A method and device for judging optical fiber network failure based on feedback decision mechanism

Abstract

The present application relates to the field of optical fiber network technology, and in particular to an optical fiber network failure judgment method and device based on a feedback decision mechanism, which constructs an optical fiber network, and sets an intelligent optical fiber wiring device at the intermediate node of the optical fiber network, and sets a service probe at the terminal node of the optical fiber network; obtains feedback information; extracts multi-source features from the obtained feedback information, including optical power data feature extraction, OTDR data feature extraction, and service probe data feature extraction; and uses a rule-based and model-based dual failure judgment mechanism to analyze the extracted multi-source features to obtain a decision result of the current network status. Thus, the rule-based and model-based dual failure judgment mechanism comprehensively analyzes the multi-source information fed back by the optical fiber network, improves the comprehensiveness, accuracy and efficiency of failure judgment, and gives corresponding repair measures based on the fault judgment result of the decision tree model, greatly shortens the response time, and improves network availability.

Description

Technical Field

[0001] The present application relates to the field of optical fiber network technology, and in particular to an optical fiber network failure judgment method and device based on a feedback decision mechanism. Background Art

[0002] With the development of information technology, fiber-optic communication networks need to continuously expand in scale and upgrade in technology to meet the growing business needs. Modern fiber-optic communication networks are large in scale and complex in structure, and the network environment and application requirements in different regions vary greatly. Fiber-optic networks are responsible for real-time, uninterrupted data transmission in many key areas (such as financial transactions, telecommunications operations, cloud computing data centers, etc.). Once a failure occurs, timely repair is essential. In complex network topologies, such as multi-level fiber-optic networks in large data centers or multi-node interconnected networks in metropolitan area networks, the propagation and manifestation of failures are more complex and diverse.

[0003] Existing fiber-optic communication networks mainly rely on manual intervention to identify faults and implement recovery strategies, which is time-consuming and inefficient. It is difficult to accurately locate the cause of the fault and take the most effective recovery measures when dealing with complex network faults. Traditional fault alarms and optical cable fault location methods are often not accurate and timely, and rely more on manual experience and judgment. When faced with complex faults, it takes a lot of time to investigate and locate, resulting in a long interruption of the optical cable connection. Moreover, the existing repair strategies are relatively fixed and lack flexibility, making it difficult to adapt to the dynamically changing network environment and the optimal processing requirements for various types of faults. Summary of the invention

[0004] In order to overcome the deficiencies in the prior art, the present application provides a method and device for determining failure of an optical fiber network based on a feedback decision mechanism, which can collect and analyze feedback information from each node in the optical fiber network and quickly locate the root cause of the fault.

[0005] In a first aspect, the present application provides a method for determining failure of an optical fiber network based on a feedback arbitration mechanism, the method comprising the following steps:

[0006] Constructing a fiber optic network, and installing intelligent fiber optic wiring devices at the middle nodes of the fiber optic network, and installing service probes at the end nodes of the fiber optic network;

[0007] Acquiring feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber distribution device, and the flow data, packet loss rate and delay information fed back by the service probe;

[0008] Perform multi-source feature extraction on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction, and service probe data feature extraction;

[0009] A dual failure judgment mechanism based on rules and models is used to analyze the extracted multi-source features to obtain the judgment result of the current network status; wherein, the extracted multi-source features are first preliminarily screened based on the rule failure judgment mechanism, and then the abnormal situations that meet the rules preliminarily screened out are deeply analyzed based on the model failure judgment mechanism to obtain the judgment result of the current network status.

[0010] In a possible implementation manner, the extracting multi-source features from the acquired feedback information includes the following steps:

[0011] Extraction of optical power data features; the ratio of the difference between the optical power values ​​at adjacent time points to the time interval is used as the short-term rate of change of optical power; the linear regression method is used to perform segmented fitting on the optical power data to obtain the long-term trend of optical power changes; the standard deviation of the optical power data is calculated to evaluate the stability of optical power;

[0012] OTDR data feature extraction; the average loss, the location of the maximum loss point, the skewness and kurtosis of the loss distribution are calculated to obtain the fiber link loss characteristics; the reflection peak intensity, width, shape and contrast with adjacent points are analyzed to obtain the reflection peak characteristics;

[0013] Business probe data feature extraction; the cluster analysis method is used to perform pattern recognition on traffic data to obtain the traffic pattern; the difference method is used to calculate the difference in packet loss rate between adjacent time points to obtain the packet loss rate trend; the average delay, maximum and minimum delay between adjacent time points are calculated, and delay data statistics are performed to obtain the distribution of network delay over a period of time.

[0014] In a possible implementation, a rule failure judgment mechanism is constructed in the following manner, including the following steps:

[0015] Construct optical power-related rules, including combining sudden optical power drops with OTDR data to determine optical fiber link failures, and combining long-term optical power drop trends with packet loss rates to determine optical fiber aging or minor damage.

[0016] Construct OTDR-related rules, including judging the fiber quality based on the average loss of the fiber link, and judging the fiber breakage or abnormal connection based on the intensity of the reflection peak;

[0017] Construct business probe-related rules, including combining packet loss rate and delay to determine network congestion, and combining data flow and delay to determine link bandwidth bottlenecks.

[0018] In a possible implementation, a model failure judgment mechanism is constructed in the following manner, including the following steps:

[0019] A decision tree model is constructed with the optical power data features, OTDR data features, and service probe data features as inputs and the network failure judgment results and corresponding repair suggestions as outputs.

[0020] The constructed decision tree model is trained, and the trained decision tree model is used to construct a model failure judgment mechanism.

[0021] In a possible implementation, the training of the constructed decision tree model includes the following steps:

[0022] P1. Initialize the root node of the decision tree and use the training set as the data set of the root node; wherein the training set includes data under normal conditions and under optical fiber network failure conditions;

[0023] P2. Calculate the information gain ratio of each feature in the training set, and select the feature with the largest information gain ratio as the split feature of the current node; for each feature, first divide the training set into multiple subsets according to its value, and then calculate the information gain ratio according to the formula;

[0024] P3. According to the value of the selected feature, the data set of the current node is divided into multiple sub-data sets; each sub-data set corresponds to a value of the feature;

[0025] P4. For each sub-dataset, if the samples in the sub-dataset all belong to the same category, mark the node corresponding to the sub-dataset as a leaf node and assign the corresponding category label; otherwise, repeat steps P2-P4 for the sub-dataset and continue to build the subtree until the stopping condition is met; the stopping condition is that the number of samples of the node, the information gain ratio or the depth of the tree meets the set threshold.

[0026] In a possible implementation, the information gain ratio is calculated by the following formula:

[0027] Assume that the data set D, the fiber network failure set is C = {c1,c 2, …,c K}, K represents the number of categories in the fiber network failure set C, and feature A has V different values ​​{a1, a2, …, a V}; Divide the data set D into V subsets D1, D2, …, D according to the value of feature A V ; The information gain Gain(A) of feature A to data set D is:

[0028]

[0029] Among them, Ent(D) is the information entropy of data set D, and the calculation formula is:

[0030]

[0031] Among them, p k It is the data set D that belongs to category c k The proportion of samples; the information gain ratio calculation formula is:

[0032]

[0033] Among them, IV(A) is the intrinsic value of feature A, and the calculation formula is:

[0034] .

[0035] In a possible implementation, the obtained decision result of the current network status includes the fault type and the corresponding repair suggestion;

[0036] The fault types include normal, fiber optic link failure, equipment failure or network congestion; the corresponding repair suggestion for the fiber optic link failure is to build and enable a backup fiber optic link based on the intelligent fiber optic wiring device; the corresponding repair suggestion for equipment failure or network congestion is to send a notification to the maintenance personnel.

[0037] In a second aspect, the present application provides a device for determining failure of an optical fiber network based on a feedback arbitration mechanism, the device comprising:

[0038] A construction module is used to construct an optical fiber network, and to set up an intelligent optical fiber wiring device at an intermediate node of the optical fiber network, and to set up a service probe at an end node of the optical fiber network;

[0039] An acquisition module, used to acquire feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber distribution device, and the flow data, packet loss rate and delay information fed back by the service probe;

[0040] An extraction module is used to perform multi-source feature extraction on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction;

[0041] The judgment module is used to analyze the extracted multi-source features using a dual failure judgment mechanism based on rules and models to obtain a judgment result on the current network status; wherein, the extracted multi-source features are first preliminarily screened based on the rule failure judgment mechanism, and then the abnormal situations that meet the rules preliminarily screened out are deeply analyzed based on the model failure judgment mechanism to obtain a judgment result on the current network status.

[0042] In a third aspect, the present application provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the optical fiber network failure judgment method based on the feedback decision mechanism as described in any one of the first aspects are performed.

[0043] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method for determining failure of an optical fiber network based on a feedback decision mechanism as described in any one of the first aspects are executed.

[0044] The present embodiment provides a method and device for judging failure of an optical fiber network based on a feedback decision mechanism, which constructs an optical fiber network, and sets an intelligent optical fiber wiring device at the intermediate node of the optical fiber network, and sets a service probe at the terminal node of the optical fiber network; obtains feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber wiring device, and the flow data, packet loss rate and delay information fed back by the service probe; extracts multi-source features from the obtained feedback information, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction; uses a rule-based and model-based dual failure judgment mechanism to analyze the extracted multi-source features to obtain the decision result of the current network state; wherein, firstly, the extracted multi-source features are preliminarily screened based on the rule failure judgment mechanism, and then, based on the model failure judgment mechanism, the abnormal conditions that meet the rules preliminarily screened are deeply analyzed to obtain the decision result of the current network state. Thus, the multi-source information fed back by the optical fiber network is comprehensively analyzed based on the rule-based and model-based dual failure judgment mechanism, which improves the comprehensiveness, accuracy and efficiency of failure judgment, and gives corresponding repair measures according to the fault judgment result of the decision tree model, greatly shortens the response time, and improves network availability. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A flow chart of a method for determining failure of an optical fiber network based on a feedback decision mechanism according to an embodiment of the present application is shown;

[0047] Figure 2A flowchart of training a decision tree model constructed in an embodiment of the present application is shown;

[0048] Figure 3 A schematic diagram of the structure of an optical fiber network failure judgment device based on a feedback decision mechanism according to an embodiment of the present application is shown;

[0049] Figure 4 A structural block diagram of an electronic device described in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0050] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.

[0051] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0052] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.

[0053] In view of the technical problems raised by the background technology, the present application provides a method and device for determining failure of an optical fiber network based on a feedback decision mechanism, which can collect and analyze feedback information of each node in the optical fiber network and quickly locate the root cause of the fault.

[0054] In one embodiment, see the attached specification Figure 1 The present application provides a method for determining failure of an optical fiber network based on a feedback decision mechanism, the method comprising the following steps:

[0055] S1. Build an optical fiber network, set up intelligent optical fiber wiring devices at the middle nodes of the optical fiber network, and set up service probes at the end nodes of the optical fiber network;

[0056] S2. Obtain feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber distribution device, and the flow data, packet loss rate and delay information fed back by the service probe;

[0057] S3, performing multi-source feature extraction on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction;

[0058] S4. Use a dual failure judgment mechanism based on rules and models to analyze the extracted multi-source features to obtain the judgment result of the current network status; wherein, firstly perform a preliminary screening of the extracted multi-source features based on the rule failure judgment mechanism, and then perform an in-depth analysis of the abnormal situations that meet the rules based on the preliminary screening based on the model failure judgment mechanism to obtain the judgment result of the current network status.

[0059] Specifically, in step S1, in order to clearly understand the technical solution of the embodiment of the present application, the application scenario of the optical fiber network can be illustrated first. In an optical fiber network, there are end nodes and intermediate nodes. The end node is usually at the user end or the terminal service side, and is the starting point for the transmission of business monitoring data; the intermediate node is usually at the access layer and the aggregation layer. In an optical fiber network with a multi-layer topology, the end node sends data to the intermediate node of the upper level, and the intermediate node integrates and forwards the data of multiple end nodes. The intermediate node can also serve as a backup path. When the direct link between the lower-level node and other nodes (such as the aggregation node) fails, the intermediate node can provide an alternative transmission path to ensure the transmission of data. In the network topology, the intermediate node is connected to multiple end nodes and other intermediate nodes, so that the network forms a more complex and robust connection system. Therefore, when the optical fiber link in a certain area fails, there is a possibility that the optical fiber path can be re-established through the intermediate node to ensure that the data can be smoothly transmitted to the central server or other nodes.

[0060] In this application, in order to collect and analyze the feedback information of each node in the optical fiber network, an intelligent optical fiber distribution device is set at the intermediate node, and a service probe is set at the end node. The intelligent optical fiber distribution device connects the optical cable cores from the intermediate nodes or end nodes at different levels to form a multi-layer topology optical fiber network. The intelligent optical fiber distribution device monitors the optical fiber link status through optical power and OTDR data, and accepts instructions from the intelligent optical fiber distribution management platform to complete the connection of optical cable cores in different directions, and realizes the automatic construction of the optical fiber path of the intermediate node; the service probe is used to detect and analyze the working status of the end node service equipment and the service data transmission performance with the service end.

[0061] In step S2, feedback information is obtained through an intelligent optical fiber distribution device and a service probe arranged in the optical fiber network.

[0062] The information fed back by the intelligent fiber optic wiring device includes the optical power monitoring value and the link connection status (OTDR test data). Optical power monitoring value: monitors the power of the optical signal in the optical fiber. A weak optical signal may be caused by excessive optical fiber loss, poor connection, or equipment failure. The built-in optical power monitoring module of the intelligent fiber optic wiring device monitors the power of the optical signal received in the optical fiber. When the optical signal is transmitted in the optical fiber, its power will be affected by many factors, such as optical fiber loss, connector loss, splitter loss, etc. In actual applications, if the optical power from the intelligent fiber optic wiring device to a certain optical cable core drops abnormally, it means that the optical fiber link connected to this intelligent fiber optic wiring device may have problems such as breakage, excessive bending, or loose connectors. Intelligent fiber optic wiring devices are deployed on a large scale in the optical cable network. According to the network topology, the overall performance of the optical fiber network can be further monitored to detect potential fault hazards in a timely manner;

[0063] Link connection status (OTDR test data): Check whether the connection of the optical fiber link is normal, whether it is loose, broken, etc. The built-in OTDR test module of the intelligent optical fiber distribution device sends a light pulse to the optical fiber under test. The optical pulse will have an optical signal reflected back to the OTDR on the optical fiber itself and each characteristic point. The OTDR module collects the reflected optical signal to form an OTDR test curve. The intelligent optical fiber distribution device analyzes the OTDR test curve to obtain the optical fiber length and the corresponding attenuation along the length of the optical cable. Based on this information, the fault location and type of optical fiber breakage and excessive loss can be analyzed and determined.

[0064] The service probes installed at the end nodes are used to detect the link status from the application level through the transmission of service test data between the service probes and the server side, including parameters such as bit error rate, signal loss, and transmission delay. These parameters reflect the actual performance of the optical fiber link during data transmission, including traffic data, packet loss rate, and delay information. Traffic data: Statistics on the number, size, and transmission rate of data packets within a time period, and calculate indicators such as the average transmission rate, maximum data packet size, and total number of data packets within the time period, which are used to analyze dynamic changes in network load. For example, in a fiber optic network of a data center, service probes can monitor data traffic between different servers and understand the traffic distribution in each area;

[0065] Packet loss rate: This is the ratio of packets lost during transmission, which is an important indicator for judging the quality of network links. The moving average method is used to calculate the average packet loss rate in the past T seconds to avoid misjudgment caused by instantaneous network fluctuations. If the packet loss rate of a link continues to increase, it may indicate that the link has a potential fault.

[0066] Latency: Business probes measure the transmission time of data packets from end nodes to server nodes by adding timestamps to data packets. High latency may lead to degradation of network application performance, such as video freezes and concurrent packet loss of large data.

[0067] In step S3, multi-source feature extraction is mainly performed on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction.

[0068] The extracted optical power data features include the optical power change trend and optical power stability. Among them, the optical power change trend includes the calculation of the short-term change rate and long-term change trend of the optical power. The short-term change rate is the ratio of the difference between the optical power values ​​at adjacent time points (for example, 10 seconds apart) to the time interval, reflecting the instantaneous fluctuation of the optical power, which may indicate sudden optical fiber link problems (such as the optical fiber being squeezed instantly, causing the optical power to drop); the long-term change trend uses the linear regression method to perform segmented fitting on the optical power data to obtain the trend curve of the optical power changing with time, judge the long-term change trend of the optical power, and discover the aging of the optical fiber or the gradual loss problem (such as the trend line slope is negative, indicating that the optical power is gradually decreasing);

[0069] The stability of optical power is expressed by calculating the standard deviation of the optical power data. The smaller the standard deviation, the more stable the optical power. Unstable optical power may be caused by loose fiber connections, equipment power fluctuations, etc. At the same time, analyze the fluctuation frequency of optical power (such as calculating the main frequency components in the optical power data through Fourier transform). High-frequency fluctuations may be related to equipment failure or electromagnetic interference.

[0070] The extracted OTDR data features include fiber link loss features and reflection peak features. Among them, for the fiber link loss features, in addition to calculating the average loss and the location of the maximum loss point, the loss distribution is further analyzed. The skewness and kurtosis of the loss are calculated. The skewness can reflect the asymmetry of the loss distribution, and the kurtosis can reflect the concentration of the loss distribution. For example, a positively skewed loss distribution may indicate that there are local severe loss points in the fiber link; a high peak may mean that the loss is concentrated in certain specific locations and requires special attention;

[0071] For the reflection peak in the OTDR test curve, we not only pay attention to its intensity, but also analyze the width and shape of the reflection peak. A wider reflection peak may indicate a larger gap or unevenness at the fiber connection point; an irregularly shaped reflection peak may indicate damage to the fiber end face or contamination of the connector. At the same time, calculate the contrast between the reflection peak and the adjacent points (the ratio of the reflection peak intensity to the average intensity of the surrounding fiber segments). A high-contrast reflection peak is more likely to be a manifestation of a fiber break or serious connection problem.

[0072] The extracted business probe data features include traffic patterns, packet loss rate trend features, and delay data statistical features. For traffic patterns, cluster analysis and other methods are used to perform pattern recognition on traffic data. The traffic data is divided into peak traffic, normal traffic, and low-peak traffic patterns, and the frequency and duration of different patterns are analyzed. High traffic peaks may indicate that the network is under a heavy load during this period, while frequent traffic changes may indicate that the network application is highly dynamic.

[0073] Packet loss rate trend characteristics: The difference method is used to calculate the difference in packet loss rates at adjacent time points, and the changing trend of the packet loss rate is calculated. The continuously rising trend of packet loss rate indicates that the quality of the network link is deteriorating, and the fault needs to be promptly checked;

[0074] Statistical characteristics of delay data: Calculate the average delay between adjacent time points, the maximum and minimum delays, and the distribution of network delays over a period of time to evaluate the stability and anomalies of network delays.

[0075] In step S4, the present application adopts a dual failure judgment mechanism based on rules and models to analyze the extracted multi-source features and obtain the judgment result of the current network status. That is, on the one hand, a series of intuitive threshold rules based on optical power, OTDR, business probe and other data are set to quickly conduct preliminary screening of obvious fault signs; on the other hand, a multi-source information decision tree model is constructed, and the information gain ratio is used to select the optimal feature splitting node. Through training of historical data, it can learn the complex relationship between different feature combinations and network failure conditions, and realize in-depth judgment of complex fault scenarios. The two work together, rule judgment provides pre-screening for decision trees, and decision trees make up for complex and fuzzy situations that rules are difficult to deal with. The advantages complement each other, greatly improving the comprehensiveness, accuracy and efficiency of failure judgment, effectively reducing the probability of misjudgment and missed judgment, and greatly improving the accuracy and reliability of failure judgment.

[0076] Specifically, the rule-based failure judgment includes the set optical power related rules, OTDR related rules and service probe related rules. Among them, the optical power related rules: if the optical power suddenly drops by more than the preset threshold OP min , and the duration exceeds the preset time T OPM (For example, the optical power suddenly drops by 3dB and lasts for more than 30s). Combined with the OTDR test data to see if the link has increased loss or abnormal reflection peaks, it can be judged that the optical fiber link may be broken or seriously damaged. If the optical power shows a downward trend for a long time (preset time TOPL) (for example, the optical power drops by more than 1dB within 10 consecutive minutes), and the service probe feedback shows that the packet loss rate of the link begins to increase, it may indicate that the optical fiber is slowly aging or slightly damaged, and further inspection is required.

[0077] OTDR-related rules: When the OTDR test curve shows that the loss of a certain section of the optical fiber link exceeds the preset average chain loss threshold (such as 0.5dB / km, set according to the optical fiber type and standard), and the length of this section exceeds a certain length (such as 100 meters), it is judged that there is a quality problem with this section of optical fiber and it may need to be replaced or repaired; if the OTDR detects an obvious reflection peak, and the reflection peak intensity far exceeds the reflection peak intensity of the normal connection point (for example, the reflection peak intensity is 5 times that of the normal connection point), it is judged that there is a fiber break or abnormal connection here;

[0078] Business probe related rules: When the packet loss rate continues to be higher than the threshold set by the network application requirements (for example, 5%) for a period of time (for example, 30 seconds), and the delay exceeds the set threshold (for example, 200ms, determined according to the real-time requirements), and the optical power monitoring does not find obvious abnormalities, it may be network congestion or server-side problems, which require further investigation; if the traffic data shows that the traffic of a link is close to or exceeds its bandwidth limit for a long time (such as 5 minutes), and the packet loss rate begins to rise and the delay increases, it is judged that the link may have a bandwidth bottleneck, and it is necessary to consider traffic optimization or link upgrade.

[0079] Secondly, based on the failure judgment of the model, the decision tree model is constructed with the optical power data features, OTDR data features and service probe data features as input, and the judgment results of network failure and corresponding repair suggestions as output. Each node of the decision tree is split according to a feature. For example, the root node can be split according to whether the short-term change rate of optical power exceeds the threshold, and the branch node can continue to split according to the skewness of OTDR loss, the temperature change rate of the equipment, and other features. The leaf nodes of the decision tree represent the judgment results of network failure (such as optical fiber link failure, equipment failure, network congestion, etc.) and the corresponding repair suggestions (such as switching backup links, restarting equipment, optimizing server configuration, etc.). The decision tree model is trained with a large amount of historical data (including data under normal and fault conditions) so that the model can learn the relationship between different feature combinations and network failure conditions.

[0080] See the instruction manual Figure 2 In one embodiment, the constructed decision tree model is trained in the following manner, including the following steps:

[0081] P1. Initialize the root node of the decision tree and use the training set as the data set of the root node; wherein the training set includes data under normal conditions and under optical fiber network failure conditions;

[0082] P2. Calculate the information gain ratio of each feature in the training set, and select the feature with the largest information gain ratio as the split feature of the current node; for each feature, first divide the training set into multiple subsets according to its value, and then calculate the information gain ratio according to the formula; for example, if it is calculated that the short-term change rate feature of optical power has the largest information gain ratio, then use this feature as the split feature of the root node;

[0083] P3. Divide the data set of the current node into multiple sub-data sets according to the value of the selected feature; each sub-data set corresponds to a value of the feature; for example, if the short-term change rate of optical power has three value ranges (low, medium, and high), the training set is divided into three sub-data sets;

[0084] P4. For each sub-dataset, if the samples in the sub-dataset all belong to the same category, mark the node corresponding to the sub-dataset as a leaf node and assign the corresponding category label; otherwise, repeat steps P2-P4 for the sub-dataset and continue to build the subtree until the stopping condition is met. The stopping condition can be: (1) When the number of samples in the node is less than a certain threshold, stop splitting. For example, when the number of samples in a node is less than 10, it will no longer be split and the node will be marked as a leaf node. The category label can be determined based on the majority category of the samples in the node; (2) When the information gain ratio is less than a certain threshold, stop splitting. This means that continuing to split the node will not improve the classification accuracy much, and the node will be marked as a leaf node at this time; (3) When the depth of the tree reaches the preset maximum depth, stop splitting. For example, set the maximum depth of the decision tree to 5 layers. When the depth of the tree reaches 5 layers, stop splitting regardless of whether other conditions are met and mark the current node as a leaf node.

[0085] Among them, the information gain ratio is calculated by the following formula:

[0086] Assume that the data set D, the fiber network failure set is C = {c1,c 2, …,c K}, K represents the number of categories in the fiber network failure set C, and feature A has V different values ​​{a1, a2, …, a V}; Divide the data set D into V subsets D1, D2, …, D according to the value of feature A V ; The information gain Gain(A) of feature A to data set D is:

[0087]

[0088] Among them, ENT(D) is the information entropy of data set D, and the calculation formula is:

[0089]

[0090] Among them, p k It is the data set D that belongs to category c k The proportion of samples; the information gain ratio calculation formula is:

[0091]

[0092] Among them, IV(A) is the intrinsic value of feature A, and the calculation formula is:

[0093] .

[0094] When the multi-source information decision tree model is trained, it will be deployed on the server side with the above rules. The intelligent fiber optic wiring device and service probe will send the collected data to the server side. The extracted multi-source features will be preliminarily screened based on the rule failure judgment mechanism, and then the multi-source features corresponding to the abnormal conditions that meet the rules will be input into the trained decision tree model. The decision tree model makes judgments on the nodes of the tree based on these features. Starting from the root node, the data will be divided into different branches according to the splitting features of each node (such as whether the short-term change rate of optical power exceeds a certain threshold) until it reaches the leaf node. The category label of the leaf node (such as "normal", "fiber link failure", "equipment failure" or "network congestion") outputs the decision tree's decision result on the current network status.

[0095] Furthermore, according to the result of the ruling, the corresponding network self-healing repair measures are automatically triggered. Including, if it is judged that the optical fiber link is faulty, the system calculates the backup optical fiber link through the network topology, completes the connection of the optical fiber link through the intelligent optical fiber wiring device set at different location nodes, builds and enables the backup optical fiber link, and notifies the maintenance personnel to carry out on-site repairs; if it is a device failure or network congestion, the maintenance personnel are notified in the form of a work order to optimize the network configuration. At the same time, data is continuously collected and analyzed to verify the effectiveness of the repair measures until the network returns to normal operation.

[0096] It can be seen that the optical fiber network failure judgment method based on the feedback judgment mechanism provided in the present application analyzes the intensity changes of signal transmission through optical power data, accurately locates the physical loss and connection status of the optical fiber link through OTDR data, reflects the actual performance of the network from the application level through business probe data, and performs feature extraction and multi-source data collaborative analysis based on optical power data, thereby avoiding misjudgment or missed judgment that may be caused by a single data source, and realizing all-round monitoring of the optical fiber network. In addition, a dual failure judgment mechanism based on rules and models is adopted, which greatly improves the comprehensiveness, accuracy and efficiency of failure judgment, effectively reduces the probability of misjudgment and missed judgment, and greatly improves the accuracy and reliability of failure judgment. In addition, according to the fault judgment result, the corresponding precise repair measures can be automatically triggered, which greatly shortens the response time, improves network availability, and realizes intelligent and automated repair.

[0097] Based on the same inventive concept, an optical fiber network failure judgment device based on a feedback decision mechanism is also provided in an embodiment of the present application. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned optical fiber network failure judgment method based on a feedback decision mechanism in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0098] As the instruction manual Figure 3As shown, an embodiment of the present application provides a device for determining failure of an optical fiber network based on a feedback decision mechanism, the device comprising:

[0099] Construction module 301, used to construct an optical fiber network, and set up an intelligent optical fiber wiring device at an intermediate node of the optical fiber network, and set up a service probe at an end node of the optical fiber network;

[0100] The acquisition module 302 is used to acquire feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber distribution device, and the flow data, packet loss rate and delay information fed back by the service probe;

[0101] An extraction module 303 is used to perform multi-source feature extraction on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction;

[0102] The judgment module 304 is used to analyze the extracted multi-source features using a dual failure judgment mechanism based on rules and models to obtain a judgment result of the current network status; wherein, the extracted multi-source features are first preliminarily screened based on the rule failure judgment mechanism, and then the abnormal conditions that meet the rules preliminarily screened out are deeply analyzed based on the model failure judgment mechanism to obtain a judgment result of the current network status. The judgment result of the current network status obtained includes the fault type and the corresponding repair suggestion; the fault type includes normal, fiber link failure, equipment failure or network congestion; the corresponding repair suggestion for the fiber link failure is to build and enable a spare fiber link based on the intelligent fiber optic wiring device; the corresponding repair suggestion for equipment failure or network congestion is to send a notification to the maintenance personnel.

[0103] In one embodiment, the extraction module 303 performs multi-source feature extraction on the acquired feedback information, including: optical power data feature extraction; wherein the ratio of the difference between the optical power values ​​at adjacent time points and the time interval is used as the short-term change rate of the optical power; a linear regression method is used to perform segmented fitting on the optical power data to obtain the long-term change trend of the optical power; the standard deviation of the optical power data is calculated to evaluate the stability of the optical power; OTDR data feature extraction; wherein the average loss, the position of the maximum loss point, the skewness and kurtosis of the loss distribution are calculated to obtain the loss characteristics of the optical fiber link; the intensity, width, shape and contrast of the reflection peak with adjacent points are analyzed to obtain the reflection peak characteristics; service probe data feature extraction; wherein a cluster analysis method is used to perform pattern recognition on the traffic data to obtain the traffic pattern; a differential method is used to calculate the difference in packet loss rate at adjacent time points to obtain the packet loss rate trend; the average delay, the maximum and minimum delay values ​​are calculated to perform delay data statistics.

[0104] In one embodiment, the device further comprises:

[0105] A construction module is used to construct rules related to optical power, including judging optical fiber link faults by combining a sudden drop in optical power with OTDR data, and judging optical fiber aging or slight damage by combining a long-term drop in optical power with a packet loss rate; constructing OTDR related rules, including judging optical fiber quality based on the average loss of the optical fiber link, and judging optical fiber breakage or abnormal connection based on the intensity of the reflection peak; constructing business probe related rules, including judging network congestion by combining packet loss rate and delay, and judging link bandwidth bottleneck by combining data flow and delay.

[0106] In one embodiment, the device further comprises:

[0107] The construction module is used to construct a decision tree model with optical power data features, OTDR data features and service probe data features as inputs and network failure judgment results and corresponding repair suggestions as outputs; the constructed decision tree model is trained, and the trained decision tree model is used to construct a model failure judgment mechanism.

[0108] In one embodiment, the construction module trains the constructed decision tree model, including: P1, initializing the root node of the decision tree, and using the training set as the data set of the root node; wherein the training set includes data under normal conditions and under optical fiber network failure conditions; P2, calculating the information gain ratio of each feature in the training set, and selecting the feature with the largest information gain ratio as the split feature of the current node; wherein, for each feature, first divide the training set into multiple subsets according to its value, and then calculate the information gain ratio according to a formula; P3, according to the value of the selected feature, divide the data set of the current node into multiple sub-datasets; wherein each sub-dataset corresponds to a value of the feature; P4, for each sub-dataset, if the samples in the sub-dataset all belong to the same category, mark the node corresponding to the sub-dataset as a leaf node and assign a corresponding category label; otherwise, repeat steps P2-P4 for the sub-dataset, and continue to build the subtree until the stopping condition is met; wherein the stopping condition is that the number of samples of the node, the information gain ratio or the depth of the tree meets the set threshold.

[0109] The present application provides a fiber optic network failure judgment device based on a feedback decision mechanism. The fiber optic network is constructed by a construction module, and an intelligent fiber optic distribution device is set at the intermediate node of the fiber optic network, and a service probe is set at the end node of the fiber optic network; feedback information is obtained by an acquisition module, including optical power monitoring values ​​and OTDR data information fed back by the intelligent fiber optic distribution device, and traffic data, packet loss rate and delay information fed back by the service probe; multi-source feature extraction is performed on the obtained feedback information by an extraction module, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction; the extracted multi-source features are analyzed by a judgment module using a dual failure judgment mechanism based on rules and models to obtain a decision result of the current network status; wherein, the extracted multi-source features are first preliminarily screened based on the rule failure judgment mechanism, and then the abnormal situations that meet the rules preliminarily screened out are deeply analyzed based on the model failure judgment mechanism to obtain a decision result of the current network status. Therefore, the dual failure judgment mechanism based on rules and models comprehensively analyzes the multi-source information fed back by the optical fiber network, improves the comprehensiveness, accuracy and efficiency of failure judgment, and gives corresponding repair measures based on the fault judgment results of the decision tree model, greatly shortens the response time, improves the comprehensive analysis of network availability information dimensions, and realizes more accurate optical fiber network failure judgment based on the feedback judgment mechanism.

[0110] Based on the same concept of the present invention, the specification is attached Figure 4 As shown, the structure of an electronic device 400 provided in an embodiment of the present application includes: at least one processor 401, at least one network interface 404 or other user interface 403, a memory 405, and at least one communication bus 402. The communication bus 402 is used to realize the connection and communication between these components. The electronic device 400 optionally includes a user interface 403, including a display (for example, a touch screen, LCD, CRT, holographic imaging (Holographic) or projection (Projector), etc.), a keyboard or a pointing device (for example, a mouse, a trackball (trackball), a touch pad or a touch screen, etc.).

[0111] The memory 405 may include a read-only memory and a random access memory, and provides instructions and data to the processor 401. A portion of the memory 405 may also include a non-volatile random access memory (NVRAM).

[0112] In some implementations, the memory 405 stores the following elements, executable modules or data structures, or a subset thereof, or an extended set thereof:

[0113] Operating system 4051, including various system programs for implementing various basic services and processing hardware-based tasks;

[0114] The application module 4052 includes various application programs, such as a launcher, a media player, a browser, etc., which are used to implement various application services.

[0115] In an embodiment of the present application, by calling the program or instructions stored in the memory 405, the processor 401 is used to execute steps in a fiber optic network failure judgment method based on a feedback decision mechanism, which can collect and analyze feedback information from each node in the fiber optic network and quickly locate the root cause of the fault.

[0116] The present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the method for determining failure of an optical fiber network based on a feedback decision mechanism are executed.

[0117] Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, etc. When the computer program on the storage medium is executed, the above-mentioned optical fiber network failure judgment method based on the feedback decision mechanism can be executed.

[0118] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0119] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0120] In addition, each functional unit in the embodiments provided in the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0121] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0122] Finally, it should be noted that the above embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the aforementioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the aforementioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application. They should all be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for determining failure of an optical fiber network based on a feedback decision mechanism, characterized in that: The method comprises the following steps: Constructing a fiber optic network, and installing intelligent fiber optic wiring devices at the middle nodes of the fiber optic network, and installing service probes at the end nodes of the fiber optic network; Acquiring feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber distribution device, and the flow data, packet loss rate and delay information fed back by the service probe; Perform multi-source feature extraction on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction, and service probe data feature extraction; A dual failure judgment mechanism based on rules and models is used to analyze the extracted multi-source features to obtain the judgment result of the current network status; wherein, the extracted multi-source features are first preliminarily screened based on the rule failure judgment mechanism, and then the abnormal situations that meet the rules preliminarily screened out are deeply analyzed based on the model failure judgment mechanism to obtain the judgment result of the current network status.

2. The method for determining failure of an optical fiber network based on a feedback decision mechanism according to claim 1, characterized in that: The multi-source feature extraction of the acquired feedback information comprises the following steps: Extraction of optical power data features; the ratio of the difference between the optical power values ​​at adjacent time points to the time interval is used as the short-term rate of change of optical power; the linear regression method is used to perform segmented fitting on the optical power data to obtain the long-term trend of optical power changes; the standard deviation of the optical power data is calculated to evaluate the stability of optical power; OTDR data feature extraction; the average loss, the location of the maximum loss point, the skewness and kurtosis of the loss distribution are calculated to obtain the fiber link loss characteristics; the reflection peak intensity, width, shape and contrast with adjacent points are analyzed to obtain the reflection peak characteristics; Business probe data feature extraction; the cluster analysis method is used to perform pattern recognition on traffic data to obtain the traffic pattern; the difference method is used to calculate the difference in packet loss rate between adjacent time points to obtain the packet loss rate trend; the average delay, maximum and minimum delay between adjacent time points are calculated, and delay data statistics are performed to obtain the distribution of network delay over a period of time.

3. The method for determining failure of an optical fiber network based on a feedback decision mechanism according to claim 2, characterized in that: The rule failure judgment mechanism is constructed in the following way, including the following steps: Construct optical power-related rules, including combining sudden optical power drops with OTDR data to determine optical fiber link failures, and combining long-term optical power drop trends with packet loss rates to determine optical fiber aging or minor damage. Construct OTDR-related rules, including judging the fiber quality based on the average loss of the fiber link, and judging the fiber breakage or abnormal connection based on the intensity of the reflection peak; Construct business probe-related rules, including combining packet loss rate and delay to determine network congestion, and combining data flow and delay to determine link bandwidth bottlenecks.

4. The method for determining failure of an optical fiber network based on a feedback decision mechanism according to claim 3, characterized in that: The model failure judgment mechanism is constructed in the following way, including the following steps: A decision tree model is constructed with the optical power data features, OTDR data features, and service probe data features as inputs and the network failure judgment results and corresponding repair suggestions as outputs. The constructed decision tree model is trained, and the trained decision tree model is used to construct a model failure judgment mechanism.

5. The method for determining failure of an optical fiber network based on a feedback decision mechanism according to claim 4, characterized in that: The training of the constructed decision tree model comprises the following steps: P1. Initialize the root node of the decision tree and use the training set as the data set of the root node; wherein the training set includes data under normal conditions and under optical fiber network failure conditions; P2. Calculate the information gain ratio of each feature in the training set, and select the feature with the largest information gain ratio as the split feature of the current node; for each feature, first divide the training set into multiple subsets according to its value, and then calculate the information gain ratio according to the formula; P3. According to the value of the selected feature, the data set of the current node is divided into multiple sub-data sets; each sub-data set corresponds to a value of the feature; P4. For each sub-dataset, if the samples in the sub-dataset all belong to the same category, mark the node corresponding to the sub-dataset as a leaf node and assign the corresponding category label; otherwise, repeat steps P2-P4 for the sub-dataset and continue to build the subtree until the stopping condition is met; the stopping condition is that the number of samples of the node, the information gain ratio or the depth of the tree meets the set threshold.

6. The method for determining failure of an optical fiber network based on a feedback arbitration mechanism according to claim 5, characterized in that: The information gain ratio is calculated by the following formula: Assume that the data set D, the fiber network failure set is C = {c1,c 2, …,c K }, K represents the number of categories in the fiber network failure set C, and feature A has V different values ​​{a1, a2, …, a V }; Divide the data set D into V subsets D1, D2, …, D according to the value of feature A V ; The information gain Gain(A) of feature A to data set D is: Among them, Ent(D) is the information entropy of data set D, and the calculation formula is: Among them, p k It is the data set D that belongs to category c k The proportion of samples; the information gain ratio calculation formula is: Among them, IV(A) is the intrinsic value of feature A, and the calculation formula is: 。 7. The method for determining failure of an optical fiber network based on a feedback arbitration mechanism according to claim 6, characterized in that: in, The judgment result of the current network status includes the fault type and the corresponding repair suggestions; The fault type includes normal, optical fiber link failure, equipment failure or network congestion; The corresponding repair suggestion for fiber optic link failure is to build and enable a backup fiber optic link based on the intelligent fiber optic wiring device; the corresponding repair suggestion for equipment failure or network congestion is to send a notification to the maintenance personnel.

8. An optical fiber network failure judgment device based on a feedback decision mechanism, characterized in that: The device comprises: A construction module is used to construct an optical fiber network, and to set up an intelligent optical fiber wiring device at an intermediate node of the optical fiber network, and to set up a service probe at an end node of the optical fiber network; An acquisition module, used to acquire feedback information, including the optical power monitoring value and OTDR data information fed back by the intelligent optical fiber distribution device, and the flow data, packet loss rate and delay information fed back by the service probe; An extraction module is used to perform multi-source feature extraction on the acquired feedback information, including optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction; The judgment module is used to analyze the extracted multi-source features using a dual failure judgment mechanism based on rules and models to obtain a judgment result on the current network status; wherein, the extracted multi-source features are first preliminarily screened based on the rule failure judgment mechanism, and then the abnormal situations that meet the rules preliminarily screened out are deeply analyzed based on the model failure judgment mechanism to obtain a judgment result on the current network status.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the optical fiber network failure judgment method based on the feedback decision mechanism as described in any one of claims 1 to 7 are performed.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the optical fiber network failure judgment method based on the feedback decision mechanism as claimed in any one of claims 1 to 7 are executed.

Images