transform of time series.

[0056] Step S202: perform data processing on the timestamp information to generate network element timestamp features and / or module timestamp features, and determine the faulty network element and / or the faulty module based on the preset machine learning algorithm, the network element timestamp features and / or the module timestamp features.

[0057] In a specific embodiment, the extracted time series features are sent to a lightweight machine learning algorithm for reasoning, and the reasoning result is whether the timestamp information (T1-T4) is abnormal, and the result is pushed to the management background. The machine learning algorithm is a supervised learning algorithm, including but not limited to tree model algorithms (such as random forests, gradient boosting trees), support vector machines, artificial neural networks, naive Bayes and other algorithms; one or more algorithms can be used for prediction to obtain the prediction results of each algorithm, and then combined with the weight coefficients, the final prediction results are obtained by weighting. The reasoning results of each network element are combined with the time synchronization link topology to determine the network element to which the fault belongs.

[0058] like Figure 3 As shown, Figure 3A schematic diagram of a faulty network element delimiting method for time synchronization fault delimitation based on machine learning provided in an embodiment of the present invention. If the selected link contains a total of 5 network elements A, B, C, D, and E, each network element will push the inference result to the management background, and the direction of the time synchronization link is from A to E. When B fails and causes a timestamp exception, the timestamp of network element C, which uses B as the master clock for synchronization, will also be abnormal. Similarly, network elements D and E will also have timestamp exceptions. Among the network elements that generate timestamp exceptions, B is at the upstream of the link, and B is determined to be a faulty network element. The network element that has a T1 or T4 timestamp exception and is located at the upstream of the link is a faulty network element, and this is a Master instance exception; or the network element that has a T2 or T3 timestamp exception and is located at the upstream of the link is a faulty network element, and this is a Slave instance exception.

[0059] The embodiment of the present invention provides a method, device and storage medium for time synchronization fault delimitation based on machine learning, which belongs to the field of communication technology. The method for time synchronization fault delimitation based on machine learning includes: determining a monitored time synchronization link; obtaining the timestamp information of the monitored time synchronization link when receiving a network element-level delimitation instruction and / or a module-level delimitation instruction; performing data processing on the timestamp information to generate a network element timestamp feature and / or a module timestamp feature, and determining the faulty network element and / or the faulty module based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature. The present invention performs fault analysis on the time synchronization link that needs to be monitored by using a preset machine learning algorithm, realizes network element-level delimitation, module-level delimitation and network element-module two-level delimitation, and improves the reliability of fault delimitation in the communication system.

[0060] based on Figure 2 In the embodiment shown, in this embodiment, step S201 includes:

[0061] Determining a preset number of message timestamps based on the monitoring time synchronization link;

[0062] A preset number of message timestamps and corresponding sampling point information are read to generate the timestamp information.

[0063] In a specific embodiment, the fault module is determined by taking the SPN product as an example. After receiving the request, the network element configures the underlying high-speed chip to latch the message timestamp (T1~T4) of the corresponding 1588 instance; the data acquisition module reads the latched timestamp and caches multiple sampling points to form timestamp information, and performs data processing. The data processing module cleans and features the data to extract important features that can characterize the characteristics of timestamp information changes. The steps of feature engineering include but are not limited to differential processing, time series statistic feature extraction, time domain / frequency domain feature extraction of time series, feature normalization, etc.

[0064] based on Figure 2 In the embodiment shown, in this embodiment, before step S202, the following steps are included:

[0065] Based on the preset machine learning algorithm, determining whether the network element timestamp feature and / or the module timestamp feature is abnormal;

[0066] When it is determined that the network element timestamp feature and / or the module timestamp feature is abnormal, the corresponding network element or module is determined to be an abnormal network element or abnormal module.

[0067] Specifically, the network element timestamp features and the module timestamp features include: 1. Time domain features, such as differential mean and differential variance. 2. Statistical features, such as mean, variance, peak, etc. 3. Frequency domain features, such as Fourier transform or wavelet transform of time series. The network element timestamp features and the module timestamp features are manually detected. When the network element timestamp features and the module timestamp features are obtained, these feature values ​​are compared with the standard timestamp features in the preset configuration library. When the two match, the network element timestamp features and / or the module timestamp features are judged to be normal.

[0068] The extracted time series features are sent to a lightweight machine learning algorithm for inference. The inference result is whether the timestamp information (T1-T4) is abnormal, and the result is pushed to the management backend. Machine learning algorithms include but are not limited to gradient boosting trees, support vector machines, naive Bayes and other algorithms; one or more algorithms can be used for prediction to obtain the prediction results of each algorithm, and then combined with the weight coefficients, the final prediction results are obtained by weighting.

[0069] based on Figure 2 In the embodiment shown, in this embodiment, step S202 includes the steps of:

[0070] In the case where the detection result of each network element or each module of the time synchronization link is detected to be abnormal, determining at least one abnormal network element or at least one abnormal module;

[0071] The abnormal network element most upstream in the time synchronization link is determined as the faulty network element.

[0072] Please refer to Figure 4 , Figure 4 A flowchart of a time synchronization fault demarcation method based on machine learning is provided for the third embodiment of the present invention.

[0073] based on Figure 1 In the embodiment shown, in this embodiment, Figure 4 As shown, step S20 further includes step S203 to step S204.

[0074] Step S203: when receiving the network element-module delimitation instruction, index the configuration instance corresponding to the monitoring time synchronization link;

[0075] Specifically, the management backend indexes the 1588 configuration instances of the network element nodes according to the selected synchronization link, including but not limited to the Master and Slave configuration instances; and forms an analysis request for the configuration instance, which is sent to the corresponding network element to trigger network element-level reasoning. The TSN network performs time synchronization based on the IEEE 802.1AS protocol.

[0076] Step S204: Determine the faulty network element based on the configuration instance and the preset machine learning algorithm, and determine the faulty module in the faulty network element.

[0077] In a specific embodiment, after receiving the request, the network element configures the underlying high-speed chip to latch the message timestamp (T1~T4) of the corresponding 1588 instance; collects timestamp information, 1588 configuration instance information, and diagnostic information of related modules, and performs data processing. The above static configuration data and dynamic time series data are integrated, and also analyzed and cleaned, and feature engineering related processing is performed to extract important features that can reflect business characteristics. The categories of extracted features include but are not limited to the following categories: 1. Time domain features, such as differential mean and differential variance. 2. Statistical features, such as mean, variance, peak, etc. 3. Frequency domain features, such as Fourier transform or wavelet transform of time series.

[0078] The extracted business features are sent to the lightweight machine learning algorithm for reasoning. The reasoning result is the module number to which the fault belongs, and the result is pushed to the management background. Fault modules include but are not limited to crystal oscillators, phase-locked loop chips, FPGAs, switching chips, etc. The machine learning algorithm is a supervised learning algorithm, including but not limited to tree model algorithms (such as random forests, gradient boosting trees), support vector machines, artificial neural networks, naive Bayes and other algorithms; one or more algorithms can be used for prediction to obtain the prediction results of each algorithm, and then combined with the weight coefficients, the weighted results are obtained to obtain the final prediction results.

[0079] The management backend receives the inference results of the faulty network elements, indexes the relevant causes and repair suggestions by the fault block number, and pushes them to the user.

[0080] The embodiments of the present invention provide a method, device and storage medium for time synchronization fault delimitation based on machine learning, which belongs to the field of communication technology. The method for time synchronization fault delimitation based on machine learning includes: determining a monitored time synchronization link; when a network element-module delimitation instruction is received, indexing a configuration instance corresponding to the monitored time synchronization link; based on the configuration instance and the preset machine learning algorithm, determining the faulty network element, and determining the faulty module in the faulty network element. The present invention performs fault analysis on the time synchronization link that requires monitoring by using a preset machine learning algorithm, realizes network element-level delimitation, module-level delimitation and network element-module two-level delimitation, and improves the reliability of fault delimitation in the communication system.

[0081] based on Figure 4 In the embodiment shown, in this embodiment, step S204 includes:

[0082] Based on the configuration example, determining a standard network element timestamp feature and a standard module timestamp feature;

[0083] By using the preset machine learning algorithm, the network element timestamp feature of each network element in the monitoring time synchronization link is compared with the standard network element timestamp feature to determine the faulty network element;

[0084] The module timestamp feature of the faulty network element is extracted, and the module timestamp feature is compared with the standard module timestamp feature through the preset machine learning algorithm to determine the faulty module.

[0085] In a specific embodiment, a lightweight supervised machine learning algorithm is deployed in the network element, including but not limited to tree model algorithms (such as random forests, gradient boosting trees), support vector machines, artificial neural networks, naive Bayes and other algorithms; multi-classification is performed on multivariate features, and the fault module number is output. The fault module includes but is not limited to crystal oscillators, phase-locked loop chips, FPGAs, switching chips, etc., and finally the results are pushed to the data channel.

[0086] Based on all the above embodiments, in this embodiment, after step S20, the following steps are further included:

[0087] The fault causes and repair suggestions matching the faulty network element and / or the faulty module are indexed and pushed to the user end.

[0088] Specifically, the inference result pushed by the faulty network element, that is, the faulty module number, is received, and the relevant causes and repair suggestions are indexed and pushed to the user.

[0089] See also Figure 5 , Figure 5 A schematic block diagram of the structure of a time synchronization fault demarcation device based on machine learning provided in an embodiment of the present invention.

[0090] like Figure 5 As shown, the time synchronization fault delimitation device 300 based on machine learning includes a processor 301 and a memory 302, and the processor 301 and the memory 302 are connected via a bus 303, such as an I2C (Inter-integrated Circuit) bus.

[0091] Specifically, the processor 301 is used to provide computing and control capabilities to support the operation of the entire machine learning-based time synchronization fault delimitation device. The processor 301 can be a central processing unit (CPU), and the processor 301 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0092] Specifically, the memory 302 may be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a mobile hard disk.

[0093] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a partial structure related to the embodiment of the present invention, and does not constitute a limitation on the time synchronization fault delimitation device based on machine learning to which the embodiment of the present invention is applied. The specific time synchronization fault delimitation device based on machine learning may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0094] The processor 301 is used to run a computer program stored in the memory 302, and implement any one of the time synchronization fault demarcation methods based on machine learning provided in the embodiments of the present invention when executing the computer program.

[0095] In one embodiment, the processor 301 is used to run a computer program stored in a memory, and implement the following steps when executing the computer program:

[0096] Determine monitoring time synchronization link;

[0097] The monitoring time synchronization link is subjected to fault analysis by a preset machine learning algorithm, and the corresponding faulty network element and / or faulty module is determined to achieve fault demarcation.

[0098] In one embodiment, the monitoring time synchronization link is subjected to fault analysis by a preset machine learning algorithm to determine the corresponding faulty network element and / or faulty module to implement fault demarcation, which is used to implement:

[0099] When receiving the network element level delimitation instruction and / or the module level delimitation instruction, obtaining the timestamp information of the monitoring time synchronization link;

[0100] The timestamp information is processed to generate a network element timestamp feature and / or a module timestamp feature, and the faulty network element and / or the faulty module is determined based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature.

[0101] In one embodiment, when a network element level delimitation instruction and / or a module level delimitation instruction is received, the timestamp information of the monitoring time synchronization link is obtained to implement:

[0102] Determining a preset number of message timestamps based on the monitoring time synchronization link;

[0103] A preset number of message timestamps and corresponding sampling point information are read to generate the timestamp information.

[0104] In one embodiment, before performing data processing on the timestamp information to generate a network element timestamp feature and / or a module timestamp feature, and determining the faulty network element and / or the faulty module based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature, it is also used to implement:

[0105] Based on the preset machine learning algorithm, determining whether the network element timestamp feature and / or the module timestamp feature is abnormal;

[0106] When it is determined that the network element timestamp feature and / or the module timestamp feature is abnormal, the corresponding network element or module is determined to be an abnormal network element or abnormal module.

[0107] In one embodiment, data processing is performed on the timestamp information to generate a network element timestamp feature and / or a module timestamp feature, and based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature, the faulty network element and / or the faulty module is determined to achieve:

[0108] In the case where the detection result of each network element or each module of the time synchronization link is detected to be abnormal, determining at least one abnormal network element or at least one abnormal module;

[0109] The abnormal network element most upstream in the time synchronization link is determined as the faulty network element.

[0110] In one embodiment, the performing fault analysis on the monitoring time synchronization link by using a preset machine learning algorithm to determine the corresponding faulty network element and / or faulty module to achieve fault demarcation is also used to achieve:

[0111] In case of receiving a network element-module delimitation instruction, indexing a configuration instance corresponding to the monitoring time synchronization link;

[0112] Based on the configuration instance and the preset machine learning algorithm, the faulty network element is determined, and the faulty module in the faulty network element is determined.

[0113] In one embodiment, based on the configuration instance and the preset machine learning algorithm, the faulty network element is determined, and the faulty module in the faulty network element is determined, so as to implement:

[0114] Based on the configuration example, determining a standard network element timestamp feature and a standard module timestamp feature;

[0115] By using the preset machine learning algorithm, the network element timestamp feature of each network element in the monitoring time synchronization link is compared with the standard network element timestamp feature to determine the faulty network element;

[0116] The module timestamp feature of the faulty network element is extracted, and the module timestamp feature is compared with the standard module timestamp feature through the preset machine learning algorithm to determine the faulty module.

[0117] In one embodiment, after performing fault analysis on the monitoring time synchronization link by using a preset machine learning algorithm and determining the corresponding faulty network element and / or faulty module to implement fault demarcation, it is used to implement:

[0118] The fault causes and repair suggestions matching the faulty network element and / or the faulty module are indexed and pushed to the user end.

[0119] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working process of the time synchronization fault delimitation device based on machine learning described above can refer to the corresponding process in the aforementioned embodiment of the time synchronization fault delimitation method based on machine learning, and will not be repeated here.

[0120] An embodiment of the present invention also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement any one of the time synchronization fault delimitation methods based on machine learning provided in the description of the embodiment of the present invention.

[0121] The storage medium may be an internal storage unit of the machine learning-based time synchronization fault delimitation device described in the aforementioned embodiment, such as a hard disk or memory of the machine learning-based time synchronization fault delimitation device. The storage medium may also be an external storage device of the machine learning-based time synchronization fault delimitation device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the machine learning-based time synchronization fault delimitation device.

[0122] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware embodiment, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that communication media typically contain 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.

[0123] It should be understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0124] The serial numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A time synchronization fault demarcation method based on machine learning, characterized in that: include: Determine monitoring time synchronization link; The monitoring time synchronization link is subjected to fault analysis by a preset machine learning algorithm, and the corresponding faulty network element and / or faulty module is determined to achieve fault demarcation.

2. The method for time synchronization fault demarcation based on machine learning according to claim 1, characterized in that: The performing fault analysis on the monitoring time synchronization link by using a preset machine learning algorithm to determine the corresponding faulty network element and / or faulty module to achieve fault demarcation includes: When receiving the network element level delimitation instruction and / or the module level delimitation instruction, obtaining the timestamp information of the monitoring time synchronization link; The timestamp information is processed to generate a network element timestamp feature and / or a module timestamp feature, and the faulty network element and / or the faulty module is determined based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature.

3. The method for time synchronization fault demarcation based on machine learning according to claim 2, characterized in that: The acquiring the timestamp information of the monitoring time synchronization link when receiving the network element level delimitation instruction and / or the module level delimitation instruction includes: Determining a preset number of message timestamps based on the monitoring time synchronization link; A preset number of message timestamps and corresponding sampling point information are read to generate the timestamp information.

4. The method for time synchronization fault demarcation based on machine learning according to claim 2, characterized in that: Before the data processing of the timestamp information to generate a network element timestamp feature and / or a module timestamp feature, and determining the faulty network element and / or the faulty module based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature, the method further includes: Based on the preset machine learning algorithm, determining whether the network element timestamp feature and / or the module timestamp feature is abnormal; When it is determined that the network element timestamp feature and / or the module timestamp feature is abnormal, the corresponding network element or module is determined to be an abnormal network element or abnormal module.

5. The method for time synchronization fault demarcation based on machine learning according to claim 4, characterized in that: The performing data processing on the timestamp information to generate a network element timestamp feature and / or a module timestamp feature, and determining the faulty network element and / or the faulty module based on the preset machine learning algorithm, the network element timestamp feature and / or the module timestamp feature, includes: In the case where the detection result of each network element or each module of the time synchronization link is detected to be abnormal, determining at least one abnormal network element or at least one abnormal module; The abnormal network element most upstream in the time synchronization link is determined as the faulty network element.

6. The method for time synchronization fault demarcation based on machine learning according to claim 1, characterized in that: The performing fault analysis on the monitoring time synchronization link by using a preset machine learning algorithm to determine the corresponding faulty network element and / or faulty module to achieve fault demarcation also includes: In case of receiving a network element-module delimitation instruction, indexing a configuration instance corresponding to the monitoring time synchronization link; Based on the configuration instance and the preset machine learning algorithm, the faulty network element is determined, and the faulty module in the faulty network element is determined.

7. The method for time synchronization fault demarcation based on machine learning according to claim 6, characterized in that: The determining the faulty network element based on the configuration instance and the preset machine learning algorithm, and determining the faulty module in the faulty network element, comprises: Based on the configuration example, determining a standard network element timestamp feature and a standard module timestamp feature; By using the preset machine learning algorithm, the network element timestamp feature of each network element in the monitoring time synchronization link is compared with the standard network element timestamp feature to determine the faulty network element; The module timestamp feature of the faulty network element is extracted, and the module timestamp feature is compared with the standard module timestamp feature through the preset machine learning algorithm to determine the faulty module.

8. The method for time synchronization fault demarcation based on machine learning according to any one of claims 1 to 7, characterized in that: After performing fault analysis on the monitoring time synchronization link by using a preset machine learning algorithm and determining the corresponding faulty network element and / or faulty module to achieve fault demarcation, the method includes: The fault causes and repair suggestions matching the faulty network element and / or the faulty module are indexed and pushed to the user end.

9. A time synchronization fault demarcation device based on machine learning, characterized in that: The machine learning-based time synchronization fault delimitation device includes a processor, a memory, a computer program stored in the memory and executable by the processor, and a data bus for realizing connection communication between the processor and the memory, wherein the computer program, when executed by the processor, implements the steps of the machine learning-based time synchronization fault delimitation method as described in any one of claims 1 to 8.

10. A storage medium for computer-readable storage, characterized in that: The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the method for time synchronization fault delimitation based on machine learning as described in any one of claims 1 to 8.

Images