Abnormality factor location identification system and abnormality factor location identification method for optical transmission system

Abstract

To identify an abnormal factor location of an optical transmission system by optical signal monitoring information that can be acquired from an existing commercial system.SOLUTION: An abnormality factor location identification system 1 includes an abnormality detection part 16 that detects an error, an abnormality factor determination part 18 that holds, for each factor of an abnormality, a classification model that classifies the presence or absence of an abnormality from a plurality of types of optical signal quality data provided for each optical path, an abnormality section narrowing-down part 22 that detects a suspected section narrowed down on the basis of optical path accommodation information 20 and error information, and an abnormality location determination part 24 that determines a final abnormality location on the basis of the suspected section and the abnormality factor determined by the abnormality factor determination part 18. The abnormality detection part 16 inputs the optical signal quality data of a reception end of the optical path in which the error is detected to the abnormality factor determination part 18 provided in the optical path. The abnormality factor determination part 18 sequentially inputs the optical signal quality data to a classification model for each factor of the abnormality, and determines the factor of the abnormality from the probability indicating the presence or absence of the abnormality acquired for each classification model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a system and method for identifying an abnormality cause location in an optical transmission system. [Background technology]

[0002] In optical transmission systems, multiple optical transmission devices are interconnected by optical transmission paths, and optical physical characteristics and analog control characteristics interact in complex ways, making it difficult to identify the location and cause of a failure when one occurs. In optical transmission systems, there is a known technology for detecting errors from data indicating the quality of optical signals, such as Q-factors (hereinafter referred to as optical signal quality data). In optical transmission systems, correction is generally performed at the optical receiver using forward error correction (FEC). The Q-factor, calculated by conversion from the pre-FEC bit error rate (Pre-FEC BER), is one of the most important indicators (transmission quality indicators) for evaluating the quality of optical transmission. The Q-factor decreases due to various types of degradation of the optical signal and the influence of noise on the optical signal. If the Q-factor falls below the limit that can be corrected by FEC, packet loss and other issues will affect communication services.

[0003] FIG. 14 is a schematic diagram showing an example of optical signal quality data (Q-factor). The horizontal axis of the graph represents time, and the vertical axis represents Q-factor (dB). The Q-factor (dB) on the vertical axis is an average value over 15 minutes, and the horizontal axis ranges from several hours, several days, or even several tens of days. At times t1, t2, and t3, Q-factor data was measured, showing a significant drop from the steady-state value. This phenomenon, in which the Q-factor suddenly drops temporarily and then recovers, is called a Q-drop. It is known that a Q-drop may be a sign of a malfunction. In this example, a malfunction occurs at time t4, and then the signal falls below the correction limit (Q-limit). The possibility that the Q-drop may be a sign of a malfunction is discussed, for example, in Reference 1 below, and therefore will not be discussed here. (Reference 1) Monia Ghobadi and Ratul Mahajan, “Optical Layer Failures in a Large Backbone,” Proceedings of the 2016 Internet Measurement Conference, pp. 461-467, Santa Monica, CA, USA, Nov. 2016.

[0004] In order to quickly identify the fault location in an optical transmission system, a technique has been proposed (see Patent Document 1) that narrows down suspected locations based on packet loss information in upper layers and optical path accommodation relationships. Also, a technique has been proposed (see Patent Document 2) that narrows down suspected locations based on optical path accommodation relationships and optical signal monitor information in relay sections. Furthermore, a technique has been proposed (see Patent Document 3) that narrows down suspected locations based on optical path accommodation relationships and optical signal characteristics at the receiving end of the optical path. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2018-064160 [Patent Document 2] International Publication No. 2023 / 162187 [Patent Document 3] International Publication No. 2020 / 110787 Summary of the Invention [Problem to be solved by the invention]

[0006] The technology described in Patent Document 1 can only narrow down the detection to the OMS (Optical Multiplex Section) section where the accommodation relationship changes, and cannot identify the fault location. The technology described in Patent Document 2 can identify the fault location, but requires optical signal monitor information that cannot be obtained from existing commercial systems, resulting in the introduction cost of adding monitors. The technology described in Patent Document 3 uses only optical signal monitor information (optical signal quality data) that can be obtained from existing systems, but is limited to extracting optical physical characteristics that deteriorate simultaneously with optical signal quality and is unable to identify suspected fault locations. Therefore, these conventional technologies have room for improvement.

[0007] The present invention has been made in view of the above circumstances, and an object of the present invention is to identify the location of an abnormality in an optical transmission system using optical signal monitor information that can be acquired from an existing commercial system. [Means for solving the problem]

[0008] The system for identifying the location of an abnormality factor in an optical transmission system according to the present invention comprises an abnormality detection unit that detects errors based on optical signal quality data at the receiving end of an optical path; an abnormality factor determination unit that is provided for each optical path and holds, for each abnormality factor, a classification model that classifies the presence or absence of an abnormality from multiple types of optical signal quality data, and determines the cause of the abnormality based on the input optical signal quality data; an abnormality section narrowing unit that detects a suspected section by narrowing down the abnormal section of the optical path based on optical path accommodation information and detected error information; and an abnormality location determination unit that determines the final abnormality location based on the suspected section and the abnormality factor determined by the abnormality factor determination unit, wherein the abnormality detection unit inputs optical signal quality data at the receiving end of the optical path in which an error has been detected to an abnormality factor determination unit provided for each optical path, and the abnormality factor determination unit sequentially inputs the input optical signal quality data to the classification models for each abnormality factor and determines the cause of the abnormality from the probability indicating the presence or absence of an abnormality obtained for each classification model. [Effects of the Invention]

[0009] According to the present invention, it is possible to identify the location of an abnormality in an optical transmission system using only optical signal monitor information that can generally be obtained from existing commercial systems. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a schematic configuration diagram of an abnormality cause location identification system according to an embodiment of the present invention; [Figure 2] 1 is a diagram illustrating an example of the arrangement of each part of an abnormality cause location identification system according to an embodiment of the present invention; [Figure 3] FIG. 1 is a diagram illustrating an example of the arrangement of multiple optical paths and receiving end transponders. [Figure 4] FIG. 2 is a block diagram showing an example of the configuration of a transponder arranged at the receiving end of each optical path. [Figure 5] FIG. 1 is a conceptual diagram of a classification model. [Figure 6] 10 is a flowchart showing a flow of processing by the abnormality cause location identification system according to the present embodiment. [Figure 7] 7 is a flowchart showing the flow of the abnormality cause determination process of FIG. 6. [Figure 8] FIG. 1 illustrates an example of the configuration of an abnormality cause location identification system. [Figure 9] FIG. 10 is an explanatory diagram of a process for narrowing down an abnormal section. [Figure 10] 9 is an explanatory diagram of a suspected section of a fault estimated in the abnormality cause localization system of FIG. 8. FIG. [Figure 11] FIG. 9 is an explanatory diagram of a cause of a failure identified in the abnormality cause location identifying system of FIG. 8. [Figure 12] FIG. 9 is an explanatory diagram of a classification model used in the abnormality cause location identifying system of FIG. 8. [Figure 13] 2 is a hardware configuration diagram illustrating an example of a computer that realizes the functions of each unit of the abnormality cause location identification system according to the present embodiment. FIG. [Figure 14] FIG. 10 is a schematic diagram showing an example of optical signal quality data (Q value). DETAILED DESCRIPTION OF THE INVENTION

[0011] The abnormality cause location identifying system according to this embodiment will be described in detail below with reference to the drawings. [Configuration overview of the system for identifying abnormality causes] An abnormality cause location identification system 1 shown in FIG. 1 is a system for identifying an abnormality cause location of an optical path (optical wavelength path) in an optical transmission system 6, as an example.

[0012] In this example, the optical transmission system 6 is configured by a plurality of nodes 5i to 5k interconnected by optical fiber 7, which is an optical transmission path, and further includes node controllers 3i to 3k and a network controller 2. Transponders 4i and 4k are denoted as "TRPD" in each figure. The transponder 4i included in the node 5i and the transponder 4k included in the node 5k are communicatively connected by an optical path LP. The optical path LP is a logical communication path connecting the transponder 4i and the transponder 4k via the optical fiber 7, which is a physical communication path, and the node 5j. Hereinafter, when there is no need to distinguish between the nodes 5i to 5k, they will simply be referred to as node 5. When there is no need to distinguish between the transponders 4i and 4k, they will simply be referred to as transponder 4.

[0013] The node controller 3i controls and monitors the transponder 4i. The node controller 3j controls and monitors the node 5j. The node controller 3k controls and monitors the transponder 4k. Hereinafter, when there is no need to distinguish between the node controllers 3i to 3k, they will simply be referred to as the node controllers 3. The network controller 2 is connected to the node controllers 3i to 3k via a data communication network, and controls and monitors the node controllers 3i to 3k.

[0014] The abnormality cause location identification system 1 includes an abnormality detection unit 16, an abnormality cause determination unit 18, an abnormal section narrowing unit 22, and an abnormality location determination unit 24. The abnormality detection unit 16 detects an error based on optical signal quality data at the receiving end of an optical path LP (Light path). The abnormality factor determination unit 18 is provided for each optical path LP, and holds a classification model 8 for each abnormality factor that classifies the presence or absence of an abnormality from multiple types of optical signal quality data, and determines the abnormality factor based on the input optical signal quality data.

[0015] The abnormal section narrowing unit 22 detects a suspect section by narrowing down the abnormal section of the optical path LP based on the optical path accommodation information 20 and the detected error information. The abnormality location determining unit 24 determines the final abnormality location based on the suspected section and the abnormality cause determined by the abnormality cause determining unit 18.

[0016] At this time, the abnormality detection unit 16 inputs optical signal quality data of the receiving end of the optical path in which the error was detected to the abnormality factor determination unit 18 provided for each optical path. The abnormality factor determination unit 18 sequentially inputs the input optical signal quality data to classification models 8 for each abnormality factor, and determines the factor of the abnormality from the probability indicating the presence or absence of an abnormality obtained for each classification model 8.

[0017] In detail, the anomaly detection unit 16 is disposed in any of the transponder 4, the node 5, the node controller 3, or the network controller 2. The anomaly detection unit 16 may be disposed either internally or externally to the device. The anomaly detection unit 16 detects errors using optical signal quality data such as the Q factor, pre-FEC BER, and error second. The error second is the number of seconds in which an error occurs. The anomaly detection unit 16 detects an error, for example, when the pre-FEC BER deteriorates. Alternatively, the anomaly detection unit 16 detects an error when the count value of the error second exceeds a threshold.

[0018] The abnormality factor determination unit 18 is placed in either the transponder 4, the node 5, the node controller 3, or the network controller 2. The abnormality factor determination unit 18 may be placed inside or outside the device. The abnormality factor determination unit 18 determines the abnormality factor using the monitor value of the optical signal quality data of the transponder 4 at the receiving end.

[0019] The monitored values ​​of optical signal quality data can include Q-factor, optical reception power, error seconds, polarization-mode dispersion (PMD), polarization-dependent loss (PDL), chromatic dispersion, etc. Polarization mode dispersion (PMD) is a physical phenomenon that occurs when the speed of light differs between two orthogonal axes during transmission, causing distortion in the optical signal. PMD is monitored as the amount of PM compensation. The polarization dependent loss is monitored as the OSNR difference between X and Y. Chromatic dispersion is a physical phenomenon that occurs when different wavelengths travel at different speeds during transmission of an optical signal, causing distortion in the optical signal. Chromatic dispersion is monitored as the amount of chromatic dispersion compensation.

[0020] The abnormality factor determination unit 18 performs the determination using, for example, multiple 1vsREST (One-Virus-Rest) models. 1vsREST is a classification model in which the same number of classification models as the number of classes are prepared and whether or not a signal falls into a certain class is determined. One abnormality factor determination unit 18 is provided for each optical path LP.

[0021] The abnormal section narrowing unit 22 is arranged on the network controller 2 side. The abnormal section narrowing unit 22 narrows down the abnormal section using optical path accommodation information held by the network controller 2. Although details will be described later, the abnormal section narrowing unit 22 can narrow down the abnormal section using, for example, a routing matrix and an alarm vector.

[0022] The abnormality location determination unit 24 is arranged on the side of the network controller 2. The abnormality location determination unit 24 determines a final abnormality location based on the abnormal section detected by the abnormal section narrowing unit 22 and the abnormality factor determined by the abnormality factor determination unit 18, and presents the result to the user.

[0023] [Configuration of an abnormality location identification system incorporated into an optical transmission system] As an example, the abnormality cause location identification system 1B shown in FIG. 2 includes an abnormality detection unit 16 and an abnormality cause determination unit 18 in the transponder 4k, and an abnormal section narrowing unit 22 and an abnormality location determination unit 24 in the network controller 2. Furthermore, the terminal 21 is connected to the network controller 2 and notifies the user of, for example, the location and cause of the failure. The abnormality location determining unit 24 may be arranged in the terminal 21, so that the terminal 21 determines the final abnormality location and presents it to the user.

[0024] 1 and 2 show an example of one optical path LP, but multiple optical paths LP can be set as shown in Fig. 3. Note that the network controller 2 is not shown. Here, an outline of the network configuration of the fault cause localization system 1C shown in FIG. 3 will be described. The fault cause localization system 1C includes a plurality of nodes 5a, 5b, 5c, 5d, 5e, and 5f, each of which is formed, for example, by an optical cross-connect device or a multiplexer / demultiplexer. Node 5a multiplexes optical signals from transponders 4a and 4b. Node 5a and node 5b are connected by optical fiber 7a. Node 5b is connected to node 5c by optical fiber 7b, and branches off and is connected to node 5e by optical fiber 7c. Node 5c is connected to node 5d by optical fiber 7d, and branches off and is connected to node 5e by optical fiber 7e. Node 5d demultiplexes the optical signal to transponders 4c and 4d. Node 5e is connected to node 5f by optical fiber 7f. Node 5f demultiplexes the optical signal to transponders 4e and 4f.

[0025] The transponder 4a is the transmitting end of the optical path LP1, and the receiving end of this optical path LP1 is the transponder 4c. The transponder 4b is the transmitting end of the optical path LP2, and the receiving end of this optical path LP2 is the transponder 4f. The transponder 4d is the transmitting end of the optical path LP3, and the receiving end of this optical path LP3 is the transponder 4e.

[0026] In the abnormality cause location identification system 1C, an abnormality detection unit 16 and an abnormality cause determination unit 18 are provided corresponding to each of the optical paths LP1, LP2, and LP3. In addition, an abnormality detection unit 16 and an abnormality cause determination unit 18 are provided for each of the transponders 4c, 4e, and 4f. In Fig. 3, an alarm is illustrated for the transponders 4e and 4f.

[0027] As shown in FIG. 4, the transponder 4e disposed at the receiving end of the optical path LP3 includes, for example, an optical signal quality data acquiring unit 14e, an abnormality detecting unit 16e, and an abnormality factor determining unit 18e. Furthermore, as shown in FIG. 4, the transponder 4f arranged at the receiving end of the optical path LP2 also includes, for example, an optical signal quality data acquiring unit 14f, an abnormality detecting unit 16f, and an abnormality factor determining unit 18f. When the optical signal quality data acquiring units 14e and 14f are not distinguished from each other, they are simply referred to as the optical signal quality data acquiring unit 14. When the abnormality detecting units 16e and 16f are not distinguished from each other, they are simply referred to as the abnormality detecting unit 16. When the abnormality factor determining units 18e and 18f are not distinguished from each other, they are simply referred to as the abnormality factor determining unit 18. The optical signal quality data acquiring unit 14 acquires optical signal quality data about the optical signal received by the transponder 4. The types of optical signal quality data are as described above. The optical signal quality data is acquired from time-series data of the optical signal received over a predetermined period of time (for example, 15 minutes).

[0028] As an example, the abnormality factor determination unit 18e is equipped with three classification models 8A, 8B, and 8C. As an example, the abnormality factor determination unit 18f is equipped with three classification models 8A, 8B, and 8D. When there is no need to distinguish between the classification models 8A, 8B, 8C, and 8D, they are simply referred to as classification models 8. Each abnormality factor determination unit 18 may be equipped with the same classification model 8 or different classification models 8.

[0029] 3, in the conventional technology, it is not possible to determine whether abnormality A has occurred in optical fiber 7f, or whether abnormality B has occurred in node 5f, or whether abnormality C has occurred in node 5e. In contrast, in this embodiment, the abnormality factor determination unit 18e of optical path LP2 and the abnormality factor determination unit 18f of optical path LP3 determine the abnormality factor, thereby making it possible to identify the location of the abnormality factor.

[0030] Next, an example of a determination method performed by the abnormality factor determination unit 18 will be described with reference to FIGS. 4 and 5 (and also with reference to FIG. 3 as needed). Classification model 8A (Fig. 4(a)) for abnormality A classifies abnormality A from other abnormalities (rest) using the first monitor value (monitor value a1) as the minimum optical reception power value and the second monitor value (monitor value a2) as the Q value. In Fig. 5(a), abnormality A classified in advance training is indicated by ◎, and other abnormalities (rest) are indicated by ●. Note that ☆ indicates test data to be classified.

[0031] Classification model 8B (Fig. 4(a)) for abnormality B classifies abnormality B from other abnormalities (rest) using the first monitor value (monitor value b1) as the error second and the second monitor value (monitor value b2) as the Q value. In Fig. 5(b), abnormality B classified in advance training is indicated by a circle, and other abnormalities (rest) are indicated by a black circle. Note that a star indicates the test data to be classified.

[0032] Classification model 8C for anomaly C classifies anomaly C from other anomalies (rest) using specific monitor values ​​c1 and c2. In Figure 5(c), pre-classified anomalies C are indicated by ◇, and other anomalies (rest) are indicated by ●. Note that ☆ indicates data to be classified.

[0033] Each classification model 8 is trained in advance using data on a certain type of anomaly and other data. If only one classification model 8 among multiple classification models 8 outputs an anomaly factor, that output is used to determine the anomaly. Also, if multiple classification models 8 output an anomaly factor, the output with the highest likelihood is used to determine the anomaly. If no classification model 8 outputs an anomaly factor, it is assumed to be another anomaly factor.

[0034] For example, in the case of logistic regression, the output is a probability, and a probability greater than 0.5 is determined to be an abnormality. Furthermore, among classification models with outputs greater than 0.5, the abnormality of the model with the highest output probability is output. For example, if the abnormality factor determination unit 18 obtains an output probability of "0.8" for classification model 8A, an output probability of "0.3" for classification model 8B, and an output probability of "0.6" for classification model 8C, it determines that the abnormality in the suspected section is "Abnormality A." By using the method of this embodiment instead of multi-class classification, even when an unknown abnormality for which no classification model is available occurs, it is correctly determined to be "other abnormality factors," thereby reducing the likelihood of misclassification as a different abnormality factor that is not the correct answer.

[0035] [Operation of the system for identifying the cause of anomalies] Next, the operation of the abnormality cause location identification system 1 will be described with reference to FIG. 6 (and also with reference to FIGS. 2 to 4 as appropriate). FIG. 6 is a flowchart showing the flow of processing by the abnormality cause location identification system 1B of FIG. 2. In step S1, the optical signal quality data acquisition unit 14 of the transponder 4 arranged at the receiving end monitors the optical signal quality data. In step S2, the abnormality detection unit 16 determines whether or not an error has been detected (step S2). If no error is detected (No in step S2), the process returns to step S1.

[0036] On the other hand, if an error is detected (Yes in step S2: step of detecting an error), then the abnormality detection unit 16 notifies the abnormality cause determination unit 18 of information about the error, and inputs the monitor value of the optical signal quality data to the abnormality cause determination unit 18 (step S3: input step). Also, if an error is detected, then the abnormality detection unit 16 notifies the abnormal section narrowing down unit 22 of information about the error (step S4). Following step S4, the abnormal section narrowing down unit 22 narrows down the abnormal section based on the optical path accommodation information, and detects a suspected section (step S5: step of detecting a suspected section). Note that the processing order of steps S3 to S4 is arbitrary, and they can also be processed in parallel.

[0037] On the other hand, following step S3, the abnormality factor determination unit 18 determines the abnormality factor (step S6: determination step). Then, the abnormality factor determination unit 18 notifies the abnormality factor determination result to the abnormality location determination unit 24 (step S7). Following steps S5 and S7, the abnormality location determination unit 24 determines the abnormality factor location from the suspected section and the abnormality factor determination result (step S8: determination step). Note that the processing order of steps S5 to S6 is arbitrary.

[0038] Next, the abnormality factor determination in step S6 will be described with reference to FIG. The abnormality factor determination unit 18 sequentially inputs optical signal quality data to the 1 vs. REST classification models 8 provided for each abnormality factor (step S61). Then, the abnormality factor determination unit 18 determines whether one or more classification models 8 have determined an abnormality (step S62). If one or more classification models have determined an abnormality (step S62: Yes), the abnormality factor determination unit 18 determines the abnormality of the classification model 8 with the highest output probability (likelihood) as the determination result (step S63). On the other hand, if none of the classification models 8 have determined an abnormality (step S62: No), the abnormality factor determination unit 18 determines the abnormality as the determination result of "other abnormality" (step S64).

[0039] Note that the abnormal section narrowing unit 22 may notify the abnormal section detected in step S5 to the abnormal section determination unit 18 before the abnormal section determination unit 18 executes the abnormal section determination process. In this way, the abnormal section determination unit 18 can exclude abnormal sections that cannot occur in the notified suspect section and use only the classification model 8 of abnormal sections that need to be considered, thereby reducing the load of the determination process and enabling the determination result to be output quickly.

[0040] [Verification experiment] Next, a verification experiment carried out to confirm the effect of the abnormality cause location identifying system of this embodiment will be described with reference to FIGS. (Verification experiment system configuration) A verification experiment was conducted on the anomaly cause localization system 1C shown in Figure 3. However, the following devices were set up in station building 1 and station building 2, and station building 2 and station building 3, which are connected to each other by a long-distance (50 km) optical fiber. This system is referred to as an anomaly cause localization system 1D. Note that although the network controller 2 is not shown in Figure 8, the network controller 2 is equipped with an anomaly section narrowing unit 22 and an anomaly location determining unit 24. Furthermore, the transponder 4 at the receiving end is equipped with an optical signal quality data acquiring unit 14, an anomaly detecting unit 16, and an anomaly cause determining unit 18.

[0041] In station building 1, transponders 4a, 4b, and 4c, nodes 5a, and 5b are connected by optical fiber 7. Node 5a includes a multiplexer / demultiplexer 31. Three inputs of this multiplexer / demultiplexer 31 are connected to the transponders 4a, 4b, and 4c. Node 5b includes a wavelength selective switch (WSS) 32 and a booster amplifier 33 installed immediately after it. An input of wavelength selective switch 32 is connected to the output of multiplexer / demultiplexer 31 of node 5a.

[0042] In the exchange building 2, nodes 5c, 5d, and transponders 4d and 4e are connected by optical fiber 7. Node 5c includes a wavelength selective switch 32 and a preamplifier 34 installed immediately before the wavelength selective switch 32. The input of the preamplifier 34 in node 5c is connected via optical fiber to a booster amplifier 33 in node 5b in the exchange building 1. One output of the wavelength selective switch 32 in node 5c is connected to transponder 4d.

[0043] Node 5d includes a wavelength selective switch 32 and a booster amplifier 33 installed immediately after the wavelength selective switch 32. One input of the wavelength selective switch 32 of node 5d is connected to transponder 4e. The other input of the wavelength selective switch 32 of node 5d is connected to the other output of the wavelength selective switch 32 of node 5c.

[0044] In the exchange building 3, node 5e, node 5f, and transponders 4f, 4g, and 4h are connected by optical fiber 7. Node 5e includes a wavelength selective switch 32 and a preamplifier 34 installed immediately before it. The input of the preamplifier 34 of node 5e is connected via optical fiber to a booster amplifier 33 of node 5d in the exchange building 2. The output of the wavelength selective switch 32 of node 5e is connected to the input of a multiplexer / demultiplexer 35 included in node 5f. The three outputs of the multiplexer / demultiplexer 35 are connected to each of the transponders 4f, 4g, and 4h.

[0045] The transponder 4a is the transmitting end of the optical path LP1, and the receiving end of this optical path LP1 is the transponder 4f. Therefore, in the drawing, "TRPD1" is attached to each of the transponders 4a and 4f to indicate that they are the transmitting and receiving ends of the optical path LP1. The transponder 4b is the transmitting end of the optical path LP2, and the receiving end of this optical path LP2 is the transponder 4d. Therefore, in the drawing, "TRPD2" is attached to each of the transponders 4b and 4d to indicate that they are the transmitting and receiving ends of the optical path LP2. The transponder 4c is the transmitting end of the optical path LP3, and the receiving end of this optical path LP3 is the transponder 4g. Therefore, in the drawings, "TRPD3" is attached to each of the transponders 4c and 4g to indicate that they are the transmitting and receiving ends of the optical path LP3. The transponder 4e is the transmitting end of the optical path LP4, and the receiving end of this optical path LP4 is the transponder 4h. Therefore, in the drawings, "TRPD4" is attached to each of the transponders 4e and 4h to indicate that they are the transmitting and receiving ends of the optical path LP4.

[0046] Here, Sections I to VII indicate the sections of the link relaying between two transmission devices. Section I indicates the link connecting the transponder 4a and the multiplexer / demultiplexer 31 in the central office building 1. This is set to detect optical power abnormalities in the transponder 4a at the transmitting end. Section II indicates the link connecting the multiplexer / demultiplexer 31 and the wavelength selective switch 32 in the central office building 1.

[0047] Section III straddles central office building 1 and central office building 2. This section is set up in central office building 1 to detect, for example, fluctuations in the polarization state in the optical fiber from central office building 1 to central office building 2. Section IV shows the link connecting the wavelength selective switch 32 of node 5d with the wavelength selective switch 32 of node 5c in central office building 2. This section is set up to detect, for example, an optical power abnormality in the wavelength selective switch 32.

[0048] Section V straddles station building 2 and station building 3. This is set up in station building 2 to detect, for example, fluctuations in the polarization state in the optical fiber from station building 2 to station building 3. Section VI indicates, in station building 3, the link connecting wavelength selective switch 32 in node 5e and multiplexer / demultiplexer 35 in node 5f. This is set up to detect, for example, an optical power abnormality in wavelength selective switch 32 or preamplifier 34.

[0049] Section VII indicates a link connecting the multiplexer / demultiplexer 35 and transponder 4f in the central office building 3. Note that sections such as those between the transponder 4b and the multiplexer / demultiplexer 31 and those between the transponder 4c and the multiplexer / demultiplexer 31 in the central office building 1 were omitted in this verification experiment, but in reality such sections are assigned.

[0050] Among optical transmission devices, active devices such as transponders, booster amplifiers, and wavelength selective switches are more susceptible to abnormalities and failures than passive devices such as multiplexers and demultiplexers, so we simulated optical power abnormalities in these devices. Optical power abnormalities were simulated by inserting a variable optical attenuator immediately after each of these active devices. Fluctuations in the polarization state were simulated by inserting polarization scramblers into the optical fiber section between station building 1 and station building 2, and into the optical fiber section between station building 2 and station building 3. The fluctuation rate of the polarization state was set to 20,000 krad / s. These abnormalities were simulated for 5 to 10 seconds. Optical power abnormalities and fluctuations in the polarization state cause Q-drops.

[0051] (Conditions and results of the verification experiment) The routing matrix shown in Fig. 9(a) was generated from the experimental system shown in Fig. 8. In the routing matrix, "1" indicates that each lightpath passes through the corresponding section, and "0" indicates that it does not pass through. The routing matrix represents the relationship between lightpaths and sections in matrix form.

[0052] An alarm vector is an aggregation of alarms from multiple optical paths. The alarm vector shown in Fig. 9(b) was generated from the experimental system shown in Fig. 8. In Fig. 8, alarms are shown for transponders 4f, 4g, and 4h. Transponders 4f, 4g, and 4h are the receiving ends of optical paths LP1, LP3, and LP4, respectively. In other words, alarms notifying of abnormalities in optical paths LP1, LP3, and LP4 are notified to the network controller. In the alarm vector shown in Fig. 9(b), "1" indicates that an alarm has been notified from the optical path, and "0" indicates that there is no alarm.

[0053] In this example, when an alarm is detected in LP1, LP3, or LP4 as shown in the alarm vector in FIG. 9(b), the abnormal section narrowing unit 22 narrows down the abnormal section from all sections (section I to section VII) based on the routing matrix shown in FIG. 9(a). As a result, as shown by the virtual lines in FIG. 10, section V or section VI, which are common sections for each alarm, becomes the suspect section. In this way, using the routing matrix makes it possible to narrow down the sections to a certain extent. However, it is not possible to uniquely narrow down the sections using only the routing matrix.

[0054] In contrast, in the abnormality cause location identification system 1D, not only is the search narrowed down to section V or section VI, but the abnormality cause determination unit 18 of optical path LP1, the abnormality cause determination unit 18 of optical path LP3, and the abnormality cause determination unit 18 of optical path LP4 (hereinafter referred to as the abnormality cause determination unit 18 of optical path LP1, etc.) input the monitor values ​​of the measured optical signal quality data (hereinafter referred to as monitor values) into a classification model 8 for each abnormality cause.

[0055] At this time, for section V, the monitor value is input to a classification model 8 prepared for each abnormality that may occur in section V, and similarly, for section VI, the monitor value is input to a classification model 8 prepared for each abnormality that may occur in section VI. For example, in section V, since it is predicted that an abnormality in the fluctuation of the polarization state will occur, the abnormality factor determination unit 18 of the optical path LP1, etc., excludes abnormality factors that do not need to be considered in this section V and uses a classification model of abnormality factors that need to be considered. For example, in section VI, since it is predicted that an optical power abnormality will occur, the abnormality factor determination unit 18 of optical path LP1 etc. excludes abnormality factors that do not need to be considered in this section VI and uses a classification model of abnormality factors that need to be considered.

[0056] In fact, in section V, as shown in Figure 11, an abnormality in the fluctuation of the polarization state (abnormality A) occurred, and the abnormality cause determination unit 18 of optical path LP1, etc., was able to determine that the cause of the abnormality in this section V was abnormality A using the classification model 8A for abnormality A. 11, an optical power abnormality (abnormality B) occurred in section VI, and the abnormality cause determination unit 18 of the optical path LP1 and the like were able to determine that the cause of the abnormality in this section VI was abnormality B using classification model 8B for abnormality B. In other words, it was confirmed that classification model 8 can correctly classify.

[0057] Classification model 8A (FIG. 4(a)) for anomaly A classifies anomaly A from other anomalies (rest) using the first monitor value (vertical axis) as the minimum optical reception power (dB) and the second monitor value (horizontal axis) as the Q value (dB). Fig. 12(a) shows the results of an actual experiment, indicating that when the minimum optical reception power is greater than -53.5 dB, the cause of the anomaly is determined to be anomaly A, and when it is smaller, the cause of the anomaly is determined to be other anomaly (rest).

[0058] Classification model 8B (Fig. 4(b)) for anomaly B classifies anomaly B from other anomalies (rest) using the first monitor value (vertical axis) as the error second (s) and the second monitor value (horizontal axis) as the Q-factor (dB). Fig. 12(b) shows the results of an actual experiment, which shows that if the error second value is larger than E calculated by the following equation (1) using the Q-factor (Q-factor), the cause of the anomaly is determined to be anomaly B, and if it is smaller, the cause of the anomaly is determined to be other anomalies (rest).

[0059] E=0.01×Q-factor+20.1 … (1)

[0060] The present invention is not limited to the above-described embodiments, and many modifications can be made by a person skilled in the art within the technical concept of the present invention. For example, when multiple optical paths LP pass through the suspected section, the abnormality location determination unit 24 (see FIG. 2) may make a comprehensive determination using the determination results of multiple abnormality factor determination units 18. For example, the abnormality factor location identification system 1D has an abnormality factor determination unit 18 for each optical path, and in this case, the abnormality location determination unit 24 outputs a final determination result based on, for example, three determination results of the abnormality factor determination units 18 for each of the transponders 4f, 4g, and 4h (see FIG. 8) in which an error has been detected. The abnormality location determination unit 24 makes a determination, for example, as follows, using the determination results of the multiple abnormality factor determination units 18.

[0061] First example: The abnormality location determining unit 24 determines the cause of the abnormality in the suspected section by majority vote based on the cause of the abnormality determined by the abnormality cause determining unit 18 for each optical path LP. For example, if the transponders 4f and 4g determine that the cause of an abnormality in a certain suspected section is abnormality A, and the transponder 4h determines that the abnormality in the same suspected section is abnormality B, the abnormality location determination unit 24 determines that the cause of the abnormality in the suspected section is abnormality A. This improves the accuracy of determining the cause of an abnormality when the determination is unclear.

[0062] Second example: If the abnormality location determining unit 24 determines that at least one of the plurality of abnormality factor determining units 18 has detected a predetermined special abnormality that may occur in a predetermined section of the optical path, it outputs the abnormality. For example, if the cause of the anomaly in a certain suspected section of even one of the transponders 4f, 4g, and 4h in which an error has been detected is determined to be anomaly C, the anomaly location determination unit 24 determines that the cause of the anomaly in that suspected section is anomaly C. This prevents misjudgment due to the monitor not reacting and the monitor value not changing due to the short duration of the anomaly, etc. For example, this is suitable for anomalies such as polarization state fluctuations (SOP fluctuations), which are momentary.

[0063] Furthermore, in the above embodiment, the classification model uses two variables (e.g., monitor values ​​a1 and a2) as monitor values, but is not limited to two dimensions, and may be multidimensional using more variables, such as monitor values ​​a1, a2, a3, a4, .... Furthermore, in the above embodiment, a simple classification model using logistic regression is used, but a complex classification model such as random forest or deep learning may also be used. Furthermore, in the above embodiment, the abnormality detection unit 16 is provided in the transponder 4 which is the end point of the optical path, but the abnormality detection unit 16 may also be provided in the network controller 2.

[0064] [Hardware configuration] Each unit of the abnormality cause location identification system 1 according to the embodiment is realized by, for example, a computer 900 configured as shown in Fig. 13. Fig. 13 is a hardware configuration diagram showing an example of the computer 900 that realizes the functions of each unit of the abnormality cause location identification system 1 according to the present embodiment. The computer 900 has a CPU (Central Processing Unit) 901, a ROM (Read Only Memory) 902, a RAM (Random Access Memory) 903, an HDD (Hard Disk Drive) 904, an input / output I / F (Interface) 905, a communication I / F 906, and a media I / F 907.

[0065] The CPU 901 operates based on a program stored in the ROM 902 or the HDD 904. The ROM 902 stores a boot program executed by the CPU 901 when the computer 900 is started up, programs related to the hardware of the computer 900, and the like.

[0066] The CPU 901 controls an input device 910 such as a mouse or keyboard, and an output device 911 such as a display or printer, via an input / output I / F 905. The CPU 901 acquires data from the input device 910 via the input / output I / F 905, and outputs generated data to the output device 911. Note that a GPU (Graphics Processing Unit) or the like may be used as a processor together with the CPU 901.

[0067] The HDD 904 stores programs executed by the CPU 901 and data used by the programs. The communication I / F 906 receives data from other devices via a communication network (NW) 920 and outputs the data to the CPU 901, and also transmits data generated by the CPU 901 to other devices via the communication network 920.

[0068] The media I / F 907 reads a program or data stored in the recording medium 912 and outputs it to the CPU 901 via the RAM 903. The CPU 901 loads a program related to a target process from the recording medium 912 onto the RAM 903 via the media I / F 907, and executes the loaded program. The recording medium 912 is an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase Change Rewritable Disc), a magneto-optical recording medium such as an MO (Magneto Optical Disk), a magnetic recording medium, or a semiconductor memory.

[0069] For example, when the computer 900 functions as each part of the abnormality cause location identification system 1 according to the embodiment, the CPU 901 executes a program loaded onto the RAM 903 to realize the function of each part of the abnormality cause location identification system 1. Furthermore, the HDD 904 stores data in the RAM 903. The CPU 901 reads and executes a program related to a target process from the recording medium 912. In addition, the CPU 901 can also read a program related to a target process from another device via the communication network 920.

[0070] [effect] As described above, the system 1 for identifying an abnormality cause location in an optical transmission system comprises an abnormality detection unit 16 that detects errors based on optical signal quality data at the receiving end of the optical path LP; an abnormality cause determination unit 18 that is provided for each optical path LP and holds, for each abnormality cause, a classification model 8 that classifies the presence or absence of an abnormality from multiple types of optical signal quality data, and determines the cause of the abnormality based on the input optical signal quality data; an abnormality section narrowing unit 22 that detects a suspected section by narrowing down the abnormal section of the optical path LP based on optical path accommodation information 20 and detected error information; and an abnormality location determination unit 24 that determines the final abnormality location based on the suspected section and the abnormality cause determined by the abnormality cause determination unit 18. The abnormality detection unit 16 inputs optical signal quality data at the receiving end of the optical path in which an error is detected to the abnormality cause determination unit 18 provided for each optical path, and the abnormality cause determination unit 18 sequentially inputs the input optical signal quality data to the classification models 8 for each abnormality cause and determines the cause of the abnormality from the probability indicating the presence or absence of an abnormality obtained for each classification model 8.

[0071] In this manner, in the anomaly cause localization system 1, the anomaly cause determination unit 18 holds a classification model 8 that classifies the presence or absence of an anomaly based on multiple types of optical signal quality data. A classification model 8 is constructed in advance for each anomaly cause, and an anomaly cause determination unit 18 holding multiple classification models 8 is provided for each optical path LP. In addition, in the anomaly cause localization system 1, the anomaly section narrowing unit 22 detects a suspected section by narrowing down the anomaly section of the optical path LP based on the optical path accommodation information 20 and the detected error information. This allows the anomaly cause localization system 1 to determine the anomaly cause in the suspected section using optical signal quality data measured at the receiving end of the optical path. The anomaly cause localization system 1 narrows down the suspected section and performs an anomaly determination for each optical path passing through that section using the 1 vs. REST classification model for each anomaly. Therefore, the anomaly cause localization system 1 can identify the anomaly cause in the optical transmission system using optical signal quality data, i.e., optical signal monitor information that can be acquired from existing commercial systems. Furthermore, even if an unknown abnormality occurs, it is possible to prevent the occurrence of erroneous classification into a different abnormality cause that is not the correct answer.

[0072] In the system 1 for identifying the location of abnormalities in an optical transmission system, the abnormality location determination unit 24 is characterized in that when multiple optical paths LP pass through a suspected section, the abnormality location determination unit 24 determines the cause of the abnormality in the suspected section by majority vote based on the cause of the abnormality determined by the abnormality cause determination unit 18 for each optical path LP.

[0073] In this way, the abnormality cause location identification system 1 can improve the accuracy of determining the abnormality cause for an abnormality in which the determination is ambiguous for a certain suspected section.

[0074] In the system 1 for identifying the location of abnormalities in an optical transmission system, the abnormality location determination unit 24 is characterized in that, based on a predetermined special abnormality that can occur in a specified section of an optical path, when the specified section becomes a suspect section and multiple optical paths LP pass through the suspect section, if any one of the abnormality cause determination units 18 for each optical path LP determines that the cause of the abnormality is a special abnormality, the abnormality location determination unit 24 determines that the cause of the abnormality in the suspect section is a special abnormality.

[0075] By doing this, the abnormality cause location identification system 1 can prevent erroneous judgments caused by the abnormality detection unit 16 not reacting and the monitor value of the optical signal quality data not changing due to reasons such as the abnormality occurring for a short period of time. [Explanation of symbols]

[0076] 1, 1B, 1C, 1D Abnormality cause location identification system 2 Network Controller 3 Node Controller 4 Transponder (TRPD) 5 nodes 6 Optical Transmission Systems 7. Optical Fiber 8 Classification Models 14 Optical signal quality data acquisition unit 16 Abnormality detection unit 18 Abnormality factor determination unit 20 Optical path accommodation information 21 terminals 22 Abnormal section narrowing down section 24 Abnormal location determination section 31 Multiplexer / demultiplexer 32 Wavelength Selective Switch 33 Booster Amplifier 34 Preamp 35 Multiplexer / demultiplexer A,B,C abnormality LP Optical Path

Claims

1. an abnormality detection unit that detects an error based on optical signal quality data at a receiving end of the optical path; an abnormality factor determination unit provided for each optical path, which holds a classification model for classifying the presence or absence of an abnormality based on multiple types of optical signal quality data for each abnormality factor, and determines the abnormality factor based on the input optical signal quality data; an abnormal section narrowing-down unit that detects a suspected section by narrowing down an abnormal section of the optical path based on the optical path accommodation information and the detected error information; an abnormality location determination unit that determines a final abnormality location based on the suspected section and the abnormality cause determined by the abnormality cause determination unit, the anomaly detection unit inputs optical signal quality data of a receiving end of an optical path in which an error has been detected to an anomaly cause determination unit provided for each optical path; The abnormality factor determination unit sequentially inputs input optical signal quality data to the classification models for each abnormality factor, and determines the cause of the abnormality from the probability indicating the presence or absence of an abnormality obtained for each classification model.

2. The abnormality location determination unit determines the cause of the abnormality in the suspected section by majority vote based on the cause of the abnormality determined by the abnormality cause determination unit for each optical path when multiple optical paths pass through the suspected section.

3. The abnormality location determination unit determines that the cause of the abnormality in the suspect section is a special abnormality when, based on a predetermined special abnormality that may occur in a specified section of the optical path, the specified section becomes a suspect section and when multiple optical paths pass through the suspect section, any one of the abnormality cause determination units for each optical path determines that the cause of the abnormality is the special abnormality.This is a system for identifying the cause of the abnormality in an optical transmission system as described in claim 1.

4. A method for identifying an abnormality cause location by an abnormality cause location identification system for an optical transmission system, comprising: The abnormality cause location identification system includes: an abnormality factor determination unit for each optical path, which is provided with optical path accommodation information and holds a classification model for classifying the presence or absence of an abnormality from multiple types of optical signal quality data for each abnormality factor; detecting an error based on optical signal quality data at a receiving end of the optical path; detecting a suspected section by narrowing down abnormal sections of the optical path based on the optical path accommodation information and the detected error information; inputting optical signal quality data of the receiving end of the optical path in which the error is detected to an abnormality cause determination unit provided for each of the optical paths; determining an abnormality factor for each of the abnormality factor determination units; determining a final abnormality location based on the suspected section and the abnormality factor determined by the abnormality factor determination unit, The method for identifying the location of an abnormality factor in an optical transmission system is characterized in that the abnormality factor determination process sequentially inputs input optical signal quality data into the classification models for each abnormality factor, and determines the cause of the abnormality from the probability indicating the presence or absence of an abnormality obtained for each classification model.

Images