阅读说明：本技术 一种实现光链路故障识别的方法、装置及系统 (Method, device and system for realizing optical link fault identification ) 是由 肖欣 张朝 李健 高云鹏 于 2018-08-16 设计创作，主要内容包括：本申请实施例中提供一种实现光链路故障识别的方法,涉及通信技术领域,包括：获取网络设备上的性能数据,提取所述性能数据的特征参数；依据所述特征参数进行光链路中的故障模式识别；解决了光链路中因设备数量庞大、线路故障多、人工排查情况难获得导致的故障情况识别难、处理慢等问题；达到了在故障发生时快速识别故障,提高排障效率；并在光链路隐患未导致故障发生时,可提前通过特征发现性能数据劣化,识别并预警等。(The embodiment of the application provides a method for realizing optical link fault identification, which relates to the technical field of communication and comprises the following steps: acquiring performance data on network equipment, and extracting characteristic parameters of the performance data; identifying a fault mode in the optical link according to the characteristic parameters; the problems of difficult fault condition identification, slow processing and the like caused by large equipment quantity, multiple line faults and difficult manual troubleshooting in the optical link are solved; the fault is rapidly identified when the fault occurs, and the fault removing efficiency is improved; and when the hidden trouble of the optical link does not cause the fault, the performance data degradation can be found in advance through the characteristics, and the performance data degradation can be identified and early-warned.)

1. A method for implementing optical link failure identification, comprising: acquiring performance data at least containing receiving optical power on network equipment in an optical link; extracting characteristic parameters for indicating the performance data change in a preset time window, and identifying a fault mode in an optical link according to the characteristic parameters; wherein the performance data is a time series within the preset time window.

2. The method of claim 1, wherein: the performance data also includes transmitted optical power and/or fiber length.

3. The method of claim 1 or 2, wherein the characteristic parameters comprise: the characteristic parameters are used for representing the abnormal degree of the performance data and/or the characteristic parameters are used for representing the variation trend of the performance data.

4. The method of claim 3, wherein the characteristic parameters for characterizing the degree of abnormality of the performance data include one or any combination of the following parameters:

the jitter degree is the random variation degree of the received optical power in the preset time window;

the weak light ratio is the ratio of the weak light time length in the received light power to the total time length of the preset time window;

the over-strong ratio is the ratio of the strong light time length in the received light power to the total time length of the preset time window; and the number of the first and second groups,

and the light-free ratio is the ratio of the light-free time length in the received light power to the total time length of the preset time window.

5. The method according to claim 3 or 4, wherein the parameters for characterizing the performance data trend include one or any combination of the following parameters:

the rebound times are used for representing the fluctuation times of the received optical power fitting processing result;

the degradation degree is used for representing the descending trend of the received optical power fitting processing result;

the rising times are used for representing the rising trend of the received optical power fitting processing result; and the number of the first and second groups,

and the mutation times are used for representing the times of deviation of the received optical power from the historical average optical power in the preset time window.

6. The method according to claims 1 to 5, wherein said identifying a failure mode in the optical link according to said characteristic parameter comprises:

and matching the characteristic parameters with a fault mode identification model to determine a fault mode.

7. The method of claim 6, further comprising:

training a fault mode identification model with characteristic parameters corresponding to a known fault mode by using historical data of the known fault mode;

the historical data includes characteristic parameters and failure modes corresponding to the characteristic parameters.

8. The method of claim 7, wherein after identifying the failure mode based on the characteristic parameter, the method further comprises:

and comparing the identified fault mode with the actual fault mode corresponding to the characteristic parameter, and if the identified fault mode is not consistent with the actual fault mode corresponding to the characteristic parameter, adding the characteristic parameter and the actual fault mode corresponding to the characteristic parameter into a training set.

9. The method of claim 8, wherein after adding the characteristic parameter and the actual failure mode corresponding to the characteristic parameter to the training set, further comprising: training the training set to establish a new fault mode recognition model; and the training set consists of historical data during the training of the fault pattern recognition model and the added characteristic parameters and actual fault patterns corresponding to the characteristic parameters.

10. An apparatus for implementing optical link failure identification, comprising:

an obtaining unit, configured to obtain performance data at least including received optical power on a network device in an optical link; the performance data acquired by the acquisition unit is a time sequence in the preset time window;

an extraction unit configured to extract a characteristic parameter indicating a change in the performance data acquired by the acquisition unit within a preset time window;

and the identification unit is used for identifying the fault mode in the optical link according to the characteristic parameters extracted by the extraction unit.

11. The apparatus of claim 10, wherein: the performance data acquired by the acquisition unit further includes a transmitted optical power and/or a fiber length.

12. The apparatus according to claim 10 or 11, wherein the feature parameters extracted by the extraction unit include: the characteristic parameters are used for representing the abnormal degree of the performance data and/or the characteristic parameters are used for representing the variation trend of the performance data.

13. The apparatus of claims 10 to 12, wherein the identification unit comprises: a matching module and an output module; wherein the content of the first and second substances,

the matching module is used for matching the characteristic parameters extracted by the extraction unit with a fault mode identification model so as to determine a fault mode;

the output module is used for outputting the fault mode determined by the matching module to external optical network equipment.

14. The apparatus of claim 13, further comprising: the training unit is used for training a fault mode identification model with characteristic parameters corresponding to a known fault mode by utilizing historical data of the known fault mode; the historical data comprises characteristic parameters and fault modes corresponding to the characteristic parameters.

15. The apparatus of claim 14, further comprising: a comparison unit and an addition unit; wherein the content of the first and second substances,

the comparison unit is used for comparing the fault mode identified by the identification unit with the actual fault mode corresponding to the characteristic parameter and outputting a comparison result to the addition unit;

the adding unit is used for receiving the output result of the comparing unit and adding the characteristic parameters and the actual fault modes corresponding to the characteristic parameters into the training set when the fault modes identified by the identifying unit are inconsistent with the actual fault modes corresponding to the characteristic parameters.

16. The apparatus of claim 15, further comprising: the establishing unit is used for training the training set to establish a new fault mode recognition model; and the training set consists of historical data during the training of the fault pattern recognition model and the added characteristic parameters and actual fault patterns corresponding to the characteristic parameters.

17. A system for realizing optical link fault recognition comprises optical link network equipment and a big data analysis online platform, and is characterized in that:

the big data analysis online platform is provided with the device according to claims 10 to 16, and receives performance data at least containing receiving optical power uploaded by the optical link network equipment.

Technical Field

The present application relates to the field of communications technologies, and in particular, to a method, an apparatus, and a system for identifying an optical link failure.

Background

With the increase of complexity and transmission capacity of network systems, optical fiber communication becomes the current main wired communication mode by utilizing the advantages of large transmission capacity, good confidentiality and the like. However, the optical fiber generally adopts glass as a waveguide, so that the optical fiber is brittle in texture and poor in mechanical strength, and the optical fiber is bent and broken in the use process. In addition, optical fiber connectors are generally used for reducing the cost of connection between optical fibers, but the phenomena of loose joints, too loose connection and the like occur in the long-time connection of the optical fiber connectors, so that the fault proportion of an optical link in optical fiber communication is increased, the reliability of the optical link is reduced, and the user experience is poor, so that fault repair needs to be performed in time to maintain the reliability of the optical link.

The conventional alarm system is used for establishing a threshold alarm system aiming at optical link network equipment or service performance data respectively, judging the size of a reference threshold value by utilizing the optical link network equipment or service performance data index in an observation system respectively to monitor whether the minimum working requirement is met, and if not, sending an alarm prompt to achieve the purpose of repairing the fault in time to maintain the reliability of the optical link. Because the network system is complex, the number of alarm prompts in the network system is huge, and a large number of other problems are mixed in the alarm prompts, only from the sent alarm prompts, network operation and maintenance personnel cannot obtain effective alarm information, cannot obtain an exact fault mode from the alarm prompts, and only can adopt a mode of manually troubleshooting, but manually troubleshooting is adopted, the processing efficiency is low, the fault repairing delay is too strong, finally, a user is affected for a long time and is complex to operate, and the judgment is carried out by utilizing the single-point time performance data of network equipment and services and a manually set baseline threshold value, so that alarm leakage or alarm mistake is easily caused, and the accuracy is low.

Content of application

The embodiment of the invention provides a method for realizing optical link fault identification, which solves the problems of low alarm prompt effectiveness, low fault processing efficiency, high fault repair delay, long time affected by a user, complex operation and the like caused by the fact that an exact fault mode cannot be obtained in the alarm prompt in the operation and maintenance of an optical link network system.

In order to achieve the purpose, the technical scheme is as follows:

in a first aspect, a method for implementing optical link failure identification is provided, including: acquiring performance data at least containing receiving optical power on network equipment in an optical link; extracting characteristic parameters for indicating the performance data change in a preset time window, and identifying a fault mode in an optical link according to the characteristic parameters; wherein the performance data is a time series within the preset time window.

In one possible implementation, the performance data further includes a transmitted optical power and/or a fiber length.

In one possible implementation, the characteristic parameters include: the characteristic parameters are used for representing the abnormal degree of the performance data and/or the characteristic parameters are used for representing the variation trend of the performance data.

In a possible implementation manner, the characteristic parameter for characterizing the degree of abnormality of the performance data includes one or any combination of the following parameters:

the jitter degree is the random variation degree of the received optical power in the preset time window;

the weak light ratio is the ratio of the weak light time length in the received light power to the total time length of the preset time window;

the over-strong ratio is the ratio of the strong light time length in the received light power to the total time length of the preset time window; and the number of the first and second groups,

and the light-free ratio is the ratio of the light-free time length in the received light power to the total time length of the preset time window.

In a possible implementation manner, the parameter for characterizing the performance data variation trend includes one or any combination of the following parameters:

the rebound times are used for representing the fluctuation times of the received optical power fitting processing result;

the degradation degree is used for representing the descending trend of the received optical power fitting processing result;

the rising times are used for representing the rising trend of the received optical power fitting processing result; and the number of the first and second groups,

and the mutation times are used for representing the times of deviation of the received optical power from the historical average optical power in the preset time window.

In a possible implementation manner, the identifying a failure mode in the optical link according to the characteristic parameter specifically includes:

and matching the characteristic parameters with a fault mode identification model to determine a fault mode.

In one possible implementation, the method further includes:

training a fault mode identification model with characteristic parameters corresponding to a known fault mode by using historical data of the known fault mode;

the historical data includes characteristic parameters and failure modes corresponding to the characteristic parameters.

In a possible implementation manner, after the fault pattern recognition is performed according to the characteristic parameters, the method further includes:

and comparing the identified fault mode with the actual fault mode corresponding to the characteristic parameter, and if the identified fault mode is not consistent with the actual fault mode corresponding to the characteristic parameter, adding the characteristic parameter and the actual fault mode corresponding to the characteristic parameter into the training set.

In a possible implementation manner, after the adding the feature parameters and the failure modes identified by the feature parameters to the training set, the method further includes: training the training set to establish a new fault mode recognition model; and the training set consists of historical data during the training of the fault pattern recognition model and the added characteristic parameters and actual fault patterns corresponding to the characteristic parameters.

In another aspect, an apparatus for implementing optical link failure identification is further provided, including: an obtaining unit, configured to obtain performance data at least including received optical power on a network device in an optical link; the performance data acquired by the acquisition unit is a time sequence in the preset time window; an extraction unit configured to extract a characteristic parameter indicating a change in the performance data acquired by the acquisition unit within a preset time window; and the identification unit is used for identifying the fault mode in the optical link according to the characteristic parameters extracted by the extraction unit.

In a possible implementation manner, the performance data acquired by the acquiring unit further includes a transmission optical power and/or an optical fiber length.

In a possible implementation manner, the feature parameters extracted by the extraction unit include: the characteristic parameters are used for representing the abnormal degree of the performance data and/or the characteristic parameters are used for representing the variation trend of the performance data.

In one possible implementation manner, the identification unit includes: a matching module and an output module; wherein the content of the first and second substances,

the matching module is used for matching the characteristic parameters extracted by the extraction unit with a fault mode identification model so as to determine a fault mode;

the output module is used for outputting the fault mode determined by the matching module to external optical network equipment.

In one possible implementation, the apparatus further includes: the training unit is used for training a fault mode identification model with characteristic parameters corresponding to a known fault mode by utilizing historical data of the known fault mode; the historical data comprises characteristic parameters and fault modes corresponding to the characteristic parameters.

In one possible implementation, the apparatus further includes: a comparison unit and an addition unit; wherein the content of the first and second substances,

the comparison unit is used for comparing the fault mode identified by the identification unit with the actual fault mode corresponding to the characteristic parameter and outputting a comparison result to the addition unit;

the adding unit is used for receiving the output result of the comparing unit and adding the characteristic parameters and the actual fault modes corresponding to the characteristic parameters into the training set when the fault modes identified by the identifying unit are inconsistent with the actual fault modes corresponding to the characteristic parameters.

In one possible implementation, the apparatus further includes: the establishing unit is used for training the training set to establish a new fault mode recognition model; and the training set consists of historical data during the training of the fault pattern recognition model and the added characteristic parameters and actual fault patterns corresponding to the characteristic parameters.

On the other hand, a system for realizing optical link fault identification is provided, which comprises an optical link network device and a big data analysis online platform: the big data analysis online platform is provided with the device and receives performance data at least containing receiving optical power uploaded by the optical link network equipment.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments of the present application will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive exercise.

FIG. 1 is a schematic diagram illustrating an optical link structure in an optical fiber communication system;

FIG. 2 is a diagram illustrating an application scenario a according to an embodiment of the present application;

fig. 3 is a schematic diagram of another application scenario b provided in an embodiment of the present application;

FIG. 4 is one of the schematic flow charts provided by one embodiment of the present application for implementing a method for optical link failure identification;

FIG. 5 is a graphical representation of IMSR fit results provided in accordance with an embodiment of the present application;

fig. 6 is a second schematic flowchart of a method for implementing optical link failure identification according to an embodiment of the present application;

FIG. 7 is a schematic flow chart diagram of a method for implementing optical link failure identification through a failure tree rule according to an embodiment of the present application;

FIG. 8 is a schematic flow chart diagram of a method for implementing optical link failure identification through historical data according to an embodiment of the present application;

fig. 9 is a third schematic flowchart of a method for implementing optical link failure identification according to an embodiment of the present application;

fig. 10 is a schematic structural diagram of an apparatus for implementing optical link failure identification according to an embodiment of the present application;

fig. 11 is a schematic structural diagram of a data processing portion in a system for implementing optical link failure identification according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

The term "and/or" herein is merely an association describing an associated object, meaning that three relationships may exist, e.g., a and/or B, may mean: a exists alone, A and B exist simultaneously, and B exists alone.

The terms "first" and "second," and the like, in the description and in the claims of the embodiments of the present application are used for distinguishing between different objects and not for describing a particular order of the objects. For example, the first target object and the second target object, etc. are specific sequences for distinguishing different target objects, rather than describing target objects.

In the embodiments of the present application, words such as "exemplary" or "for example" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "e.g.," is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, use of the word "exemplary" or "such as" is intended to present concepts related in a concrete fashion.

In the description of the embodiments of the present application, the meaning of "a plurality" means two or more unless otherwise specified. For example, a plurality of processing units refers to two or more processing units; the plurality of systems refers to two or more systems.

The technical solution in the present application will be described below with reference to the accompanying drawings.

At present, compared with spread spectrum communication and satellite communication, optical fiber communication is used as wired communication, and has the advantages of large communication capacity, long transmission distance, small attenuation, small volume, good anti-interference performance, convenient expansion and capability of saving a large amount of nonferrous metal resources. For example: the potential bandwidth of one fiber in fiber optic communications can be up to 20 THz.

In the optical fiber communication system, an optical link consists of an optical transmitter, an optical fiber, an optical receiver and other necessary optical devices, the principle is that the interconnection of networks is realized by an optical method to obtain optical interconnection networks with various topological structures, and the optical link transmits sound, images and data signals by using an optical fiber communication technology.

Fig. 1 is a schematic diagram of an optical link structure in an optical fiber communication system, as shown in fig. 1, the optical link includes: the optical transmitter, the optical relay amplifier, the optical receiver and the connecting optical fiber are multiple; the optical transmitter includes: the optical monitoring system comprises N optical repeaters, an optical multiplexer, a power Amplifier (BA for short) and an optical supervisory channel transmitter, wherein the N optical repeaters are used for receiving N light waves and have a photoelectric conversion function and a signal processing function; the optical combiner is used for combining N optical signals with different wavelengths in one optical fiber for transmission; the BA is used for carrying out power boosting and transmission on the N wavelength signals of the combined wave; the optical monitoring channel transmitter is used for monitoring the transmission condition of each channel in the system and transmitting signals. The optical relay amplification comprises: the system comprises an optical monitoring channel receiver, a Line-Amplifier (LA) and an optical monitoring channel transmitter, wherein the optical monitoring channel receiver is used for monitoring the transmission condition of each channel in the system and receiving signals; wherein, the LA is used for periodically compensating the transmission loss of the line; wherein the optical supervisory channel transmitter is for use in accordance with an optical supervisory channel transmitter in the optical transmitter. The optical receiver includes: the optical relay Amplifier comprises an optical monitoring channel receiver, a preamplifier (Pre-Amplifier, abbreviated as PA), an optical demultiplexer and N optical receiving ends, wherein the function of the optical monitoring channel receiver is consistent with the function of the optical monitoring channel receiver in the optical relay Amplifier; the PA is used for Signal amplification and improving the sensitivity of the receiver (for example, when an Optical Signal Noise Ratio (OSNR) meets requirements (the required Noise index is small), the Noise of the receiver can be suppressed by larger input power and the receiving sensitivity is improved); the optical branching filter is used for separating a plurality of optical signals of different sections from each other on the space of light intensity and wavelength; the N optical receiving ends are configured to receive the separated optical signals with N wavelengths; the optical fiber is used for connecting the optical transmitter and the optical repeater, and the optical repeater and the optical receiver are connected to form an optical link.

It should be noted that the optical link structure should be understood as an example of an implementation manner of the optical link structure in the optical fiber communication system, and is not limited to the optical link structure in the optical fiber communication system.

In the above description, optical fiber is used as a transmission medium because optical fiber generally uses glass as a waveguide to transmit information in the form of light from one end to the other end, and low loss glass optical fiber is hardly subject to bandwidth limitation, making optical fiber a convenient transmission tool. In addition, the optical fiber is used as a transmission tool and has high sensitivity and is not interfered by electromagnetic noise; small volume, light weight, long service life and low price; the cable is insulating, high-pressure resistant, high-temperature resistant and corrosion resistant, and is suitable for working in special environments; the geometric shape can be adjusted according to the requirement of the environment, and the signal transmission is easy; high bandwidth, large communication quantity, small attenuation and long transmission distance; the signal crosstalk is small, and the transmission quality is high; the security is high; convenient laying and carrying of raw materials and the like. However, since the optical fiber uses glass as the waveguide, the optical fiber is brittle and has poor mechanical strength, and the optical fiber is bent and broken during use. In addition, optical fiber connectors are generally used for connecting optical fibers to reduce the cost, but the phenomena of loose joints, over-loose connections and the like occur when the optical fiber connectors are connected for a long time, so that faults occur in optical links.

It should be noted that each component element except for the optical fiber in the network system can be simply understood as a device in the network.

Specifically, the Network device in the Optical link may be an Optical Line Terminal (OLT), an Optical Network Unit (ONU), or the like. The OLT is a network device connected with an optical distribution network consisting of a plurality of ONUs through passive optical cables, optical splitting/combining devices and the like.

It should be understood that the network device in the optical link may be a hardware device or the like having a transmission function and capable of executing a software program, and is not limited to the above-listed device or terminal.

The method for realizing the optical link fault identification can be applied to an OLT (optical line terminal) and an ONU (optical network unit) and can also be applied to a big data analysis online platform. Firstly, acquiring performance data at least containing received optical power on network equipment; extracting characteristic parameters for indicating the performance data in a preset time window; and finally, identifying the characteristic parameters to identify a fault mode.

The devices for realizing optical link failure identification in an optical link are different in terms of functions, performance, reliability and the like according to the difference of chips, software programs and connection positions carried in network equipment. For example: as shown in fig. 3, the OLT carries a feature parameter extraction model and a fault pattern recognition model, so that the OLT can perform feature parameter extraction and fault pattern recognition; as shown in fig. 2, the characteristic parameter extraction model and the fault pattern recognition model are disposed on the big data analysis online platform, and the big data analysis online platform performs the characteristic parameter extraction and the fault pattern recognition operations, and the OLT performs the performance data acquisition and uploading operations.

Specifically, as shown in fig. 2, in a scenario (a), an ONU transmits KPI data to an OLT, and a big data analysis online platform provided with a feature parameter extraction model and a fault pattern recognition model periodically acquires the KPI data from the OLT (for example, periodically for 15min), and after the big data analysis online platform obtains the KPI data, the big data analysis online platform extracts feature parameters of the KPI data by using a feature parameter extraction model or an algorithm, so as to obtain feature parameters of the KPI data, and then performs fault pattern recognition according to the feature parameters of the KPI data by using the fault pattern recognition model, so as to determine a fault pattern of an optical link. As shown in fig. 3, in the scenario (b), the ONU transmits KPI data to the OLT including a feature parameter extraction model and a fault pattern recognition model, and after the OLT obtains the KPI data, the OLT performs feature extraction on the KPI data by using a feature parameter extraction model or an algorithm, and obtains feature parameters of the KPI data, and then performs fault pattern recognition according to the feature parameters of the KPI data by using a fault pattern recognition model, and reports the recognized fault pattern to the big data analysis online platform. It should be noted that, in the present embodiment, the feature parameter extraction may be implemented by a parameter extraction method in the prior art, in addition to the extraction method in the following embodiments, and the present embodiment is not particularly limited.

The method obtains the performance data transmitted on the network equipment in the optical link, extracts the characteristic parameters of the performance data, so as to rapidly identify the fault mode in the optical link according to the performance characteristics, and reports the identified fault mode.

Fig. 4 is a schematic flowchart of a method for implementing optical link failure identification according to an embodiment of the present application. The method comprises the following steps:

401, acquiring performance data at least containing received optical power on network equipment in an optical link;

in this embodiment, the performance data at least includes a received optical power on the optical link network device, and certainly may also include at least one of a transmitted optical power on the optical link network device and optical fiber length data connected to the optical link network device; of course, the performance data may be performance data to be analyzed collected in the optical link network device, or may be KPI data acquired in real time and having a short period. It should be noted that, in this embodiment, the performance data on the network device is a time sequence collected from the network device within a preset time window, where the received optical power included in the performance data may be represented by (time, optical power value), for example: (2018-01-06:00:00:00, -20dB) which indicates that the optical power collected at 06 days 00:00 sec in 01 month in 2018 is-20 dB.

402, extracting characteristic parameters for indicating the performance data change in a preset time window;

specifically, in an embodiment of the present application, the feature extraction algorithm is to analyze performance data on the network device within a preset time window by using a big data technology to extract feature parameters; the characteristic parameters are used for representing the change condition of the performance data on the network equipment in a preset time window; it should be noted that, in this embodiment, the characteristic parameters may include a parameter for characterizing the degree of abnormality of the performance data and/or a parameter for characterizing the trend of change of the performance data, and the like; and the parameter for representing the abnormal degree of the performance data at least comprises one of a jitter degree, a weak light ratio, an over-strong ratio and a non-light ratio; the parameters for representing the variation trend of the performance data at least comprise one of rebound times, deterioration degrees, rising times and mutation times. The expression form of the characteristic parameter is not limited, and may be represented by a table, a graph or a character. For example, the vector form of the feature parameters can be expressed as: within the time windowThe characteristic parameter ═ x₁，x₂，x₃，x₄，x₅，x₆，x₇，x₈，x₉…); wherein x is₁Is the degree of jitter, x₂Is the ratio of weak light to x₃Is over-strong ratio, x₄Is a dark duty ratio, x₅Is the number of rebounds, x₆Is the number of deterioration times, x₇Is the degree of deterioration, x₈Is the number of rises, x₉The number of mutations, etc.

In this embodiment, the meaning of each feature value of the feature parameter may be represented as:

the jitter degree is a random variation degree of the received optical power within a preset time window, and is generally expressed by an overall standard deviation;

the weak light duty ratio is the ratio of the weak light duration of the received light power on the network equipment to the total duration of the preset time window in the preset time window, namely the weak light duty ratio is used for representing the weak light rate in the preset time window; specifically, by configuring a low-light threshold, calculating the proportion of a part of which the light power value is lower than the low-light threshold and is greater than the no-light threshold in a preset time window; for example: the weak light threshold is configured as follows: 30dB, the light power value corresponding to 30 time points in the window is lower than a low light threshold value and is greater than a no light threshold value, and the low light ratio is 0.3 when 100 data time points are totally arranged in the window;

the over-strong ratio is the ratio of the strong light time length of the receiving light power on the network equipment to the total time length of the preset time window in the preset time window, namely the over-strong ratio is used for expressing the strong light rate in the preset time window; specifically, the ratio of the portion of the optical power value higher than the strong light threshold value in a preset time window is calculated by configuring the strong light threshold value; for example, the highlight threshold is configured as: 8dB, the optical power value corresponding to 40 time points in the window is higher than the strong light threshold value, and the over-emphasis ratio is 0.4 when 100 data time points are in the window;

the dark duty ratio is a ratio of a dark time length of the received optical power on the network device to a total time length of a preset time window in the preset time window, namely, the dark duty ratio is used for representing the dark rate in the preset time window; specifically, the occupation ratio of the part of the optical power value lower than the lightless threshold value in a preset time window is calculated by configuring the lightless threshold value; for example, the configured matte threshold is: 35dB, the light power value corresponding to 50 time points in the window is lower than the lightless threshold value, and the lightless ratio is 0.5 when the window has 100 data time points.

It should be noted that, the unit of configuring the weak light threshold is the unit of performance data on the network device; in addition, the above configuration of the weak light threshold, the strong light threshold, and the no light threshold is merely an example.

In this embodiment, when analyzing the variation trend of the performance data, fitting processing needs to be performed on the received optical power of the network device, and the variation trend of the performance data is represented by a fitting processing result; in this embodiment, a linear fitting operation is taken as an example for explanation, and the specific application is not limited to this, and the linear fitting result in this embodiment is represented in the form of a curve: wherein the content of the first and second substances,

acquiring the bounce times according to the positive and negative conditions between two adjacent sections in the linear fitting result, wherein the bounce times are used for representing the fluctuation times of the received optical power data fitting processing result; obtaining a degradation degree according to the number of sections which are counted as negative numbers in each section of the linear fitting result, wherein the degradation degree is used for representing the descending degree of the linear fitting result in a preset time window, namely the descending trend of the received optical power fitting result; acquiring the rising times according to the number of sections counted as positive real numbers in each section of the linear fitting result, wherein the rising times are used for expressing the rising degree of the linear fitting result in a preset time window, namely the rising trend of the received optical power fitting processing result; the mutation times are used for representing the times of the received light power data within a preset time window and deviating from the historical average light power;

the following illustrates that the linear fitting operation is performed on the received optical power, and each characteristic parameter for representing the change trend of the performance data is obtained, but the method is not limited to this in the practical application process:

researching an Iterative piecewise Regression model corresponding to N unknown turning points by configuring the maximum number N of Iterative turning points by using an Iterative piecewise Regression (IMSR) algorithm, and mining Multi-section trend characteristics between adjacent turning points; and the number N of unknown turning points is less than or equal to the number N of iterative turning points. Specifically, as shown in fig. 5, a fitting result of fitting operation on received optical power data within a preset time window by using an IMSR algorithm is shown, where a dark gray point is an original optical power time series, a light gray line is an IMSR fitting result, and a inflection point number n is 8 (including a start point and an end point):

each time segment has an inflection point of

Calculating the minimum value of Sum of squared residuals (SSE) of the fitted curve corresponding to all possible inflection points among 1 to N iterative inflection points by using Argmin function, wherein the minimum value is represented by the following formula 1, and each inflection point T is obtained when the Sum of squared residuals (SSE) of the fitted curve corresponding to all possible inflection points is the minimum value of the squared residuals among N (between 1 and N) iterative inflection points calculated by using Argmin function_nTaking values:

specifically, the calculation result of the SSE in formula 1, i.e. the calculation interval [ T ] is obtained by the following formula 2_i,T_i+1]Minimum of the sum of the squared residuals of the fitted model in (d), where both T1 and Tn are inflection times:

SSE(T₁,…,T_n)＝min∑T_i,T_i+1SSE

in formula 1 and formula 2, the number n of iterative inflection points represents the number of all possible inflection points, and an iterative piecewise regression model is selected to calculate the sum of the residual squares of n unknown inflection points, namely the sum of the residual squares of the fitting models for all inflection point intervals; wherein the residual is the difference of corresponding positions of a data line and a regression line statistically, and the sum of squares of the residual is the sum of the squares of each residual, which represents the effect of random errors. When the value satisfies the minimum sum of SSEs, the corresponding inflection point number n and the inflection point of each time segment

The fitting effect is best; interval [ T_i,T_i+1]Inner fitting model may be advantageousUsing linear fitting; and linear fitting when the sum of the minimum SSEs is minimum enables the error of the fitting model and the actual observation point to be minimum.

The linear equation fitted in the iterative piecewise regression model of equation 3 below:

Y_t＝A_it+K_i，T_i≤t≤T_i+1

wherein, Y_tThe value is performance data at a certain moment in the corresponding interval; wherein A is_iTrend coefficients for the fitted line segments; wherein, K_iIs the trend intercept of the fitted line segment; wherein, the iterative piecewise regression model is used to output data as the trend coefficient A of linear fitting corresponding to each inflection point T and each segment_i. Wherein, T is a value of time corresponding to the performance data corresponding to the iteration inflection point.

According to a fitting result of linear fitting operation of the IMSR algorithm on the received optical power in a preset time window, quantitative values of rebound times, degradation times, rising times and mutation times are determined, and relevant parameters are configured, wherein the specific method comprises the following steps:

the rebound times are as follows: and judging whether the situation of the opposite directions of the trend coefficients of the adjacent sections exists or not according to the fitted trend coefficients of the sections, and if the trend coefficient in the ith section is a negative real number (indicating that the section descends), and the trend coefficient in the (i + 1) th section nearby is a positive real number (indicating that the section ascends), recording the two sections as one rebound. An up/down trend coefficient threshold may be further configured and the degree of rebound identified may be controlled based on a comparison of the trend coefficient in section i with the threshold. The receiving optical power variation range threshold of each section can also be configured, and only the bounce times when the receiving optical power variation is larger than the threshold are counted through the receiving optical power variation range threshold. The number of bounces finally calculated is an integer.

The number of deterioration times: and calculating the number of the sections with the trend coefficients being negative real numbers according to the fitted trend coefficients of the sections. A trend coefficient degradation threshold value with which the number of sections in which the trend coefficient is lower than the degradation threshold value is counted may also be further configured. The threshold of the optical power variation range of each segment can also be configured, that is, only the number of segments with the optical power variation larger than the threshold of the optical power range is counted.

Degree of deterioration: for all the sections with degradation (in the same manner as the above-described manner of determining the number of degradation times), the trend coefficient in the later preset time section in the preset time window is output as the degradation degree.

The rising times are as follows: and calculating the number of the sections with the trend coefficients being positive real numbers according to the fitted trend coefficients of the sections. The trend coefficient rise threshold may also be further configured, i.e. only the number of segments for which the trend coefficient is higher than the rise threshold is counted. The threshold of the optical power range of each segment can also be configured, that is, only the number of segments with the optical power variation larger than the threshold of the optical power range is counted.

The mutation frequency is as follows: if two adjacent inflection points exist in the inflection points and are adjacent moments, and the absolute value of the optical power difference value corresponding to the two points is greater than the configured deviation threshold, marking as a sudden change. Can be expressed in terms of the number of times of '+/-', where '+' denotes a rising deviation and '-' denotes a falling deviation.

403, identifying a failure mode in the optical link according to the characteristic parameters;

in one embodiment of the application, the characteristic parameters are analyzed through a fault mode identification algorithm, and fault identification is performed by combining the characteristic parameters of the existing fault conditions. Specifically, the fault pattern recognition algorithm is to match the characteristic parameters with the fault pattern recognition model based on the characteristic parameters of the existing fault patterns to determine the fault patterns, so as to perform fault pattern recognition.

For example: according to the characteristic parameters of the received optical power data obtained in step 402: the jitter degree, the weak light ratio, the over-strong ratio, the no-light ratio, the rebound times, the deterioration times, the rising times, the mutation times and the like; and identifying a fault mode according to each variable value contained in the characteristic parameter of the optical power.

In an embodiment of the present application, a plurality of network devices exist in the optical link, and in order to quickly obtain performance data on the network devices in the optical link, the performance data of the network devices in the optical link may be monitored in real time; the monitoring can be real-time monitoring of online transmission data or periodic monitoring of the transmission data; for example: performance data on the network device is acquired every 15 seconds interval. The specific implementation of the monitoring is not specifically limited herein. The method and the device have the advantages that the problem that the fault mode in the optical link is difficult to obtain due to large equipment data volume, large line fault proportion and manual troubleshooting conditions in a network system is solved by acquiring the performance data on the network equipment and identifying the fault mode by using a fault mode identification algorithm, and the fault mode in the optical link can be timely and accurately identified under the condition that an alarm threshold does not need to be configured.

The fault mode identification provided by the application has two modes: a fault tree rule is configured in advance, and a fault condition is obtained by matching characteristic parameters with the fault tree rule; one is to establish a fault mode identification model with characteristic parameters corresponding to fault modes by using a data modeling mode, train the fault mode identification model by using fault mode historical data containing known fault conditions, apply the trained model on line and identify the fault conditions of the existing network.

FIG. 8 is a schematic flow chart diagram of a method for implementing optical link failure identification through historical data according to an embodiment of the present application; the method comprises the following steps:

801, training a fault mode identification model with characteristic parameters corresponding to a fault mode according to historical data of the known fault mode;

the training fault mode recognition model is used for extracting the characteristic parameters according to historical data of known fault modes and establishing the fault mode recognition model according to the characteristic parameters; the historical data comprises characteristic parameters and fault modes corresponding to the characteristic parameters.

Establishing a fault identification model according to the historical data by utilizing a classification algorithm or a regression algorithm with a discrimination function, training and establishing a fault mode identification model; specifically, the classification algorithm or the regression classification algorithm with the discrimination function may be a Gradient Boosting Decision Tree (GBDT) or a random forest algorithm.

When the algorithm is a GBDT algorithm, generating a weak classifier by multiple iterations of the historical data, training each classifier on the basis of the residual error of the last classifier, and finally obtaining a total classifier by weighting and summing the weak classifiers obtained by each iteration of training; wherein the weak classifiers generated in each round have a loss function; fitting each regression tree by using the value of the negative gradient of the loss function in the current model as an approximate value of a residual error in the regression problem lifting algorithm, and fitting the negative gradient of the loss function in the current model in each iteration so that the loss function is as small as possible, thereby enabling the GBDT algorithm to accurately generate the characteristic parameters of the historical data; and the characteristic parameters are characteristic combinations corresponding to the historical data. And then, establishing a fault mode identification model according to the characteristic parameters.

When the algorithm is a random forest algorithm, repeatedly and randomly extracting K samples in the N training sample sets in the historical data in a replacing mode to generate a new training sample set through a self-help resampling technology, then generating K classification trees according to the self-help sample set to form a random forest, and the characteristic parameters are determined according to the voting scores of the classification trees; and then establishing a fault mode identification model corresponding to the historical data according to the determined characteristic parameters.

802, performing online application on the trained fault mode identification model, and identifying the fault mode in the optical link according to the characteristic parameters;

the trained model is the fault mode identification model in the step 801; wherein, the failure mode identification model can be applied to step 403 in fig. 4, and the failure mode identification model is applied to identify the characteristic parameters in step 403.

According to an embodiment of the present application, after training the fault pattern recognition model according to the historical data of the known fault patterns, the method further includes: and taking the historical data as a training set and correspondingly storing the historical data and the fault pattern recognition model.

In one embodiment of the present application, the fault pattern recognition model is obtained by training according to historical data of the known fault patterns; therefore, the characteristic parameters to be identified by the fault pattern identification model are consistent with the characteristic parameters determined by the historical data through the algorithm, so that the fault pattern identification model and the historical data have a corresponding relation; and the fault pattern recognition model is obtained by training according to the historical data, so the historical data is used as a training set for training the fault pattern recognition model. The correspondingly storing the training set and the fault pattern recognition model may be storing the training set and the fault pattern recognition model in a storage unit having a corresponding relationship, or storing the training set and the fault pattern recognition model after making a corresponding label or tag, etc. It should be noted that, the way of storing the training set and the failure mode identification model correspondingly is not specifically limited.

According to one embodiment of the application, the fault mode identification model is an identification model established according to fault tree rules configured in advance through artificial experience.

FIG. 7 is a schematic flow chart diagram of a method for implementing optical link failure identification through a failure tree rule according to an embodiment of the present application; specifically, the implementation steps of the method include:

701, defining a fault tree rule in advance by using manual experience, and establishing a fault mode identification model according to the fault tree rule: the fault tree rule is defined in advance by utilizing manual experience, namely a fault tree defined according to a fault mode known by manual experience, and then the fault tree rule is defined according to the fault tree; then, a fault pattern recognition model is established according to the fault tree rules. The fault tree is a special inverted tree logic cause and effect relationship diagram, and event symbols, logic gate symbols and transition symbols are used for describing cause and effect relationships among various events in the system. The input event of a logic gate is the "cause" of the output event and the output event of the logic gate is the "effect" of the input event.

Specifically, in the embodiment of the present application, a fault tree is obtained through manual experience according to a known fault mode in an optical link by using a known fault in the optical link (for example, a fiber bend in the optical link, a direct connection without an optical splitter, a fiber break, etc.), and a fault tree rule is formulated.

For example: the optical fiber bending and the optical splitter-free direct connection are known failure modes in an optical link, wherein the obvious characteristic rules corresponding to the optical fiber bending can be that the rebound times are greater than 0, the degradation times are greater than 0, the rising times are greater than 0, the degradation degree is greater than a preset threshold value, and the jitter degree is greater than a certain threshold value. If the extracted features satisfy the rules, identifying the fiber as bent; the certain threshold is a numerical value preset according to manual experience; the obvious feature rule corresponding to the optical splitter-free direct connection can be that the over-emphasis ratio is greater than a certain threshold, the rising frequency is greater than 0, the mutation frequency is greater than 0, and the deterioration frequency is 0, and if the extracted features meet the rule, the optical splitter-free direct connection is identified; the certain threshold is set according to the characteristics of the fault condition, each fault mode has a specific rule, and different thresholds may be corresponding to the certain threshold according to manual experience.

And 702, performing online application on the established fault mode identification model, and performing fault mode identification in the optical link according to the characteristic parameters.

The failure mode identification model may be applied to step 403 in fig. 4, and the failure mode identification model is applied to identify the characteristic parameters in step 403. The specific process of identifying the fault mode in the optical link according to the characteristic parameters comprises the following steps: matching the fault tree rule with the extracted characteristic parameters; if the match is successful, a failure mode may be identified.

For example: as shown in fig. 6, when the performance data is the time sequence in the preset time window, and the failure mode in the optical link is optical fiber bending, the failure tree rule set for optical fiber bending according to manual experience is as follows: the rebound frequency is larger than 0, the deterioration frequency is larger than 0, the rise frequency is larger than 0, the deterioration degree is larger than 0.2, and the jitter degree is larger than 0.5. When the characteristic parameters of the optical power data are: the rebound frequency is 2, the degradation frequency is 2, the rise frequency is 2, the degradation degree is 2, and the jitter degree is 1.92; according to the fault tree rule of the optical fiber bending setting, the rebound times are larger than 0, the degradation times are larger than 0, the rise times are larger than 0, the degradation degree is larger than 0.2, the jitter degree is larger than 0.5, and the fault mode is obtained as optical fiber bending.

In an embodiment of the application, after the identifying the fault pattern of the characteristic parameter by using the fault pattern identifying algorithm, the method further includes: and comparing the identified fault mode with the actual fault mode corresponding to the characteristic parameter, and if the identified fault mode is not consistent with the actual fault mode corresponding to the characteristic parameter, adding the characteristic parameter and the actual fault mode corresponding to the characteristic parameter into the training set.

Fig. 9 is a schematic flow chart of another method for identifying an optical link failure according to an embodiment of the present application; as shown in fig. 9, during online application, characteristic parameters are extracted from the collected performance data, and the fault pattern recognition model is trained for online recognition. The maintenance engineer can further obtain an actual fault mode according to the checking, feed back the accuracy of the fault mode and update the feedback data to a training set so as to improve the accuracy of model training. And as can be seen from fig. 9, the feedback supervision of the fault identification model is in an offline mode, whereas the fault identification is in an online mode, it can be seen that the supervision is a supervision of the accuracy of the fault identification model.

The actual fault mode is actually processed when a maintenance engineer maintains the fault mode in the optical link; the fault mode is identified according to the acquired performance data on the network equipment; in order to confirm the accuracy of establishing the identification result of the fault mode identification model, comparing the correspondingly stored fault mode with the actual fault mode in a maintenance engineer or an actual fault mode database to obtain the consistency of the fault mode and the actual fault mode, and if the fault mode is consistent with the actual fault mode, indicating that the identification result of the fault mode identification model is accurate; and if the fault mode is inconsistent with the actual fault mode, the identification result of the fault mode identification model is inaccurate, and the fault mode identification model needs to be reestablished.

When the fault mode recognition model needs to be reestablished, the performance data stored corresponding to the fault mode needs to be fed back to a training set, and the training set is trained to establish a new fault mode recognition model; wherein the training set consists of the historical data when training the fault pattern recognition model before no feedback, and consists of the historical data when training the fault pattern recognition model and the performance data after feedback.

The embodiment of the application provides a method for realizing optical link fault identification, which comprises the following steps: monitoring performance data in real time for network equipment in an optical link; acquiring performance data from the network device; extracting characteristic parameters of the performance data by using a characteristic extraction algorithm; carrying out fault mode identification on the characteristic parameters by using a fault mode identification model to obtain an identification result; the fault mode identification model consists of two parts, namely an offline fault mode identification model and an online updated fault mode identification model; the offline fault mode identification model is a fault mode identification model obtained by training with historical data corresponding to the existing fault condition in the optical link before online identification; and the online updating of the fault mode identification model is to take performance data corresponding to the fault mode identified by the fault mode identification model as feedback data, update the feedback data into a training set, train the feedback data by using the updated training set to obtain a new fault mode identification model, and update the new fault mode identification model into an online program or equipment. And the online system is not interrupted when acquiring real-time performance data, and the fault mode identification model is not interrupted when identifying the characteristic parameters of the acquired performance data.

The embodiment of the application provides a method for realizing optical link fault identification, which can be applied to a platform or optical link network equipment for big data analysis.

As shown in fig. 2, when applied to a big data analysis online platform, the big data analysis online platform identifies a fault part in the optical link, and receives KPI data transmitted by an ONU through an OLT and a fault pattern identification model of a big data analysis training platform; the big data analysis training platform trains a fault mode recognition model according to historical performance data or characteristic parameters, and the big data analysis training platform applied to the big data analysis online platform trains the data of the recognized fault mode to obtain a new fault mode recognition model and then updates the new fault mode recognition model to the big data analysis online platform; wherein the ONU and the OLT are connected through an optical fiber.

The detailed process is described as follows:

the ONU sends the KPI data to the OLT, the OLT receives the KPI data and periodically sends the KPI data to a big data analysis online platform, the big data analysis online platform extracts the characteristic parameters of the received KPI data and identifies the characteristic parameters of the KPI data by using the fault mode identification model after obtaining the characteristic parameters of the KPI data; and transmitting the KPI data after identifying the fault mode to the big data analysis training platform for training to obtain a new fault mode identification model, and updating the new fault mode identification model to the big data analysis online platform.

As shown in fig. 3, when the method is applied to an optical link network device, where an OLT is a part of an optical link that identifies a fault, the OLT receives the KPI data transmitted by the ONU, and transmits the KPI data to a big data analysis training platform through the big data analysis online platform to train a fault pattern identification model; the big data analysis online platform is used for receiving a fault mode recognition result reported by the OLT, transmitting feedback data to the big data analysis training platform and updating a new fault mode recognition model obtained by training into the OLT; the new fault mode identification model is trained by the big data analysis training platform according to feedback data; the OLT comprises a chip, and the chip is used for fault identification.

The detailed process is described as follows:

the ONU sends KPI data to the OLT, the OLT receives the KPI data, performs fault identification on the KPI data by using a feature mining and fault mode identification model, reports the fault mode to the big data analysis online platform, the big data analysis online platform transmits the data after the identification mode to the big data analysis training platform, the big data analysis training platform transmits a new fault mode identification model to the big data analysis online platform, and the big data analysis online platform updates a new fault mode identification model to the OLT; and the OLT and the ONU are connected through optical fibers.

And the ONU sends the KPI data to the OLT periodically, the OLT carries out feature mining or feature extraction on the KPI data, obtains the performance features of the KPI data, and carries out fault identification based on a performance feature fault mode identification model. After the diagnosis result is obtained, the OLT reports the fault mode in the diagnosis result to a big data analysis online platform. In addition, model training is carried out on the big data analysis training platform regularly, and a model is provided for the big data analysis online platform, and the big data analysis online platform updates the model to the OLT for fault diagnosis.

Based on the same technical concept, the embodiment of the present application further provides a device for implementing optical link fault identification, so as to implement the above method embodiment.

As shown in fig. 10, in the embodiment of the present application, an apparatus 1000 for implementing optical link failure identification is provided, including: an obtaining unit 1010, configured to obtain performance data at least including received optical power on a network device in an optical link; the performance data acquired by the acquiring unit 1010 is a time sequence in the preset time window; an extracting unit 1020 configured to extract a characteristic parameter indicating a change in the performance data acquired by the acquiring unit 1010 within a preset time window; an identifying unit 1030, configured to perform failure mode identification in the optical link according to the characteristic parameters extracted by the extracting unit 1020.

Wherein the performance data acquired by the acquiring unit 1010 further includes a transmission optical power and/or an optical fiber length. Wherein the feature parameters extracted by the extraction unit 1020 include: the characteristic parameters are used for representing the abnormal degree of the performance data and/or the characteristic parameters are used for representing the variation trend of the performance data.

Further, the identifying unit 1030 may include: a matching module and an output module; the matching module is used for matching the characteristic parameters extracted by the extraction unit 1020 with a fault mode identification model to determine a fault mode; the output module is used for outputting the fault mode determined by the matching module to external optical network equipment.

The apparatus for implementing optical link failure identification provided in another embodiment of the present application may further include: the training unit is used for training a fault mode identification model with characteristic parameters corresponding to a known fault mode by utilizing historical data of the known fault mode; the historical data comprises characteristic parameters and fault modes corresponding to the characteristic parameters.

The apparatus for implementing optical link failure identification provided in another embodiment of the present application may further include: a comparison unit and an addition unit; the comparison unit is used for comparing the fault mode identified by the identification unit with the actual fault mode corresponding to the characteristic parameter and outputting a comparison result to the adding unit; the adding unit is used for receiving the output result of the comparing unit and adding the characteristic parameters and the actual fault modes corresponding to the characteristic parameters into the training set when the fault modes identified by the identifying unit are inconsistent with the actual fault modes corresponding to the characteristic parameters.

The apparatus for implementing optical link failure identification provided in another embodiment of the present application may further include: the establishing unit is used for training the training set to establish a new fault mode recognition model; and the training set consists of historical data during the training of the fault pattern recognition model and the added characteristic parameters and actual fault patterns corresponding to the characteristic parameters.

Based on the same technical concept, the embodiment of the application also provides a system for realizing optical link fault identification, and the system structure can be similar to the structure of the existing optical link system and can comprise optical network equipment, an optical line terminal and a big data analysis online platform; the difference is that the apparatus for implementing optical link failure identification mentioned in the above embodiment may be disposed in the optical line terminal, such as scenario b, or may be disposed in the big data analysis online platform, such as scenario a, and the connection and corresponding related operations between the apparatus and other units have been described in the above embodiment, and are not described herein again.

It should be noted that the apparatus for implementing optical link failure identification provided in the embodiment of the present application may include an extraction unit and an identification unit implemented by a processor, where the extraction unit and the identification unit are connected with other modules. As shown in fig. 11, the processing portion of the optical link failure identification system may include a processor 111 and a memory 112; the memory 112 may be used for storing indication information, and may also be used for storing codes, instructions and the like executed by the processor 111.

The storage unit may be a memory, for example. When the network device comprises a storage unit, the storage unit is used for storing computer execution instructions, the processing unit is connected with the storage unit, and the processing unit executes the computer instructions stored by the storage unit so as to enable the optical link fault identification system to identify the characteristic parameters.

Optionally, if the optical link fault identification system is composed of a plurality of network devices; optionally, if the optical link failure identification system is a chip that collects software programs of a plurality of network devices, the chip includes:

a processing module, which may be implemented by a processor. The processing module may execute computer-executable instructions stored by the memory unit. The memory cell is a memory cell in the chip, such as: a register, a cache, and the like, and the storage unit may also be a storage unit located outside the chip in the terminal, such as a read-only memory (ROM) or another type of static storage device that can store static information and instructions, a Random Access Memory (RAM) and the like.

Optionally, the storage unit is a storage unit in the chip, such as a register, a cache, and the like, and the storage unit may also be a storage unit located outside the chip in the terminal, such as a read-only memory (ROM) or another type of static storage device that can store static information and instructions, a Random Access Memory (RAM), and the like. The storage unit may also be a storage unit located outside the chip in the terminal, such as a read-only memory (ROM) or other types of static storage devices that may store static information and instructions, a Random Access Memory (RAM), and the like.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the procedures or functions described in accordance with the present application are generated, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website, computer, server, or data center to another website, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state Disk).

Based on the same technical concept, embodiments of the present application further provide a computer-readable storage medium storing a computer program, where the computer program includes at least one piece of code, and the at least one piece of code is executable by a remote server to control a big data analysis online platform or an OLT to implement the above-described method embodiments.

Based on the same technical concept, the embodiment of the present application further provides a computer program, which is used to implement the above method embodiments when the computer program is executed by a remote server.

The program may be stored in whole or in part on a storage medium packaged with the processor, or in part or in whole on a memory not packaged with the processor.

Based on the same technical concept, the embodiment of the present application further provides a processor, and the processor is configured to implement the above method embodiment. The processor may be a chip.

Those skilled in the art will recognize that, in one or more of the examples described above, the functions described in the embodiments of the present application may be implemented in hardware, software, firmware, or any combination thereof. When implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the embodiments of the present application in further detail, and it should be understood that the above-mentioned embodiments are only specific embodiments of the present application, and are not intended to limit the scope of the embodiments of the present application, and any modifications, equivalent substitutions, improvements and the like made on the basis of the technical solutions of the embodiments of the present application should be included in the scope of the embodiments of the present application.