Optical fiber abnormal event detection method and device, electronic equipment and storage medium

Abstract

The invention provides an optical fiber abnormal event detection method and device, electronic equipment and a storage medium. The invention discloses a detection device applied to an optical fiber abnormal event. The device is arranged at a receiving end of a data transmission system. The method comprises the following steps: receiving a sensing signal transmitted by a transmitting end based on an optical fiber link, and carrying out digital processing on the sensing signal to obtain a digital signal; performing signal processing on the digital signal to obtain a signal polarization state and a signal quality index; combining the signal polarization state and the signal quality index into initial time sequence data, and intercepting the initial time sequence data to obtain target time sequence data; and inputting the target time sequence data into a pre-trained anomaly detection model to obtain an anomaly probability of the optical fiber anomaly event, and determining a detection result of the optical fiber anomaly event based on the anomaly probability.

Description

Technical Field

[0001] The present disclosure relates to the field of signal transmission technology, and in particular to a method, device, electronic device, and storage medium for detecting abnormal optical fiber events. Background Art

[0002] The efficient operation of fiber-optic communication systems is crucial to modern communications. If abnormal events during fiber-optic communication are not detected in a timely manner, they can lead to signal interruption, data loss, and even disrupt the entire network. However, fiber abnormality event detection based on optical performance parameters cannot provide sufficient information about the tested fiber, which can lead to false alarms and inaccurate detection results.

[0003] In view of this, how to avoid inaccurate detection results of optical fiber anomaly detection events has become a technical problem that needs to be solved urgently. Summary of the Invention

[0004] In view of this, the purpose of the present disclosure is to provide a method, device, electronic device and storage medium for detecting optical fiber abnormal events to solve or partially solve the above technical problems.

[0005] Based on the above objectives, the first aspect of the present disclosure provides a method for detecting optical fiber abnormal events, which is applied to an optical fiber abnormal event detection device, wherein the device is provided at a receiving end of a data transmission system; the method comprises:

[0006] receiving a sensor signal transmitted by a transmitter based on an optical fiber link, and digitizing the sensor signal to obtain a digital signal;

[0007] Performing signal processing on the digital signal to obtain a signal polarization state and a signal quality indicator;

[0008] Combining the signal polarization state and the signal quality indicator into initial time series data, and intercepting the initial time series data to obtain target time series data;

[0009] The target time series data is input into a pre-trained anomaly detection model to obtain an abnormality probability of an optical fiber abnormal event, and a detection result of the optical fiber abnormal event is determined based on the abnormality probability.

[0010] Based on the same inventive concept, the second aspect of the present disclosure provides a device for detecting abnormal optical fiber events, the device being provided at a receiving end of a data transmission system; the device comprising:

[0011] a digital processing module configured to receive a sensor signal transmitted by a transmitter based on an optical fiber link and digitally process the sensor signal to obtain a digital signal;

[0012] A signal processing module is configured to perform signal processing on the digital signal to obtain a signal polarization state and a signal quality indicator;

[0013] an interception processing module, configured to combine the signal polarization state and the signal quality indicator into initial time series data, and perform interception processing on the initial time series data to obtain target time series data;

[0014] The anomaly detection module is configured to input the target time series data into a pre-trained anomaly detection model to obtain an anomaly probability of an optical fiber anomaly event, and determine a detection result of the optical fiber anomaly event based on the anomaly probability.

[0015] Based on the same inventive concept, the third aspect of the present disclosure proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor implements the method described above when executing the computer program.

[0016] Based on the same inventive concept, a fourth aspect of the present disclosure proposes a non-transitory computer-readable storage medium, which stores computer instructions for causing a computer to execute the method described above.

[0017] As can be seen from the above, the present disclosure provides a method, device, electronic device and storage medium for detecting optical fiber abnormal events. A sensor signal transmitted by the transmitting end based on the optical fiber link is received, and the sensor signal is digitized to obtain a digital signal. The digital signal is processed to obtain a signal polarization state and a signal quality index. The signal polarization state and the signal quality index are combined into initial time series data, and the initial time series data is intercepted and processed to obtain target time series data. The target time series data is input into a pre-trained abnormality detection model to obtain the abnormal probability of the optical fiber abnormal event, and the detection result of the optical fiber abnormal event is determined based on the abnormal probability. In this way, the abnormal probability of the optical fiber abnormal event is determined based on the signal polarization state and the signal quality index using the trained abnormality detection model, which can achieve accurate detection of optical fiber abnormal events. In addition, the presence of abnormal events in the optical fiber link can be collaboratively monitored based on multiple parameters of the signal polarization state and the signal quality index, which can avoid the problem of inaccurate detection results caused by using a single parameter for optical fiber abnormal event detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 Flowchart of a method for detecting abnormal optical fiber events according to an embodiment of the present disclosure;

[0020] Figure 2 This is a flowchart of abnormal event detection of an optical fiber link according to an embodiment of the present disclosure;

[0021] Figure 3 A schematic diagram of a sliding window intercepting time series data according to an embodiment of the present disclosure;

[0022] Figure 4 A schematic diagram of abnormal event detection of an optical fiber link according to an embodiment of the present disclosure;

[0023] Figure 5 This is a flow chart of processing time series data based on a convolutional neural network according to an embodiment of the present disclosure;

[0024] Figure 6 Schematic diagram of the structure of a device for detecting abnormal optical fiber events according to an embodiment of the present disclosure;

[0025] Figure 7 Schematic diagram of the structure of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0027] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should have the usual meanings understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the embodiments of the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the described object changes, the relative position relationship may also change accordingly.

[0028] As background information has been provided, with the continuous advancement of fiber-optic communication technology, optical networks have become a critical component of modern communications infrastructure and are widely used worldwide. Optical networks provide high-bandwidth, low-latency, and highly reliable communication services. Optical networks utilize optical fibers to deliver high-capacity, long-distance, and highly reliable link transmission. As the core artery carrying 98% of the world's data traffic, the security of optical networks is directly related to national security and the lifeblood of the digital economy.

[0029] With the continuous advancement of scientific research, researchers have discovered that optical fiber can not only be used for data transmission but also as a special sensor. Fiber-optic sensing technology, with its advantages such as high sensitivity, ultra-long-distance sensing, stability, and low cost, is widely used in various fields. In recent years, with the advancement of computing power and the support of big data, machine learning technology has experienced rapid development. Its application in data processing, noise extraction, and resolution enhancement in fiber-optic sensing systems has achieved some promising results, providing new ideas and methods for the development of fiber-optic sensing technology. The efficient operation of fiber-optic communication systems is crucial to modern communications. If abnormal events are not detected in a timely manner, they can lead to signal interruption, data loss, and even affect the entire network. How to integrate optical fiber with current optical networks to accurately sense abnormal events occurring on fiber-optic links in a cost-effective, efficient, and intelligent manner has become a new topic.

[0030] Currently, there are three main approaches to detecting fiber anomalies. Traditionally, optical time-domain reflectometers (OTDRs) are used in optical networks as a means of detecting and locating anomalies in fiber links. This method relies on expensive, specialized hardware to detect changes in optical signal power. Coherent OTDRs employ coherent detection to capture the amplitude and phase of backscattered light, offering significant improvements over traditional OTDRs. A more advanced approach, distributed acoustic sensing (DAS), leverages Rayleigh backscattering in optical fibers to achieve high spatial resolution for detecting mechanical vibrations along the fiber's length. While this technology is highly sensitive and accurate, the equipment used is expensive and integration with traditional optical networks is challenging. While methods for detecting fiber anomalies based on optical performance monitor (OPM) data are compatible with current fiber networks, their sensitivity is insufficient to detect even minor anomalies at the physical layer. Furthermore, OPM data is susceptible to real-world events such as ambient temperature fluctuations, making it easy for this approach to confuse it with anomalies, leading to false alarms. The state of polarization (SOP) represents the direction of the electric field when a light wave propagates in an optical fiber, and includes values ​​such as Stokes parameters, which define the polarization of light. Its high sensitivity to external disturbances is usually attributed to stress-induced birefringence, which makes SOP a key attribute for environmental sensing. Therefore, SOP has become the most promising technology among various schemes for monitoring abnormal events in optical fiber networks, and has received great attention from academia and industry. However, the relationship between the change in polarization state and abnormal events may be nonlinear, which also leads to a higher complexity in its pattern recognition. The natural drift of SOP data also has a great probability of masking the real abnormal signal.

[0031] However, regardless of the optical fiber anomaly detection scheme, measuring changes in a single parameter (e.g., backscattered light phase, optical signal intensity) is limited. In many cases, measuring a single parameter does not provide sufficient information about the fiber under test, potentially leading to false alarms or misreporting of anomalies, resulting in a waste of manpower and material resources. In the field of optical fiber anomaly detection, collaborative multi-parameter measurement is urgently needed.

[0032] As mentioned above, how to avoid inaccurate detection results of optical fiber anomaly detection events has become an important research issue.

[0033] Based on the above description, if Figure 1As shown, the optical fiber abnormal event detection method proposed in this embodiment is applied to an optical fiber abnormal event detection device, which is provided at a receiving end of a data transmission system; the method includes:

[0034] Step 101: receiving a sensing signal transmitted by a transmitting end based on an optical fiber link, and digitizing the sensing signal to obtain a digital signal.

[0035] In a specific implementation, the data transmission system includes a transmitter, an optical fiber link, and a receiver. A device for detecting optical fiber abnormalities is provided at the receiver. The transmitter transmits a sensing signal via the optical fiber link to the receiver, which samples and digitizes the received sensing signal to produce a digital signal. The sensing signal is a continuous signal.

[0036] Step 102: Process the digital signal to obtain a signal polarization state and a signal quality indicator.

[0037] During specific implementation, the signal polarization state and signal quality index are extracted from the digital signal through an offline digital signal processing algorithm (DSP).

[0038] The signal's state of polarization (SOP) refers to the directional vibration characteristics of the electric field vector during light wave propagation. The signal quality indicator (Error Vector Magnitude, EVM) refers to the difference between the ideal signal and the actual received signal. EVM is an optical performance parameter.

[0039] Step 103: Combine the signal polarization state and the signal quality indicator into initial time series data, and perform interception processing on the initial time series data to obtain target time series data.

[0040] In a specific implementation, the signal polarization state and the signal quality index are combined into initial time series data, and the initial time series data is intercepted and processed using a sliding window with a preset step size to obtain target time series data.

[0041] The sliding window can adaptively adjust its step size based on the currently detected sensor signal. Specifically, it determines whether an abnormal event has been detected in the sensor signal; in response to determining that no abnormal event has been detected in the sensor signal, the initial time series data is intercepted and processed using a sliding window of a first preset step size to obtain target time series data; in response to determining that an abnormal event has been detected in the sensor signal, the initial time series data is intercepted and processed using a sliding window of a second preset step size to obtain target time series data; wherein the first preset step size is greater than the second preset step size.

[0042] In this way, when no abnormal event is detected in the sensor signal, the initial time series data is intercepted and processed using the larger first preset step size to obtain the target time series data. In this way, the amount of intercepted target time series data is small, which can reduce the amount of computation when processing the target time series data. When an abnormal event is detected in the sensor signal, the initial time series data is intercepted and processed using the smaller second preset step size to obtain the target time series data. In this way, using the smaller second preset step size can even cause the windows to overlap, allowing for more fine-grained scanning and achieving optimal performance.

[0043] Step 104 : Input the target time series data into a pre-trained anomaly detection model to obtain an abnormality probability of an optical fiber abnormal event, and determine a detection result of the optical fiber abnormal event based on the abnormality probability.

[0044] In specific implementations, the target time series data is segmented. Pre-trained anomaly detection models are capable of processing multivariate, long time series data. For example, anomaly detection models based on recurrent neural networks (RNNs), convolutional neural networks (CNNs), and transformer models all need to simultaneously model temporal dependencies and inter-variable correlations. Because of the large number of input parameters, multi-channel interactive processing is often required.

[0045] Compared to models that rely solely on signal polarization state data, multivariate models can integrate multiple factors to make decisions. While this increases the computational effort, it significantly improves the precision and accuracy of anomaly detection. In high-noise conditions, the stability of signal quality indicators mitigates the degradation of signal polarization state estimation accuracy, ensuring the robustness of the anomaly detection model.

[0046] The target time series data is input into a pre-trained anomaly detection model, and the anomaly detection model is used to determine the anomaly probability of the optical fiber anomaly event based on the signal polarization state and signal quality indicators in the target time series data, and the detection result of the optical fiber anomaly event is determined based on the anomaly probability.

[0047] When the detection result indicates that no abnormal event exists, the optical fiber abnormal event detection device is in a default state. When the detection result indicates that an abnormal event exists and the abnormal event is an eavesdropping event, an alarm is issued. The detection result can be used to determine the type of abnormal event occurring on the optical fiber at the target time.

[0048] In order to make full use of current facilities, realize real-time and intelligent detection of optical fiber abnormal events on the current optical network and distinguish between normal disturbances (for example, disturbances caused by equipment installation, etc.) and abnormal events (for example, optical fiber breakage, human eavesdropping, etc.), the embodiment of the present disclosure builds a new detection paradigm based on the traditional optical performance parameter (Optical Performance Monitor, referred to as OPM) detection, which combines the signal polarization state (State Of Polarization, referred to as SOP) with optical performance parameters (i.e., signal quality indicators) to detect optical fiber abnormal events. Combined with machine learning algorithms, it analyzes the changes in optical fiber polarization state in real time, autonomously identifies different optical fiber events, realizes intelligent diagnosis of optical fiber abnormal events, and improves the security of optical fiber networks.

[0049] Through the above embodiment, a sensor signal transmitted by the transmitting end based on the optical fiber link is received, and the sensor signal is digitized to obtain a digital signal. The digital signal is processed to obtain a signal polarization state and a signal quality index. The signal polarization state and the signal quality index are combined into initial time series data, and the initial time series data is intercepted and processed to obtain target time series data. The target time series data is input into a pre-trained anomaly detection model to obtain the anomaly probability of the optical fiber anomaly event, and the detection result of the optical fiber anomaly event is determined based on the anomaly probability. In this way, the trained anomaly detection model is used to determine the anomaly probability of the optical fiber anomaly event based on the signal polarization state and the signal quality index, which can achieve accurate detection of the optical fiber anomaly event. In addition, based on the multiple parameters of the signal polarization state and the signal quality index, the optical fiber link is collaboratively monitored to determine whether there is an abnormal event, which can avoid the problem of using a single parameter to detect optical fiber anomaly events, resulting in inaccurate detection results.

[0050] In some embodiments, step 102 includes:

[0051] Step 1021: Determine a signal polarization state according to the digital signal based on a transfer function of the optical fiber link.

[0052] Step 1022: Determine the actual symbol sequence of the receiving end from the digital signal, and determine a signal quality indicator according to the actual symbol sequence and the ideal symbol sequence of the transmitting end.

[0053] In specific implementation, the influence of the optical fiber link on the polarization state can be expressed by a transfer function, which describes the change of the polarization state of the signal during the transmission process.

[0054] Signal polarization state (State Of Polarization, referred to as SOP) refers to the vibration direction characteristics of the electric field vector of the light wave during propagation. The Jones vector of the light wave is determined based on the digital signal. The digital signal is compensated for by the transfer function of the optical fiber link using an adaptive equalizer to obtain an equalized signal, and the polarization state rotation of the optical fiber link is determined by extracting the zero-frequency component in the equalized signal. The linear horizontal polarized light after the optical fiber link conversion is determined according to the polarization state rotation, and the Stokes parameters are determined according to the linear horizontal polarized light. The real-time position information of the Stokes parameters on the Poincare sphere is used as the signal polarization state.

[0055] The signal quality indicator (Error Vector Magnitude, or EVM) refers to the difference between the ideal signal and the actual received signal. EVM is an optical performance parameter. The ideal symbol sequence at the transmitter is obtained, and the actual symbol sequence at the receiver is determined from the digital signal. The vector difference between the actual and ideal symbol sequences is used as the error vector. The square of the modulus of each error vector is determined, and the squared moduli of multiple error vectors are averaged to obtain the mean square error (MSE). The ideal symbol average power is determined based on the ideal symbol sequence. The signal quality indicator is determined based on the MSE and the ideal symbol average power.

[0056] Through this solution, the polarization state of the signal can be accurately determined from the digital signal based on the transfer function of the optical fiber link. The actual symbol sequence at the receiving end is determined from the digital signal, and the signal quality index is determined based on the actual symbol sequence and the ideal symbol sequence at the transmitting end. In this way, the optical fiber link is collaboratively monitored for abnormal events based on the multiple parameters of the polarization state and the signal quality index, avoiding the problem of inaccurate detection results caused by using a single parameter for optical fiber abnormality detection.

[0057] In some embodiments, step 1021 includes:

[0058] Step 10211: Determine the Jones vector of the light wave based on the digital signal.

[0059] Step 10212: Using an adaptive equalizer to perform compensation processing on the digital signal based on the transfer function of the optical fiber link to obtain an equalized signal, and determining the polarization rotation of the optical fiber link by extracting a zero-frequency component from the equalized signal.

[0060] Step 10213: Determine the linear horizontal polarized light after the optical fiber link conversion according to the polarization state rotation, and determine the Stokes parameters according to the linear horizontal polarized light.

[0061] Step 10214: Use the real-time position information of the Stokes parameter on the Poincare sphere as the signal polarization state.

[0062] In specific implementations, the polarization state of an optical signal can be represented by a Jones vector, which is usually a complex vector that describes the amplitude and phase of a light wave in two orthogonal polarization directions (e.g., horizontal and vertical). The Jones vector of a light wave is determined based on a digital signal.

[0063]

[0064] in, It is the complex vector of the optical signal, namely the Jones vector of the light wave, which is used to describe the polarization state of the light wave. x is the complex amplitude of the light wave in the X direction (horizontally polarized) of the fiber cross section, s y It is the complex amplitude of the light wave in the Y direction (vertical polarization) of the optical fiber cross section. The complex amplitude includes amplitude and phase information.

[0065] An adaptive equalizer is used to compensate for channel-induced distortion, including polarization mode dispersion (PMD) and rotation of state of polarization (RSOP). The adaptive equalizer compensates the digital signal based on the fiber link's transfer function to produce an equalized signal. The zero-frequency component extracted from the equalized signal determines the polarization rotation of the fiber link. RSOP describes the effect of the fiber channel on the polarization state of the optical signal.

[0066] The linearly horizontally polarized light after the fiber link conversion is determined based on the polarization state rotation, and the Stokes parameters are determined based on the linearly horizontally polarized light. The Stokes parameters describe the polarization state of light and can be represented on the Poincare sphere. The real-time position information of the Stokes parameters on the Poincare sphere is used as the signal polarization state.

[0067] The above scheme determines the Jones vector of the lightwave based on the digital signal. An adaptive equalizer compensates the digital signal based on the transfer function of the optical fiber link to generate an equalized signal. The zero-frequency component extracted from the equalized signal determines the polarization rotation of the optical fiber link, thereby accurately determining the polarization rotation effect of the optical signal. The linear horizontal polarization of the optical fiber link after conversion is determined based on the polarization rotation, and the Stokes parameters are determined based on the linear horizontal polarization. The real-time position of the Stokes parameters on the Poincare sphere is used as the signal polarization state, thereby accurately determining the signal polarization state.

[0068] In some embodiments, step 10212 includes:

[0069] Step 10212A: Using an adaptive equalizer, the digital signal is compensated based on the transfer function of the optical fiber link to obtain an equalized signal.

[0070]

[0071] in, is the equalized signal, is the transfer function of the fiber link, is the inverse function of the transfer function of the optical fiber link, and ω is the signal frequency.

[0072] Step 10212B: determining the polarization rotation of the optical fiber link by extracting the zero-frequency component from the equalized signal.

[0073]

[0074] in, is the polarization rotation of the fiber link, M xx M is the ratio of horizontal polarization light to maintain horizontal polarization after passing through the optical fiber. xy is the coupling coefficient of vertically polarized light to output vertically polarized light, M yx is the coupling coefficient of horizontal polarized light to output vertical polarized light, M yy is the ratio in which vertically polarized light remains vertically polarized.

[0075] In specific implementations, the adaptive equalizer compensates for polarization-related distortion in the fiber channel through inverse filtering to restore the polarization state information in the original signal, thereby accurately extracting the SOP and EVM data. The above formula represents the inverse operation to offset the channel's distortion of the signal's polarization state. Specifically, the fiber link causes the optical signal's polarization state to rotate (RSOP). The adaptive equalizer's role is to inversely compensate for this rotation, allowing the receiver to recover polarization state information consistent with that of the transmitter. However, the characteristics of the fiber link will change over time, and the adaptive equalizer continuously updates the filter coefficients to track changes in channel characteristics and maintain compensation accuracy.

[0076] Rotation of State of Polarization (RSOP) is a characteristic of optical fiber channels that causes changes in the polarization state. RSOP describes the effect of an optical fiber channel on the polarization state of an optical signal.

[0077] Through the above scheme, an adaptive equalizer is used to compensate the digital signal based on the transfer function of the optical fiber link to obtain an equalized signal. The polarization rotation of the optical fiber link is determined by extracting the zero-frequency component from the equalized signal, thereby accurately determining the rotation effect of the polarization state of the optical signal.

[0078] In some embodiments, step 10213 includes:

[0079] Step 10213A, determining the linear horizontal polarized light after the optical fiber link conversion according to the polarization state rotation,

[0080]

[0081] in, is the horizontally polarized light after conversion in the optical fiber link, M xx M is the ratio of horizontal polarization light to maintain horizontal polarization after passing through the optical fiber. xy is the coupling coefficient of vertically polarized light to output vertically polarized light.

[0082] Step 10213B, determining Stokes parameters based on the linear horizontal polarized light,

[0083]

[0084] Where S0 is the total power of the light wave, S1 is the power difference between the horizontally polarized light and the vertically polarized light in the light wave, S2 is the power difference between the ±45° linear polarization components in the light wave, and S3 is the power difference between the left and right circularly polarized light in the light wave. Re(·) represents the real signal component related to the polarization state of the extracted light wave, and Im(·) represents the imaginary signal component related to the polarization state of the extracted light wave. Indicates M xy The complex conjugate of .

[0085] In a specific implementation, the coupling characteristics of the horizontal and vertical polarization components in the optical fiber are extracted through complex number operations.

[0086] In the above formula, S0 represents the total power of the lightwave, while S1, S2, and S3 quantify the intensity differences between different types of polarized light. S1 describes the power difference between horizontally and vertically polarized light in the lightwave. If S1 > 0, horizontal polarization is dominant, and if S1 < 0, vertical polarization is dominant. S2 describes the power difference between the ±45° linear polarization components in the lightwave. If S2 > 0, +45° polarization is dominant, and if S2 < 0, -45° polarization is dominant. S3 describes the power difference between left- and right-handed circularly polarized light in the lightwave. If S3 > 0, right-handed circular polarization is dominant, and if S3 < 0, left-handed circular polarization is dominant. S1, S2, and S3 form the three coordinate axes of the Poincare sphere, with each point on the sphere corresponding to a polarization state.

[0087] Through this scheme, the linear horizontal polarization of the converted optical fiber link is determined based on the polarization state rotation, and the Stokes parameters are determined based on the linear horizontal polarization. The real-time position information of the Stokes parameters on the Poincare sphere is used as the signal polarization state, thereby accurately obtaining the signal polarization state.

[0088] In some embodiments, step 1022 includes:

[0089] Step 10221: Acquire the ideal symbol sequence of the transmitting end, and determine the actual symbol sequence of the receiving end from the digital signal, and use the vector difference between the actual symbol sequence and the ideal symbol sequence as an error vector.

[0090] Step 10222: determine the square of the modulus of each error vector, and average the squares of the moduli of multiple error vectors to obtain a mean square error.

[0091] Step 10223: Determine the ideal symbol average power based on the ideal symbol sequence.

[0092] Step 10224: determining the signal quality indicator according to the mean square error and the ideal symbol average power.

[0093]

[0094] Where EVM is the signal quality indicator, E is the mean square error, P a is the ideal symbol average power.

[0095] In specific implementation, there are distortions and noises in the actual symbol sequence received after DSP processing. Knowing the ideal symbol sequence of the transmitter, the vector difference between the actual symbol sequence and the ideal symbol sequence is calculated for each signal to obtain e i .

[0096] Calculate the mean square error E of the squares of all error vector moduli, and combine it with the mean power P of the ideal symbol sequence. a , calculate the EVM value:

[0097]

[0098] Among them, EVM is the signal quality indicator, E is the mean square error, P a is the ideal symbol average power.

[0099] The above scheme uses the vector difference between the actual symbol sequence and the ideal symbol sequence as an error vector. The square of the modulus of each error vector is determined, and the squared moduli of multiple error vectors are averaged to obtain the mean square error (MSE). The ideal symbol average power is determined based on the ideal symbol sequence. The mean square error and the ideal symbol average power can be used to accurately determine the signal quality indicator.

[0100] In some embodiments, step 104 includes:

[0101] Step 1041 : Input the target time series data into a pre-trained anomaly detection model to obtain a first anomaly probability of a fracture anomaly, a second anomaly probability of a bending anomaly, and a third anomaly probability of a vibration anomaly.

[0102] Step 1042: Determine a target abnormal event with the highest abnormal probability from among the fracture abnormality, the bending abnormality, and the vibration abnormality based on the first abnormality probability, the second abnormality probability, and the third abnormality probability.

[0103] Step 1043: Use the target abnormal event and the target abnormal probability corresponding to the target abnormal event as the detection result of the optical fiber abnormal event.

[0104] In specific implementation, the target time series data is input into a pre-trained anomaly detection model to obtain the first anomaly probability for a break anomaly, the second anomaly probability for a bend anomaly, and the third anomaly probability for a vibration anomaly. When the third anomaly probability is the highest, the target anomaly is determined to be a vibration anomaly. The vibration anomaly event and the third anomaly probability are used as the fiber anomaly detection results.

[0105] The anomaly detection model is trained using a convolutional neural network. The anomaly detection model processes the target time series data intercepted by the sliding window and can accurately detect different types of abnormal events and the corresponding anomaly probabilities.

[0106] Specifically, the anomaly detection model constructs a two-dimensional matrix from the target time series captured by a sliding window with a preset step size. The convolutional layer uses 64 filters to output a 64-channel feature map to extract primary features. 1×2 max pooling is performed on each feature map to reduce redundant information and retain primary features. By stacking two convolutional layers, the number of channels is expanded from 64 to 128, allowing the extraction of higher-level and more complex time series. 2×2 pooling is performed simultaneously in the time and feature dimensions, halving both the time step and feature dimension, reducing computational complexity while preserving key features. The flattening layer converts the multidimensional features into a one-dimensional vector for input into the fully connected layer. The first fully connected layer projects the 2028-dimensional features into a 256-dimensional space. The second fully connected layer maps the 256-dimensional features into multiple categories. A dropout layer is inserted in between to prevent overfitting. Finally, a softmax function is used to generate the anomaly probability for each anomaly event. The softmax function converts the event type into a more intuitive probability distribution, simultaneously displaying the occurrence and type of an anomaly to the user.

[0107] Using the above scheme, the target time series data is input into a pre-trained anomaly detection model to obtain a first anomaly probability for a break anomaly, a second anomaly probability for a bend anomaly, and a third anomaly probability for a vibration anomaly. Based on the first, second, and third anomaly probabilities, the target anomaly event with the highest probability among the break, bend, and vibration anomalies is determined. The target anomaly event and its corresponding target anomaly probability are used as the fiber anomaly event detection result. This way, the detected fiber anomaly event results in the target anomaly event with the highest probability and its corresponding target anomaly probability, allowing users to quickly and intuitively identify the target anomaly event present in the fiber link.

[0108] Through the above embodiment, a sensor signal transmitted by the transmitting end based on the optical fiber link is received, and the sensor signal is digitized to obtain a digital signal. The digital signal is processed to obtain a signal polarization state and a signal quality index. The signal polarization state and the signal quality index are combined into initial time series data, and the initial time series data is intercepted and processed to obtain target time series data. The target time series data is input into a pre-trained anomaly detection model to obtain the anomaly probability of the optical fiber anomaly event, and the detection result of the optical fiber anomaly event is determined based on the anomaly probability. In this way, the trained anomaly detection model is used to determine the anomaly probability of the optical fiber anomaly event based on the signal polarization state and the signal quality index, which can achieve accurate detection of the optical fiber anomaly event. In addition, based on the multiple parameters of the signal polarization state and the signal quality index, the optical fiber link is collaboratively monitored to determine whether there is an abnormal event, which can avoid the problem of using a single parameter to detect optical fiber anomaly events, resulting in inaccurate detection results.

[0109] It should be noted that the embodiments of the present disclosure may be further described in the following manner:

[0110] The embodiments of the present disclosure are aimed at the security requirements of optical fiber networks. Taking into account the defect that in actual events, abnormal optical fiber events will be confused with normal events such as disturbances, and it is impossible to accurately identify whether abnormal events such as eavesdropping have occurred. The embodiments of the present disclosure propose a method based on collaborative monitoring of SOP and optical performance parameters. Combined with machine learning algorithms, SOP and OPM data are analyzed in real time and intelligently, and the type of optical fiber abnormal event or normal disturbance is accurately detected. The following uses the combination of EVM data and SOP as an example to introduce the embodiments of the present disclosure.

[0111] Figure 2 FIG. 1 is a flow chart of abnormal event detection of optical fiber link according to an embodiment of the present disclosure. Figure 2 As shown in the figure, optical fiber link abnormal event detection includes:

[0112] Step 1: Sampling and digitizing the captured sensor signal. Continuous sensor signals are captured at the receiving end of the system, sampled, and digitized to obtain digital signals.

[0113] Step 2: Use offline DSP to extract the SOP and EVM data contained in the sensor signal. The digitized digital signal is passed through an adaptive equalizer to continuously update the filter coefficients.

[0114] Assume that the complex vector of the transmitted optical signal is expressed as:

[0115]

[0116] in, It is the complex vector of the optical signal, namely the Jones vector of the light wave, which is used to describe the polarization state of the light wave. x is the complex amplitude of the light wave in the X direction (horizontally polarized) of the fiber cross section, s y It is the complex amplitude of the light wave in the Y direction (vertical polarization) of the optical fiber cross section. The complex amplitude includes amplitude and phase information.

[0117] Polarization dependent transfer function through the optical fiber channel After that, in the received signal recovery, the adaptive equalizer for:

[0118]

[0119] By extracting The zero-frequency component of the optical fiber link can be used to analyze the rotation of polarization state (RSOP). The RSOP is expressed as:

[0120]

[0121] It can be inferred that linear horizontal polarization (LHP) light is converted by the fiber link into:

[0122]

[0123] Then, the Stokes parameter can be expressed as:

[0124]

[0125] The position information of the Stokes parameters on the Poincare sphere is recorded as a dynamic SOP change parameter updated in real time.

[0126] The EVM data is estimated based on the signal recovered by DSP processing. The error vector e of each symbol is used i , calculate the mean square error E of the squares of all error vector moduli, and combine it with the mean value P of the ideal symbol average powera , calculate the EVM value:

[0127]

[0128] Step 3: Adaptively intercept the time series data using a sliding window. Combine the SOP data (S1, S2, S3) with the EVM data to form new multivariate time series data. Use a fixed-size sliding window to intercept the time series. Figure 3 This is a schematic diagram of the sliding window intercepting time series data in the embodiment of the present disclosure. Figure 3 As shown in the figure, the sliding window dynamically adjusts its step size based on the currently detected signal, a process known as adaptive step size. When the detection result is "no anomaly," a larger step size is used to reduce computational effort. Once the detection result indicates a suspected anomaly, a smaller step size is used, even allowing windows to overlap for more fine-grained scanning and optimal performance.

[0129] like Figure 3 As shown in the figure, when the detection result is no anomaly, the step size is set to 4, and there is no overlap between window 1 and window 2, which can reduce the amount of calculation. When the detection result is anomaly, the step size is set to 3, and there is an overlap between window 1 and window 2, which can achieve fine-grained scanning and thus accurate detection.

[0130] Step 4: Deep Learning Model. Use a deep learning model capable of processing multivariate, long-time series data to process the segmented data. For example, models such as recurrent neural networks (RNNs), convolutional neural networks (CNNs), and Transformers all need to simultaneously model both temporal dependencies and correlations between variables. Due to their large number of parameters, they often require multi-channel interactive processing. Compared to models targeting single SOP data, multivariate models can integrate multiple factors to make decisions. While this increases the computational effort, it greatly improves the precision and accuracy of anomaly detection. Under high-noise conditions, the stability of the EVM mitigates the decline in the accuracy of SOP estimation, ensuring model robustness.

[0131] Step 5: Determine the event type. The event type occurring on the fiber at a given moment is determined. If the event is normal, the system remains in its default state. If the event is determined to be an illegal event such as eavesdropping, an alarm is generated.

[0132] Figure 4 FIG. 1 is a schematic diagram of detecting abnormal events in an optical fiber link according to an embodiment of the present disclosure. Figure 4 As shown, the transmitting end transmits the signal to the receiving end through the optical fiber link, and the optical fiber link abnormal event detection device is set at the receiving end. The event occurs in area A between the optical fiber links. The following steps are used to detect whether the event is an abnormal event.

[0133] (1) A signal is sent from a transmitting end to a receiving end via an optical fiber link. The receiving end uses the optical fiber abnormal event monitoring system B of the present invention. At a certain time, an abnormal event occurs on the optical fiber link in area A.

[0134] (2) System B collects the signal at the receiving end and samples and digitizes it.

[0135] (3) The digital signal processing module records the parameters of the Stokes parameters on the Poincare sphere and the affected EVM data.

[0136] (4) The sliding window intercepts the time series data containing SOP and EVM features.

[0137] (5) Using the CNN model trained on the dataset, the segmented time series data captured by the sliding window is processed to accurately detect different event types. Figure 5 Flowchart of processing time series data based on convolutional neural network according to the embodiment of the present disclosure. Figure 5 As shown, time series data is processed based on convolutional neural networks, including:

[0138] 1. The fixed-length segments captured by the sliding window are organized into a two-dimensional matrix. The convolutional layer uses 64 filters to output a 64-channel feature map to extract primary features.

[0139] 2. Perform 1×2 maximum pooling on each feature map to reduce redundant information and retain salient features.

[0140] 3. By stacking two convolutional layers, the number of channels is expanded from 64 to 128, extracting higher-level and more complex temporal features.

[0141] 4. Perform 2×2 pooling in both the time and feature dimensions, halving the time step and feature dimension, reducing the amount of computation while retaining key features.

[0142] 5. The flattening layer converts the multi-dimensional features into a one-dimensional vector to adapt to the fully connected layer input.

[0143] 6. The first fully connected layer projects the 2028-dimensional features into a 256-dimensional space. The second fully connected layer maps the 256-dimensional features into multiple categories. A dropout layer is inserted in between to prevent overfitting. Finally, a softmax function is used to generate the probability of each event.

[0144] (6) The softmax function converts the event type into a more intuitive probability distribution, and at the same time shows the user whether an abnormal event occurs and the event type.

[0145] Through the above-described embodiments, real-time monitoring of the SOP and OPM characteristic data in optical fibers over time is achieved. Using a deep learning model that processes multivariate data, abnormal events can be detected more accurately. The disclosed embodiments are compatible with current optical network architectures, requiring no additional equipment and without disrupting normal communications. They can be directly applied to existing networks through software upgrades, enabling intelligent fiber sensing across the entire network.

[0146] Multi-parameter collaborative monitoring of fiber link anomalies. This system integrates SOP and OPM data to collaboratively monitor multiple parameters of fiber links, overcoming the limitations of single parameters in detecting anomalies. Adaptive segmentation of time series data. Using a fixed-size sliding window, the system can adaptively adjust the step size for different events, dynamically adjusting the amount of data input to the model. Deep learning model processing of fiber time series data. Using a deep learning model tailored to multivariate, long time series data, the system processes feature data containing fiber link information, enabling accurate detection of a variety of fiber anomalies.

[0147] It should be noted that the method of the embodiments of the present disclosure can be performed by a single device, such as a computer or server. The method of the embodiments of the present disclosure can also be applied in a distributed scenario, where multiple devices cooperate to perform the method. In such a distributed scenario, one of the multiple devices may only perform one or more steps of the method of the embodiments of the present disclosure, and the multiple devices will interact with each other to complete the method.

[0148] It should be noted that the above description is limited to some embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in an order different from that described in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0149] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure further provides a device for detecting abnormal optical fiber events.

[0150] refer to Figure 6 , the optical fiber abnormal event detection device, the device is arranged at the receiving end of the data transmission system; the device comprises:

[0151] The digital processing module 301 is configured to receive a sensor signal transmitted by a transmitter based on an optical fiber link and digitally process the sensor signal to obtain a digital signal;

[0152] The signal processing module 302 is configured to perform signal processing on the digital signal to obtain a signal polarization state and a signal quality indicator;

[0153] An interception processing module 303 is configured to combine the signal polarization state and the signal quality indicator into initial time series data, and perform interception processing on the initial time series data to obtain target time series data;

[0154] The anomaly detection module 304 is configured to input the target time series data into a pre-trained anomaly detection model to obtain an anomaly probability of an optical fiber anomaly event, and determine a detection result of the optical fiber anomaly event based on the anomaly probability.

[0155] In some embodiments, the signal processing module 302 includes:

[0156] a signal polarization state determining unit, configured to determine a signal polarization state according to the digital signal based on a transfer function of the optical fiber link;

[0157] The signal quality indicator determination unit is configured to determine the actual symbol sequence of the receiving end from the digital signal, and determine the signal quality indicator according to the actual symbol sequence and the ideal symbol sequence of the transmitting end.

[0158] In some embodiments, the signal polarization state determination unit includes:

[0159] a Jones vector determination subunit configured to determine a Jones vector of the light wave based on the digital signal;

[0160] a polarization state rotation determining subunit, configured to use an adaptive equalizer to perform compensation processing on the digital signal based on a transfer function of the optical fiber link to obtain an equalized signal, and determine the polarization state rotation of the optical fiber link by extracting a zero-frequency component from the equalized signal;

[0161] a Stokes parameter determination subunit, configured to determine the linear horizontal polarized light after the optical fiber link conversion according to the polarization state rotation, and determine the Stokes parameters according to the linear horizontal polarized light;

[0162] The signal polarization state determination subunit is configured to use the real-time position information of the Stokes parameter on the Poincare sphere as the signal polarization state.

[0163] In some embodiments, the polarization state rotation determining subunit is specifically configured to:

[0164] Using an adaptive equalizer, the digital signal is compensated based on the transfer function of the optical fiber link to obtain an equalized signal.

[0165]

[0166] in, is the equalized signal, is the transfer function of the fiber link, is the inverse function of the transfer function of the optical fiber link, ω is the signal frequency;

[0167] Determine the polarization rotation of the optical fiber link by extracting the zero-frequency component from the equalized signal,

[0168]

[0169] in, is the polarization rotation of the fiber link, M xx M is the ratio of horizontal polarization light to maintain horizontal polarization after passing through the optical fiber. xy is the coupling coefficient of vertically polarized light to output vertically polarized light, M yx is the coupling coefficient of horizontal polarized light to output vertical polarized light, M yy is the ratio in which vertically polarized light remains vertically polarized.

[0170] In some embodiments, the Stokes parameter determination subunit is specifically configured to:

[0171] Determine the linear horizontal polarized light after the optical fiber link conversion according to the polarization state rotation,

[0172]

[0173] in, is the horizontally polarized light after conversion in the optical fiber link, M xx M is the ratio of horizontal polarization light to maintain horizontal polarization after passing through the optical fiber. xy is the coupling coefficient of the vertically polarized light to the output vertically polarized light;

[0174] Determine the Stokes parameters based on the linear horizontal polarized light,

[0175]

[0176] Where S0 is the total power of the light wave, S1 is the power difference between the horizontally polarized light and the vertically polarized light in the light wave, S2 is the power difference between the ±45° linear polarization components in the light wave, and S3 is the power difference between the left and right circularly polarized light in the light wave. Re(·) represents the real signal component related to the polarization state of the extracted light wave, and Im(·) represents the imaginary signal component related to the polarization state of the extracted light wave. Indicates M xy The complex conjugate of .

[0177] In some embodiments, the signal quality indicator determination unit includes:

[0178] an error vector determination subunit, configured to obtain an ideal symbol sequence of the transmitting end, determine an actual symbol sequence of the receiving end from the digital signal, and use a vector difference between the actual symbol sequence and the ideal symbol sequence as an error vector;

[0179] a mean square error determination subunit, configured to determine the square of the modulus of each error vector, and average the squares of the moduli of multiple error vectors to obtain the mean square error;

[0180] an average power determination subunit, configured to determine an ideal symbol average power based on the ideal symbol sequence;

[0181] a signal quality indicator determination subunit, configured to determine the signal quality indicator according to the mean square error and the ideal symbol average power,

[0182]

[0183] Where EVM is the signal quality indicator, E is the mean square error, P a is the ideal symbol average power.

[0184] In some embodiments, the anomaly detection module 304 includes:

[0185] an abnormality probability determination unit configured to input the target time series data into a pre-trained abnormality detection model to obtain a first abnormality probability of a fracture abnormality, a second abnormality probability of a bending abnormality, and a third abnormality probability of a vibration abnormality;

[0186] a target abnormal event determining unit configured to determine a target abnormal event with the highest abnormality probability from among the fracture abnormality, the bending abnormality, and the vibration abnormality according to the first abnormality probability, the second abnormality probability, and the third abnormality probability;

[0187] The detection result determining unit is configured to use the target abnormal event and the target abnormal probability corresponding to the target abnormal event as the detection result of the optical fiber abnormal event.

[0188] For the convenience of description, the above devices are described as being functionally divided into various modules. Of course, when implementing the present disclosure, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0189] The device of the above embodiment is used to implement the corresponding optical fiber abnormal event detection method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0190] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the program, the method for detecting abnormal optical fiber events described in any of the above embodiments is implemented.

[0191] Figure 7 10 is a schematic diagram showing a more specific hardware structure of an electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other within the device via the bus 1050.

[0192] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0193] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, etc. The memory 1020 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0194] The input / output interface 1030 is used to connect input / output modules to implement information input and output. The input / output modules can be configured as components within the device (not shown in the figure) or can be externally connected to the device to provide corresponding functions. Input devices may include a keyboard, mouse, touch screen, microphone, various sensors, etc., and output devices may include a display, speaker, vibrator, indicator light, etc.

[0195] The communication interface 1040 is used to connect to a communication module (not shown) to enable communication between the device and other devices. The communication module can communicate via a wired method (e.g., USB (Universal Serial Bus), network cable, etc.) or a wireless method (e.g., mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).

[0196] The bus 1050 comprises a path for transmitting information between the various components of the device (eg, the processor 1010 , the memory 1020 , the input / output interface 1030 , and the communication interface 1040 ).

[0197] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in a specific implementation, the device may also include other components necessary for normal operation. In addition, it will be understood by those skilled in the art that the above device may only include the components necessary to implement the embodiments of this specification, and does not necessarily include all the components shown in the figure.

[0198] The electronic device of the above embodiment is used to implement the corresponding optical fiber abnormal event detection method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be described in detail here.

[0199] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the method for detecting optical fiber abnormal events as described in any of the above embodiments.

[0200] The computer-readable media of this embodiment include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0201] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the optical fiber abnormal event detection method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0202] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a computer program product, including computer program instructions. When the computer program instructions are run on a computer, the computer executes the method for detecting optical fiber abnormal events as described in any of the above embodiments, which has the beneficial effects of the corresponding method embodiments and will not be repeated here.

[0203] It is understandable that before using the technical solutions of each embodiment of the present disclosure, the type, scope of use, usage scenarios, etc. of the personal information involved will be informed to the user in an appropriate manner, and the user's authorization will be obtained.

[0204] For example, in response to a user's active request, a prompt message is sent to the user to clearly inform the user that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the electronic device, application, server, storage medium, or other software or hardware that performs the operation of the disclosed technical solution based on the prompt message.

[0205] As an optional but non-limiting implementation, in response to a user's active request, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. Furthermore, the pop-up window may also contain a selection control for the user to select "agree" or "disagree" to provide personal information to the electronic device.

[0206] It is understandable that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0207] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present disclosure is limited to these examples. Based on the concept of the present disclosure, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of simplicity.

[0208] In addition, to simplify the description and discussion, and so as not to obscure the embodiments of the present disclosure, known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided figures. In addition, devices may be shown in the form of block diagrams to avoid obscuring the embodiments of the present disclosure, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present disclosure are to be implemented (i.e., these details should be fully within the purview of those skilled in the art). Where specific details (e.g., circuits) are set forth to describe exemplary embodiments of the present disclosure, it will be apparent to those skilled in the art that the embodiments of the present disclosure may be implemented without these specific details or with variations in these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0209] Although the present disclosure has been described in conjunction with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0210] The embodiments of the present disclosure are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the present disclosure. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A method for detecting abnormal events in optical fibers, characterized in that: A device for detecting abnormal optical fiber events is provided at a receiving end of a data transmission system; the method includes: receiving a sensor signal transmitted by a transmitter based on an optical fiber link, and digitizing the sensor signal to obtain a digital signal; Performing signal processing on the digital signal to obtain a signal polarization state and a signal quality indicator; Combining the signal polarization state and the signal quality indicator into initial time series data, and intercepting the initial time series data to obtain target time series data; The target time series data is input into a pre-trained anomaly detection model to obtain an abnormality probability of an optical fiber abnormal event, and a detection result of the optical fiber abnormal event is determined based on the abnormality probability.

2. The method according to claim 1, characterized in that The processing of the digital signal to obtain a signal polarization state and a signal quality indicator includes: determining a signal polarization state according to the digital signal based on a transfer function of the optical fiber link; An actual symbol sequence of the receiving end is determined from the digital signal, and a signal quality indicator is determined according to the actual symbol sequence and an ideal symbol sequence of the transmitting end.

3. The method according to claim 2, characterized in that The determining of the signal polarization state according to the digital signal based on the transfer function of the optical fiber link includes: determining a Jones vector of the light wave based on the digital signal; Using an adaptive equalizer to perform compensation processing on the digital signal based on a transfer function of the optical fiber link to obtain an equalized signal, and determining the polarization rotation of the optical fiber link by extracting a zero-frequency component from the equalized signal; Determining linear horizontal polarized light after conversion of the optical fiber link according to the polarization state rotation, and determining Stokes parameters according to the linear horizontal polarized light; The real-time position information of the Stokes parameter on the Poincare sphere is used as the signal polarization state.

4. The method according to claim 3, characterized in that The method comprises: performing compensation processing on the digital signal based on the transfer function of the optical fiber link by using an adaptive equalizer to obtain an equalized signal, and determining the polarization rotation of the optical fiber link by extracting a zero-frequency component from the equalized signal, comprising: Using an adaptive equalizer, the digital signal is compensated based on the transfer function of the optical fiber link to obtain an equalized signal. in, is the equalized signal, is the transfer function of the fiber link, is the inverse function of the transfer function of the optical fiber link, ω is the signal frequency; Determine the polarization rotation of the optical fiber link by extracting the zero-frequency component from the equalized signal, in, is the polarization rotation of the fiber link, M xx M is the ratio of horizontal polarization light to maintain horizontal polarization after passing through the optical fiber. xy is the coupling coefficient of vertically polarized light to output vertically polarized light, M yx is the coupling coefficient of horizontal polarized light to output vertical polarized light, M yy is the ratio in which vertically polarized light remains vertically polarized.

5. The method according to claim 3, characterized in that The step of determining the linear horizontal polarized light after conversion of the optical fiber link according to the polarization state rotation, and determining the Stokes parameters according to the linear horizontal polarized light, comprises: Determine the linear horizontal polarized light after the optical fiber link conversion according to the polarization state rotation, in, is the horizontally polarized light after conversion in the optical fiber link, M xx M is the ratio of horizontal polarization light to maintain horizontal polarization after passing through the optical fiber. xy is the coupling coefficient of the vertically polarized light to the output vertically polarized light; Determine the Stokes parameters based on the linear horizontal polarized light, Where S0 is the total power of the light wave, S1 is the power difference between the horizontally polarized light and the vertically polarized light in the light wave, S2 is the power difference between the ±45° linear polarization components in the light wave, and S3 is the power difference between the left and right circularly polarized light in the light wave. Re(·) represents the real signal component related to the polarization state of the extracted light wave, and Im(·) represents the imaginary signal component related to the polarization state of the extracted light wave. Indicates M xy The complex conjugate of .

6. The method according to claim 2, characterized in that Determining the actual symbol sequence of the receiving end from the digital signal, and determining a signal quality indicator according to the actual symbol sequence and the ideal symbol sequence of the transmitting end, includes: Acquire an ideal symbol sequence of the transmitting end, and determine an actual symbol sequence of the receiving end from the digital signal, and use a vector difference between the actual symbol sequence and the ideal symbol sequence as an error vector; Determine the square of the modulus of each error vector, and average the squares of the moduli of multiple error vectors to obtain a mean square error; determining an ideal symbol average power based on the ideal symbol sequence; Determining the signal quality indicator according to the mean square error and the ideal symbol average power, Where EVM is the signal quality indicator, E is the mean square error, P a is the ideal symbol average power.

7. The method according to claim 1, characterized in that Inputting the target time series data into a pre-trained anomaly detection model to obtain an abnormality probability of an optical fiber abnormal event, and determining a detection result of the optical fiber abnormal event based on the abnormality probability, includes: Inputting the target time series data into a pre-trained anomaly detection model to obtain a first anomaly probability of a fracture anomaly, a second anomaly probability of a bending anomaly, and a third anomaly probability of a vibration anomaly; Determine, according to the first abnormality probability, the second abnormality probability, and the third abnormality probability, a target abnormality event with the highest abnormality probability from among the fracture abnormality, the bending abnormality, and the vibration abnormality; The target abnormal event and the target abnormal probability corresponding to the target abnormal event are used as the detection result of the optical fiber abnormal event.

8. A device for detecting abnormal events in optical fibers, characterized in that: The device is provided at a receiving end of a data transmission system; the device comprises: a digital processing module configured to receive a sensing signal transmitted by a transmitting end based on an optical fiber link and digitally process the sensing signal to obtain a digital signal; A signal processing module is configured to perform signal processing on the digital signal to obtain a signal polarization state and a signal quality indicator; an interception processing module, configured to combine the signal polarization state and the signal quality indicator into initial time series data, and perform interception processing on the initial time series data to obtain target time series data; The anomaly detection module is configured to input the target time series data into a pre-trained anomaly detection model to obtain an anomaly probability of an optical fiber anomaly event, and determine a detection result of the optical fiber anomaly event based on the anomaly probability.

9. An electronic device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium, characterized in that The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 7.

Images