WO2024001206A1 - Optical network health monitoring method, management unit, system and storage medium - Google Patents

Abstract

The embodiments of the present application provide an optical network health monitoring method, a management unit, a system and a storage medium. The optical network health monitoring method comprises: acquiring an optical network health performance prediction instruction (S100); according to the optical network health performance prediction instruction, acquiring optical network historical performance data within a first time period (S200); and obtaining optical network prediction performance data according to the optical network historical performance data and an optical network prediction model (S300).

Description

Optical network health monitoring method, management unit, system and storage medium

Cross-references to related applications

This application is filed based on a Chinese patent application with application number 202210750234.2 and a filing date of June 29, 2022, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is hereby incorporated by reference into this application.

Technical field

Embodiments of the present application relate to the field of optical network communication technology, and in particular to optical network health monitoring methods, management units, systems and storage media.

Background technique

In OTN (Optical Transport Network), in order to ensure the normal operation of large-capacity and high-concurrency services, engineers need to spend a lot of time, manpower, and resources on the operation, maintenance and processing of optical fibers before the equipment is put into operation. For example, to detect OTS (Optical Transmission Section, optical transmission section) optical fiber and OCh (Optical Channel layer, optical channel layer) channel faults, detect OTS through optical power meters, and detect OCh channels through bit error meters to ensure that the optical The network is running healthily. However, even if the project acceptance requirements are met, after being put into use, various operations and maintenance of the network, such as external force-triggered failures, loose connections, connection module failures, fiber core failures, etc., will cause the optical network to fail to operate normally.

In related technologies, in the case of a slow-down optical fiber failure, there is a period of performance degradation for a certain period of time but it has not yet had a major impact on the business, so it is not easy to identify the degradation trend in the early stage. Usually, the business is interrupted and complaints are raised. As a result, the operation and maintenance engineers are in a passive processing state for a long time. When the problem is discovered, large losses may have already occurred.

Contents of the invention

The following is an overview of the topics described in detail in this article. This summary is not intended to limit the scope of the claims.

Embodiments of the present application provide optical network health monitoring methods, management units, systems and storage media.

In a first aspect, embodiments of the present application provide an optical network health monitoring method, which includes: obtaining an optical network health performance prediction instruction; and obtaining optical network historical performance data within a first time period according to the optical network health performance prediction instruction; According to the optical network historical performance data and the optical network prediction model, the optical network prediction performance data is obtained.

In a second aspect, embodiments of the present application provide an optical network management unit, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following is implemented: The optical network health monitoring method described in the first aspect.

In a third aspect, embodiments of the present application provide an optical network system, including: an optical network management unit as described in the second aspect; a source endpoint device connected to the optical network management unit for sending or receiving optical network services flow; a plurality of intermediate node devices connected to the optical network management unit for forwarding optical network service flows; sink endpoint devices connected to the optical network management unit for receiving or sending optical network service flows.

In a fourth aspect, embodiments of the present application provide a computer-readable storage medium that stores computer-executable instructions, and the computer-executable instructions are used to: execute the optical network health monitoring method described in the first aspect. It can be understood that the beneficial effects of the above-mentioned second to fourth aspects compared with the related art are the same as the beneficial effects of the above-mentioned first aspect compared with the related art. Please refer to the relevant description in the above-mentioned first aspect. This will not be described again.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings in the following description are only embodiments of the present application. For some embodiments, those of ordinary skill in the art can also obtain other drawings based on these drawings without exerting any creative effort.

Figure 1 is a schematic diagram of a system architecture for executing an optical network health monitoring method according to an embodiment of the present application;

Figure 2 is a schematic diagram of a system architecture for executing an optical network health monitoring method according to another embodiment of the application;

Figure 3 is a schematic flow chart of an optical network health monitoring method provided by an embodiment of the present application;

Figure 4 is a schematic diagram of the original performance file in an embodiment of the present application;

Figure 5 is a schematic diagram of a performance data structure file provided by an embodiment of the present application;

Figure 6 is a schematic diagram of a resource data structure file provided by an embodiment of the present application;

Figure 7 is a schematic diagram of a file after performance conversion and resource merging provided by an embodiment of the present application;

Figure 8 is a schematic flow chart of an optical network health performance prediction algorithm provided by an embodiment of the present application;

Figure 9 is a schematic polyline diagram of original training data provided by an embodiment of the present application;

Figure 10 is a schematic polyline diagram of an approximate partial prediction result provided by an embodiment of the present application;

Figure 11 is a broken line schematic diagram of the details provided by an embodiment of the present application;

Figure 12 is a schematic diagram of the final polyline after fusion provided by an embodiment of the present application;

Figure 13 is a schematic flow chart of an optical network health monitoring method provided by another embodiment of the present application;

Figure 14 is a schematic diagram of the optical network health creation prediction task provided by an embodiment of the present application;

Figure 15 is a schematic diagram of an optical network health analysis creation and prediction task provided by an embodiment of the present application;

Figure 16 is a schematic diagram of the optical network health analysis startup prediction task provided by an embodiment of the present application;

Figure 17 is a schematic diagram of the results of a scheduled query task provided by an embodiment of the present application;

Figure 18 is a schematic diagram of performance data backup provided by an embodiment of the present application;

Figure 19 is a schematic flow chart of an optical network health monitoring method provided by another embodiment of the present application;

Figure 20 is a schematic diagram of service-based PRBS test management provided by an embodiment of the present application;

Figure 21 is a schematic flow chart of an optical network health monitoring method provided by another embodiment of the present application;

Figure 22 is a schematic flow chart of an optical network health monitoring method provided by another embodiment of the present application;

Figure 23 is a schematic diagram of service PRBS testing provided by an embodiment of the present application;

Figure 24 is a schematic flow chart of an optical network health monitoring method provided by another embodiment of the present application;

Figure 25 is a schematic diagram of the optical network health interface provided by an embodiment of the present application;

Figure 26 is a schematic diagram of an optical network healthy optical fiber link early warning provided by an embodiment of the present application;

Figure 27 is a schematic flow chart of an optical network health monitoring method provided by another embodiment of the present application;

Figure 28 is a schematic flowchart of an optical network health monitoring method provided by another embodiment of the present application.

Detailed ways

In the following description, details such as specific system structures and technologies are provided for the purpose of explanation rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the embodiments of the present application may be implemented in other embodiments without these details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the embodiments of the present application with unnecessary detail.

It should be noted that although a logical sequence is shown in the flowchart, in some cases, the steps shown or described may be performed in an order different from that in the flowchart. The terms "first", "second", etc. in the description, claims, and above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific sequence or sequence.

It should also be understood that reference to "one embodiment" or "some embodiments" or the like in the description of the embodiments of the present application means that the specific features described in connection with the embodiment are included in one or more embodiments of the embodiments of the present application. , structure or characteristics. Therefore, the phrases "in one embodiment", "in some embodiments", "in other embodiments", "in other embodiments", etc. appearing in different places in this specification are not necessarily References are made to the same embodiment, but rather to "one or more but not all embodiments" unless specifically stated otherwise. The terms “including,” “includes,” “having,” and variations thereof all mean “including but not limited to,” unless otherwise specifically emphasized.

With the explosive growth of global data traffic, emerging services represented by video and streaming media services have developed rapidly, making dynamic, high-bandwidth and high-quality data services the main body of network traffic, and driving the network to evolve towards packetization. In terms of transmission network, it can be seen that the traditional SDH (Synchronous Digital Hierarchy, synchronous digital system) circuit-switched network has developed to the MSTP (Multi-Service Transfer Platform) with multi-service access function, a multi-service transmission platform based on SDH ), and gradually evolved to OTN and PTN (Packet Transport Network).

Statistics show that OTS (Optical Transmission Section, optical transmission section) optical fiber failure and OCh (Optical Channel layer, optical channel layer) channel failure are the two major categories of OTN network failures. The operation and maintenance and exception handling of optical fiber and OCh channels are two important factors in maintaining OTN networks.

In the OTN network, in order to ensure the normal operation of large-capacity and high-concurrency services, engineers need to spend a lot of time, manpower, and resources on the operation, maintenance and processing of optical fibers before the equipment is put into operation, such as OTS (Optical Transmission Section, Optical transmission section) fiber and OCh (Optical Channel layer, optical channel layer) channel faults are detected, OTS is detected through optical power meters, and OCh channels are detected through bit error meters to ensure the healthy operation of the optical network. However, even if the project acceptance requirements are met, after being put into use, various operations and maintenance of the network, such as external force-triggered failures, loose connections, connection module failures, fiber core failures, etc., will cause the optical network to fail to operate normally.

In related technologies, optical network fault location is to test the optical network through optical power and bit error rate meters, analyze the optical signal, analyze the cause, and obtain whether the line is faulty. However, for complex network environments, this detection method is not easy to accurately locate faults and alarms when used in commercial projects, resulting in low detection efficiency and high engineering maintenance costs. Especially in the case of slow fiber failure, there will be a period of performance degradation for a certain period of time but it will not have a major impact on the business, so it is not easy to identify the degradation trend in the early stage. Usually, the business is interrupted and complaints are raised. As a result, the operation and maintenance engineers are in a passive processing state for a long time. When the problem is discovered, large losses may have already occurred.

Based on this, embodiments of the present application provide optical network health monitoring methods, management units, systems and storage media. The solution idea of the embodiment of the present application is to: obtain the optical network health performance prediction instruction; obtain the optical network historical performance data in the first time period according to the optical network health performance prediction instruction; according to the optical network historical performance data and the optical network prediction model , obtain optical network prediction performance data. Examples of this application By obtaining the historical performance data of the optical network, the performance of the optical network is monitored, and the optical network prediction model is used to predict the future performance data of the optical network based on the historical performance data of the optical network, thereby realizing the monitoring and control of the performance of the optical network. Prediction, which is conducive to efficient testing and fault location of each optical network service link, and early warning and prompts for possible future failures.

In order to solve the above-mentioned problems of the current optical network health, meet the efficient test fault and location of various services, and provide early warning and prompts for possible future fault problems.

The embodiment of this application uses AI (Artificial Intelligence, artificial intelligence) algorithm to analyze the historical performance data of the optical network. For example, for optical fiber, the attenuation of optical signal power is proportional to the signal attenuation of OTS services; for optical signals, the bit error rate is proportional to the degree of degradation of OCh services. By extracting performance values for a certain period of time, and then storing the performance values in a specific format, filtering short-term special extreme values, and performing sum operator scheduling calculations through AI algorithms, the future optical network performance indication value is predicted, and the system can automatically analyze The cause of the fault is automatically given and treatment suggestions are provided to quickly identify and repair the fault and detect the problem in its bud.

Referring to Figures 1 and 2, Figures 1 and 2 are schematic diagrams of a system architecture for performing an optical network health monitoring method provided by an embodiment of the present application. In the examples of Figures 1 and 2, the system architecture includes an optical network management unit, multiple service endpoint devices, and multiple intermediate node devices. As shown in Figure 1, the optical network management unit communicates with each business endpoint device and each intermediate node device to realize management and control of the OTN network and realize intelligent synchronization of data across the entire network; multiple business endpoint devices and multiple The intermediate node device is connected through optical fiber communication and is used to transmit optical network service data; for example, the optical network service data can be transmitted from the service endpoint device on the left to the service endpoint device on the right through service route 1; for another example, the optical network service data It can be transmitted from the service endpoint device on the left to the service endpoint device on the right through service route 2.

Among them, as shown in Figure 1, the optical network management unit can include network management equipment, SDN (Software Defined Network, Software Defined Network) controller, APP, database, etc. As shown in Figure 2, the network management device can be a new generation wireless network management UME (Unified Management Expert, UME) system device; the SDN controller can include an SC controller (network centralized controller) and multiple DC controllers (Domain Controller, domain Controller), each DC controller is used to manage and control the intermediate node equipment in the corresponding domain. The SDN controller communicates and connects with multiple DC controllers respectively to achieve centralized management and control; UME and SDN controller, multiple The DC controllers are respectively connected through communication; the APP is connected through communication with the SC controller for outputting interface display and providing user interaction.

In some embodiments, the optical network management unit includes but is not limited to: initialization module, task management module, task scheduling module, operator task scheduling module, operator calculation module, task calculation module, service PRBS sending module, and service PRBS receiving module. , business PRBS management module, optical network health analysis module, data storage module, interface display module and data synchronization module.

Among them, the initialization module, when the system is initialized, obtains the optical network performance value from the original performance database, including but not limited to performance values based on OTS/OCh services, such as optical power, bit error rate, PRBS test information of the service, test start time, End time, test results, etc., read various resource information, and display this information in the interface display module.

The task management module is used to create optical network health performance prediction tasks according to task specifications. Set the task name, pass the prediction interval and prediction duration, select the prediction resource file, store the task information, and build the task data space; after activating the task, parse the original performance asset data, save the original performance data of the task, and convert the original OTS/OCh business performance The data is converted into performance files and resource files required by the algorithm module and saved into the database.

The task scheduling module is used to trigger optical network tasks, retrieve and monitor task status, and push the task results to the task calculation module after the task calculation is completed.

The operator task scheduling module is used to realize the scheduling of various operators. Different operators are responsible for different calculation functions, realize the workflow of the operators, and ensure the collaborative calculation of various operators.

The operator calculation module, also called the AI algorithm module, is used to convert and analyze resource file formats, and to pass historical performance data through the underlying ARMA (auto regressive moving average model, autoregressive moving average model) + LSTM (Long Short -Term Memory, long short-term memory network) algorithm combination to generate future performance data, and after calculation, a prediction file is generated to the task calculation module. Because the optical power will change due to the influence of the external environment and is random and uncertain, the optical fiber optical power is a time series data with characteristics of nonlinearity, time variability and complexity. In view of these characteristics of optical fiber optical power, first use business PRBS to test the optical fiber passing by the business link, record the results of multiple test data, then use ARMA to predict small changes in optical power and bit error rate data, and use LSTM to predict optical power. The changing trend of power and bit error rate data, and other prediction data are superimposed through the algorithm model to obtain the optical network health prediction data.

The task calculation module analyzes and calculates the health of the optical network based on the results output by the task scheduling module and the operator calculation module, and stores the analyzed data into the database. The stored data is used by the optical network health analysis module.

The service PRBS sending module is used to send PRBS encoded signals to the source endpoint device A-->sink endpoint device Z direction and the sink endpoint device Z-->source endpoint device A direction for various optical network services.

The service PRBS receiving module is used to receive various optical network services in the direction of source endpoint device A-->sink endpoint device Z, sink endpoint device Z-->source endpoint device A direction to send PRBS encoded signals, and process the sent PRBS Compare the signals and analyze whether the PRBS signals are consistent.

The service PRBS management module is used to load various PRBS test services and display the PRBS service ID, whether to select, PRBS type, PRBS receiving status, service source endpoint device A, service sink endpoint device Z, test start time, test end time, Test result information wait. Set the PRBS type, test start time, end time, test results, etc. for data sent by the service PRBS sending module and the service PRBS receiving module.

The optical network health analysis module is used to analyze whether the optical network is normal based on the service PRBS test results, and then obtain the fiber health statistics of the entire network, the OCh health statistics of the entire network, and the health score based on the calculation results of the task calculation module, and Obtain the object optical network health monitoring curve, including historical curves and predicted curves, quickly locate the fault point, provide early warning for possible faulty links, provide which OCh and customer services are affected by sub-healthy optical fiber, and explain the cause of the fault .

The data storage module is used to store the data of the above modules in the database or local task space.

The interface display module is used to display the data of the optical network health analysis module in the BS/CS mode, including but not limited to displaying various data of the optical network health in cylinders, pie charts, curves, etc., and displaying it in special colors. There may be faulty links in the future. Display the test results of service PRBS, including but not limited to service label, service type, whether to select, service source endpoint device A, service sink endpoint device Z, PRBS type, PRBS reception status, test start time, test end time, and test results. All network elements and link connection status diagrams are displayed in the form of a GIS map. Links with problems are reminded in red, and the display can be zoomed in and out. The health history and prediction curve of the selected link can be viewed.

The data synchronization module is used to regularly update the performance data of the original database to the task management module, and update the optical network health database and local task space with the prediction task as the space to maintain data consistency. In some embodiments, data synchronization can also be triggered manually.

The service endpoint device can be a source endpoint device or a sink endpoint device. The service endpoint device can be used to provide an optical network service data access interface. In some embodiments, the service endpoint device may be a CPE (Customer Premises Equipment) device, used for accessing service data and sending or receiving optical network service flows.

The intermediate node device is used to forward optical network service flows. The intermediate node device can be connected to the service endpoint device through optical fiber, or can be connected to one or more other intermediate node devices through optical fiber. In some embodiments, the service endpoint device may be a NE (Net Element) device.

Figure 1 exemplarily illustrates the topological relationship between the network management equipment, SDN controller and node equipment (including multiple service endpoint equipment and multiple intermediate node equipment). The topology structure of the network management equipment, SDN controller and node equipment is comprehensively analyzed. Network data is intelligently synchronized. The node device reports topology resources to the network management device and SDN controller. After the network management device and SDN controller monitor the data, they update the topology data of the network management device and SDN controller and maintain the database data of the network management device and SDN controller. Node device data is synchronized and consistent. The agent module on the node device side is set up on each node device. The agent module is responsible for communicating with the single board in the node device in the south direction, and is responsible for communicating with the network management device and SDN controller in the north direction, and managing data resources.

For example, the optical network shown in Figure 2 includes 4 service endpoint devices CPE1 to CPE6 and 15 intermediate node devices NE1 to NE15. Among them, NE1 to NE5 form the DC A domain, and NE6 to NE10 form the DC C domain. , NE11~NE15 form DC B domain, CPE1 and CPE3 are connected to DC A domain, CPE2 and CPE4 are connected to DC C domain, and CPE5 and CPE6 are connected to DC B domain. The SDN controller can include an SC controller and three DC controllers. The three DC controllers are DC controller A used to manage and control DC A domain, DC controller B used to manage and control DC B domain. The DC controller C used to manage and control the DC C domain. The SC controller is connected to the three DC controllers respectively. The UME is connected to the SDN controller and the three DC controllers. The APP is connected to the SC controller. , used to output interface display and provide user interaction. In the optical network shown in Figure 2, bidirectional service 1 starts from CPE1 as the source endpoint device, passes through the intermediate nodes NE1, NE2, NE6, and NE9 in sequence, reaches CPE4 as the sink endpoint device, and then passes through CPE4 as the sink endpoint device. Starting from the intermediate nodes NE9, NE6, NE2, and NE1, it reaches CPE1 as the source endpoint device to realize bidirectional optical network communication; the unidirectional service 1 starts from CPE3 as the source endpoint device and passes through the intermediate nodes NE3, NE4, and NE11 in sequence. , NE15, and NE13, reach CPE6 as the sink endpoint device, realizing one-way optical network communication.

The system architecture and application scenarios described in the embodiments of this application are for the purpose of explaining the technical solutions of the embodiments of this application more clearly, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. Those skilled in the art will know that with the system architecture With the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

Those skilled in the art can understand that the above hardware platform does not limit the embodiments of the present application, and may include more or fewer components than shown in the figures, or combine certain components, or arrange different components.

In the above hardware platform, the optical network management unit can call its stored optical network health monitoring program to execute the optical network health monitoring method.

Based on the above system architecture, various embodiments of the optical network health monitoring method in this application are proposed.

Embodiments of the present application provide an optical network health monitoring method, which is applied to an optical network management unit. As shown in Figure 3, an optical network health monitoring method includes:

S100, obtain optical network health performance prediction instructions;

S200: Obtain the historical performance data of the optical network in the first time period according to the optical network health performance prediction instruction;

S300: Obtain optical network prediction performance data based on the optical network historical performance data and the optical network prediction model.

In some embodiments, the optical network health performance prediction instruction is triggered based on the status of the optical network health performance prediction task; the first time period is determined based on the time period selected by the user corresponding to the optical network health performance prediction task; the optical network prediction The model can be a neural network The network model may also be a combination of multiple neural network models, which is not limited in the embodiments of the present application.

The embodiment of this application proposes a calculation method based on AI operator scheduling to analyze performance data, and predict future performance data after calculation based on historical data.

For example, in some embodiments, the optical power historical value of the optical channel OTS service can be analyzed and the optical fiber health curve can be obtained through algorithm calculation, including the historical curve and the curve of the future time (the prediction time can be set), and the possible Provide timely warnings for faulty optical fiber links and ports, identify faulty optical fibers, sub-healthy optical fibers and early warning status in the entire network, view status details, and provide which OTS services are affected by sub-healthy optical fibers.

For another example, in some embodiments, the optical fiber health curve can be obtained by analyzing the historical value of the bit error rate of the optical channel OCh service and calculating it through an algorithm, including the historical curve and the curve of the future time (the prediction time can be set) , provide timely warnings for possible faulty OCh links and ports, obtain status details, check which customers are affected by sub-healthy OCh services, and give the reasons.

The first aspect of the embodiments of the present application provides an optical network health monitoring method, which includes: obtaining an optical network health performance prediction instruction; obtaining optical network historical performance data within a first time period according to the optical network health performance prediction instruction; Optical network historical performance data and optical network prediction model are used to obtain optical network prediction performance data. The embodiments of the present application realize the monitoring of the performance of the optical network by acquiring the historical performance data of the optical network, and use the optical network prediction model to predict the future performance data of the optical network based on the historical performance data of the optical network, thereby realizing the monitoring of the optical network performance. Performance monitoring and prediction will facilitate efficient testing and fault location of each optical network service link, and provide early warning and prompts for possible future failures.

In some embodiments, the optical network historical performance data includes at least one of the following: optical signal power, bit error rate, optical signal-to-noise ratio, PRBS test data, etc.

In some embodiments, the optical signal power can be obtained by measuring the OTS service on the optical network service link with an optical power meter, the bit error rate can be obtained by measuring the OCh service on the optical network service link by using a bit error meter, and the PRBS test data can be It is obtained by measuring the ODU service, OAC service or OCh service on the optical network service link by the PRBS test module.

In some implementations, S300, obtaining optical network predicted performance data based on optical network historical performance data and optical network prediction models includes:

S310: Format the optical network historical performance data to obtain formatted time series data;

S320: Input the formatted time series data into the optical network prediction model to obtain optical network prediction performance data.

In some embodiments, the historical performance data of the optical network is unstructured data and needs to be first converted into a data format required by the optical network prediction model.

In some implementations, S310, formatting the optical network historical performance data to obtain formatted time series data includes:

S311, extract the original performance files and original resource files from the historical performance data of the optical network;

S312, associate the performance original file with the performance structure file to form a performance data structure file;

S313, associate the resource original file with the resource structure file to form a resource data structure file;

S314. Merge the performance data structure file and the resource data structure file to obtain the formatted time series data.

In some embodiments, the original performance file mainly records time information, ports and their corresponding performance data. The performance data can be optical signal power, bit error rate, optical signal-to-noise ratio, PRBS test data, etc.; the original resource file mainly records services. Channel information, such as source endpoint device port information, sink endpoint device port information, intermediate node device port information, channel status, code rate, MOC (Means of Communication, communication means), etc.

In some embodiments, by constructing performance and resource model data, each operator information must first be constructed, including operator name, associated parameters, execution order, etc. The operator name is the displayed operator name. The performance structure file and the resource structure file can be set in advance, and the operator calculation module can associate the performance original file with the performance structure file through the preset association parameters to form a performance data structure file; combine the resource original file with the resource structure file Association to form a resource data structure file. Among them, the association parameter is a parameter used to associate the original performance file with the performance structure file, and associate the original resource file with the resource structure file. For example, the association parameter can be time information. The original file has no data structure. Through the time information, the original performance file and the performance structure file are associated. This becomes performance data with a data table structure. The original resource file and the resource structure file are associated. , thus becoming resource data with a data table structure; facilitating subsequent data conversion. Then, the performance data structure file and the resource data structure file are merged to obtain the formatted time series data.

For example, the data content of the performance original file is shown in Figure 4, in which the first recording time information is "2021-12-02 16:00:00"; the port is "ME{472306ac-e84d-40a0-8bb6" -abdd748365b0], EQ＝{/r＝0/sh＝1/31＝32], PTP＝{/p＝453_4)"; the performance data is "2.2600000000000002, -2.3000000000000003, -60, -60, -60, - 2.27"; The records in Articles 2 to 6 are similar and will not be repeated here.

The content of the performance structure file is shown in Figure 5. The code specifies the name tags and structure of each field. The data of the performance original file can be structured through the code shown in Figure 5.

Exemplarily, the associated resource data structure file is shown in Figure 6. The resource file conversion process may include: converting the associated resource file into Convert into resource files including but not limited to .csv format. The generated resource file related information is shown in Figure 6. The information recorded in the resource file includes: source endpoint device port information APORTID, sink endpoint device port information ZPORTID, and intermediate node device port information ANEID. and ZNEID, channel status STATE, code rate RATE, MOC (Means of Communication, communication means), etc.

Convert the associated performance files into performance files including but not limited to .csv format, and merge the converted performance files to generate a new file diagram. This file can be used directly by the algorithm, as shown in Figure 7.

In some embodiments, the optical network prediction model includes an LSTM-ARIMA model;

S320, inputting the formatted time series data into the optical network prediction model to obtain optical network prediction performance data, including:

S321, perform wavelet decomposition on the formatted time series data to obtain high-frequency component data and low-frequency component data;

S322, input the high-frequency component data into the ARIMA model for prediction, and obtain high-frequency prediction results;

S323, input the low-frequency component data into the LSTM model for prediction, and obtain the low-frequency prediction result;

S324: Superpose the high-frequency prediction result and the low-frequency prediction result to obtain the optical network prediction performance data.

In some embodiments, the LSTM-ARIMA algorithm calculation is described with reference to Figure 8, taking the optical network prediction performance data as optical power as an example. First, the processed link + performance data is formed into optical power time series data (formatted time series data); then the formatted time series data is subjected to wavelet decomposition to obtain high-frequency component data and low-frequency component data. Among them, the high-frequency component data represents the details of the light attenuation data and represents the random changes in optical power attenuation; the low-frequency component represents the trend term of light attenuation and represents the trend of light attenuation changes. The high-frequency component data is input into the ARIMA model for prediction, and high-frequency prediction results are obtained. Among them, the main steps of ARIMA model prediction include stationarity analysis (ADF test), model ordering (ACF and PACF), construction of time series data, and model prediction. The low-frequency component data is input into the LSTM model for prediction, and low-frequency prediction results are obtained. Among them, the main steps of LSTM model prediction include: data normalization, construction of prediction data set, training model, prediction data, and data denormalization. The high-frequency prediction results and the low-frequency prediction results are superimposed to obtain the optical network prediction performance data. For example, perform db5 wavelet reconstruction on the two sets of predicted data (high-frequency prediction results and low-frequency prediction results) to obtain the optical network prediction performance data.

In some embodiments, after obtaining the optical network prediction performance data, the optical network prediction performance data is output, the optical network prediction performance data is used by the optical network health analysis module, and is finally presented in the interface display module.

In some embodiments, the optical network prediction model includes a trend item prediction sub-model (such as an LSTM sub-model) and a detail item prediction sub-model (such as an ARIMA sub-model). Before using the optical network prediction model, you need to build and train the optical network prediction model.

Exemplarily, constructing and training an optical network prediction model includes the following steps S320-1 to S320-4:

S320-1: Construct a trend item prediction sub-model.

According to the optical power, bit error rate, OSNR, and PRBS test results, the historical performance data trend item {Lt, t=1, 2, 3,..., N} is divided into ranges. Lt represents the performance data at time t, and the data extremes are removed. value, and use the LSTM algorithm to predict the trend item data prediction results.

S320-2: Construct a detail item prediction sub-model. The detail item prediction sub-model may be an ARIMA-based sub-model.

S320-3: Perform model testing and prediction based on the sample data, and combine the ARMA model and the LSTM model to superimpose the predicted trend data and model data.

S320-3, perform variance calculation.

Calculate the error based on the variance, filter out a port (BN:ME{106635791}, EQ={/r=0/sh=1/sl=6}, PTP={/p=401_1}), and then perform feature engineering. From the output optical power at a specific time interval (such as 15 minutes), filter out the effective optical power information, totaling N (such as 589) pieces of data, and perform predictive analysis. The first X% (such as 95%) of the data is used as training data, and the last Y% (such as 5%) of the data is used as test data. The abscissa is the index of the data in the array, and the equivalent mapping is a specific time interval (such as 15min) time span, that is, when the abscissa span is 1, it means that the time span is also a specific time interval (such as 15min). The line chart of the training data is shown in Figure 9.

Then db5 wavelet decomposition is performed on the training data, which is decomposed into a detailed part and an approximate part. The detailed part is predicted by the ARIMA algorithm, and the approximate part is predicted by the LSTM algorithm. The line graph of the approximate part of the prediction results is shown in Figure 10, and the line graph of the detailed part is shown in Figure 11; finally, wavelet fusion is performed, and the final line graph is shown in Figure 12.

It should be noted that in Figures 8 to 16, taking optical power as an example, the horizontal axis represents time and the vertical axis represents optical power value.

As shown in Figure 13, in some embodiments, the method further includes:

S400, obtain prediction task parameters, wherein the task parameters include parameters corresponding to the first time period;

S500: Create an optical network health performance prediction task according to the prediction task parameters;

S600: Generate the optical network health performance prediction instruction according to the status of the optical network health performance prediction task.

In some embodiments, the embodiments of this application propose a task management-based method to create a prediction task. After starting the prediction, the system can automatically perform prediction on the entire optical network, start, suspend, and stop the prediction task, and can independently perform certain tasks. An optical fiber, OCh channel, and PRBS perform task management operations to achieve prediction of optical network performance within a specified prediction time range in the future, solving the problems of multi-concurrent asynchronous execution, monitoring, retrieval, and calculation of optical fiber performance prediction.

In some embodiments, the prediction task parameters further include at least one of the following:

Task name, prediction type, task type, prediction interval, prediction duration, resource files, task information, task space data parameters, etc.;

Wherein, the resource file is used to determine the resource file of the first time period, and the task data space data parameter is used to create a task data space for storing task-related data.

In some embodiments, as shown in Figure 14, it is a schematic diagram of the optical network health creation prediction task in the embodiment of the present application. The optical network health creation prediction task is used to create a new optical network health prediction task. Each task is as shown in the figure. Enter the requirements in and store it in the database after successful creation and display it in the list. During the creation process, the network management performance data must be taken out, processed into optical network performance data, and then entered into the optical network health database. The task can be started again after the task is started. Perform operations such as path calculation.

In some embodiments, after creating an optical network health performance prediction task based on the prediction task parameters, the optical network health detection method further includes:

S510, obtain original performance data from the database according to the prediction task parameters;

S520: Convert the original performance data into optical network historical performance data with original performance files and original resource files.

In some embodiments, as shown in Figure 15, a schematic diagram of a prediction task is created for optical network health analysis in this embodiment of the present application. Includes the following steps:

E111, create a prediction task, and implement the operation of the optical network health prediction process in the form of tasks. There can be multiple tasks. The input of each task is shown in Figure 10. Multiple tasks can be run in parallel. The idea of implementing the task is to combine the network Management performance data is extracted, copied to the space of each task, and then converted into the data format required by the AI algorithm. After the algorithm is calculated, the performance data required for optical network health analysis is generated and stored in the database, so that the interface display module can be based on performance Various optical network link health information required by the data generation interface, including network-wide health status, network-wide fiber health statistics, network-wide OCh health statistics, health score, object performance prediction, and warning fiber/OCH information .

E112 stores task information, extracts network management performance data, such as optical power, bit error rate, OSNR, and service PRBS data, and stores them in task space in units of tasks.

E113, parses the original performance asset data of the network, and parses the network management performance data into the data file format required by the algorithm.

E114, save the parsed task data and save the parsed task data in the task space.

E115, generate algorithm required files from the performance data in the task space.

E116, obtain the AI creation project template. The AI algorithm module provides a module for each task, which is generally defined in json file format. Each task changes this file module to its own file data, defines scheduling operator related information, and stores original data within the task. Path, performance data collection frequency and other information, the AI algorithm module uses these to generate the files required by each operator from the historical performance data of the task space.

E117, create a path calculation task, AI creates a path calculation task, starts the thread, and schedules the calculation of each operator.

E118, generate AI path calculation task information, and each operator generates the task information required for AI path calculation.

E119, update task status and store task data information.

E1110, the prediction task was created successfully.

In some embodiments, the status of the optical network health performance prediction task includes startup status, pending status, and execution completed status;

S600, generating the optical network health performance prediction instruction according to the status of the optical network health performance prediction task, including:

S610: When the optical network health performance prediction task is in the startup state, generate the optical network health performance prediction instruction.

In some embodiments, the embodiments of this application propose a task scheduling method to realize the triggering of optical network tasks, activate tasks, and regularly retrieve and monitor task status. After the task calculation is completed, it is monitored that the task status changes from waiting to waiting. When the task is completed, the task results are pushed to the task calculation module, which solves the problem of asynchronous scheduling of multiple tasks.

Exemplarily, as shown in Figure 16, it is a schematic diagram of the optical network health analysis startup prediction task in the embodiment of the present application, which includes the following steps:

E211, start the prediction task, start the prediction task in the network management system operation, and create task-related threads.

E212, modify the task status to "Executing".

E213, start the AI prediction task, start the AI prediction task execution, run AIE operator mobilization, and execute the operators separately in a row.

E214, modify the task status. After the task is executed, modify the task status to "Ended" or "Failed".

E215, executes the prediction task, executes each operator separately, and finally outputs the performance prediction data file.

E216, output prediction file, the operator is executed successfully, and the performance prediction data file is output.

E217, returns the test results, and sends the test results to the task management module for processing and storage.

In addition, in this embodiment of the present application, the optical network management unit also provides task query and task backup functions.

Illustratively, as shown in Figure 17, it is a schematic diagram of the timing query task results in the embodiment of the present application, including the following steps:

E311, start querying task status results regularly.

E312, traverse the current task;

E313, get the list of tasks being executed;

E314, query task information regularly;

E315, get task status.

E316, determine whether the task is completed, if completed, go to step 7, otherwise go to step 2;

E317, AI prediction task output file, output performance prediction data file.

E318, parse performance prediction data files.

E319, update task information and save parsed performance data.

For example, as shown in Figure 18, the schematic diagram of performance data backup in the embodiment of this application includes the following steps:

E411, query expired data.

E412, get data.

E413, packaged into files.

E414, performance data backup.

E415, clean data.

As shown in Figure 19, in some implementations, the optical network health detection method also includes:

S700, obtain PRBS test instructions;

S800: Perform a PRBS test on the optical network service link according to the PRBS test instruction, and obtain the PRBS test results of each service link.

In some embodiments, the embodiments of this application propose a service-based PRBS test, which performs PRBS tests on two-way optical network services in both directions. By returning the PRBS results, the health status of the optical network of the optical channel where the service is located is analyzed and displayed. Test Results.

In practical applications, the PRBS (Pseudo-Random Binary Sequence) test of optical network services can be used to simulate as much as possible some of the system acceptance indicators that need to be tested by hanging tables during actual OTN project acceptance (such as 5 minutes, 15 minutes, 24 hour system error test), in order to achieve the purpose of acceptance without having to carry professional instruments to the station. Having this function is the prerequisite for guiding customers to agree to remote acceptance. If deployed in large quantities on the existing network, it can greatly save the cost of project acceptance testing and help shorten the project delivery cycle.

In some embodiments, S800, perform a PRBS test on the optical network service link and obtain the PRBS test results of each service link, including:

S810, obtain the optical network service link to be tested;

S820: Determine the source endpoint device and the sink endpoint device according to the optical network service link;

S830, send the PRBS test command to the source endpoint device;

S840: Receive the PRBS test results from the sink endpoint device.

In some embodiments, as shown in Figure 2, which is a schematic diagram of the service PRBS test system in the embodiment of the present application, the service PRBS test can be a two-way service, such as two-way service 1, or a one-way service, such as one-way service 2. It can be a cross-domain service or a single pre-service; the PRBS code stream signal is sent from the source board of the service, passes through all network elements and boards where the service is located, and the PRBS code stream signal is viewed on the sink board of the service, and the network management receives Compare the PRBS code stream signals reported to the sink board to see if the PRBS code streams of the source and sink boards are consistent. If they are consistent, the optical fiber link through which the current service passes is considered healthy; if they are inconsistent, the optical fiber link passed by the current service is considered healthy. There is deterioration of the optical fiber link passing through, or there is a loss of business; the tested link information is stored in the database as part of the data basis for link analysis of the link optical network link situation.

For example, as shown in Figure 20, it is a schematic diagram of service-based PRBS test management in the embodiment of the present application. PRBS related data of various user-selected service types can be loaded into the network management system, PRBS tests are issued for the selected two-way or unidirectional services, and PRBS tests are issued for the source boards of one or more services, and the delivery is stopped. After the command is issued, check the PRBS test results and store the service test results in the database. The optical network health analysis module analyzes and checks the problematic optical links based on the stored information and gives early warnings. As shown in Figure 20, the user label can display bidirectional services separately according to forward/reverse. The service types can include OAC services, OCH services, ODU services, etc. The PRBS test data can be encoded according to the selected PRBS type.

In some implementations, the optical network service link includes at least one of the following:

ODU service link, OCA service link, OCH service link, etc.

In some embodiments, the PRBS test on the optical network service link according to the PRBS test instruction includes:

S850, obtain PRBS test parameters according to the PRBS test instruction;

S860: Perform PRBS testing on the optical network service link according to the PRBS test parameters;

Wherein, the PRBS test parameters include at least one of the following:

PRBS service ID, PRBS type, PRBS reception status, service source endpoint device, service sink endpoint device, test start time, test end time, etc.

As shown in Figure 21, in some implementations, the optical network health detection method also includes:

S900: Obtain multiple optical network prediction performance data of the optical network service link;

S1000: Predict the changing trend of the optical network service link based on a plurality of the optical network performance data.

As shown in Figure 22, in some implementations, the optical network health detection method also includes:

S1100, obtain PRBS test results of multiple optical network service links;

S1200, obtains optical network historical performance data and optical network predicted performance data of multiple optical network service links;

S1300: Obtain the optical network health level of the optical network service link based on the optical network historical performance data and optical network predicted performance data of the optical network service link;

S1400: According to the PRBS test results of each optical network service link, the optical network health level and the optical network service link Quantity, the optical network health score is calculated.

In some embodiments, embodiments of the present application propose to incorporate service-based PRBS test results, fiber health status performance prediction results, and OCh health status performance prediction results into optical network health monitoring and prediction evaluation to obtain an optical network health score.

Exemplarily, refer to Figure 23, which is a schematic diagram of the service PRBS test in the embodiment of the present application, including the following steps:

E511, delivers PRBS test, and the service PRBS management module issues PRBS test command to the selected service;

E512, traverse all issued services and take out the current service;

E513, take out the current service source board and construct the PRBS command information, such as PRBS service type, PRBS direction, receiving status, PRBS name, authorization or not, and other information.

E514, sends the PRBS code stream signal to the source board of the current service;

E515, reads the PRBS signal through the service sink board;

E516, determine whether the service source and sink PRBS signals are consistent;

E517, enter the success or failure results of the current service PRBS test into the optical network health detection database;

E518, the optical network health analysis module analyzes and predicts faulty links, and gets the prediction data into the database;

E519, the optical network health analysis module generates optical network link health information, including network-wide health status, network-wide fiber health statistics, network-wide OCh health statistics, health score, object performance prediction, and warning fiber/OCH information. .

In some embodiments, the optical network health level includes healthy, sub-healthy and abnormal.

In some embodiments, S1400, calculating the optical network health score based on the PRBS test results of each optical network service link, the optical network health level and the number of optical network service links includes:

S1410, configure a first weight for the PRBS test results, configure a second weight for the optical network health level, perform weighted calculations on the PRBS test results and optical network health levels of each optical network service link, and obtain calculation results;

S1420: Divide the calculation result by the number of optical network service links to obtain the optical network health score.

As shown in Figure 24, in some implementations, the optical network health detection method also includes:

S1500 displays the optical network health status in BS/CS mode according to at least one of the following:

PRBS test results of multiple optical network service links, optical network historical performance data, optical network predicted performance data, optical network health level, and optical network health score;;

Wherein, the display method of the optical network health status includes at least one of the following: GIS map, cylinder statistical chart, pie chart, and curve statistical chart.

In some embodiments, the embodiments of this application propose to display all network element and link information through a GIS map, and support zooming in on the map, so that the performance health curve of any link can be viewed, and the link can be obtained Provide early warning of possible problems on the road and provide repair suggestions.

In some implementations, S1500, the optical network health status is displayed in a BS/CS manner according to at least one of the following: PRBS test results of multiple optical network service links, optical network historical performance data, optical network Predicted performance data, optical network health level, and optical network health score, including:

S1510. Display the health status of each optical network service link in a GIS map according to the optical network health level of each optical network service link;

and / or,

S1520: Based on the optical network health level of each optical network service link, display the optical fiber health statistics chart of the entire network and the OCH health statistics chart of the entire network;

and / or,

S1530, display the PRBS test results of each optical network service link;

and / or,

S1540. Display the optical network health monitoring curve according to the optical network historical performance data and/or optical network predicted performance data of multiple optical network service links, where the optical network health monitoring curve includes historical curves and/or predictions. curve.

Exemplarily, as shown in Figure 25, it is a schematic diagram of the optical network health interface in the embodiment of the present application, which respectively displays the health status of the entire network, fiber health statistics of the entire network, OCh health statistics of the entire network, and health scores. Object performance prediction, fiber optic/OCH warning is in progress. The health status of the entire network is displayed as a GIS map. Grouped nodes and connected fibers are dyed according to the evaluation results. The map can be reduced and enlarged. Links that may have problems are displayed in red. When the cursor moves over a link with problems, a reminder pops up. form. The fiber health statistics of the entire network show the number of normal, sub-healthy, and abnormal fiber links. The network-wide OCh health statistics show the number of normal, sub-healthy, and abnormal OCh links. The health score calculates three types of data: business PRBS test data, fiber prediction data, and OCh prediction data using a certain algorithm to obtain an optical network health score.

For example, the object performance prediction can be scheduled with a certain algorithm based on historical performance files such as optical power, bit error rate, OSNR, etc., and then combined with algorithms, such as the LSTM algorithm, to obtain the optical power, bit error rate, Performance values such as OSNR are plotted as curves, so that the trend of fiber attenuation degradation and wavelength performance degradation can be seen more intuitively.

In some embodiments, the embodiments of the present application can obtain a sub-healthy optical fiber list and an OCH channel list based on the previous analysis, point out the failed optical fiber link, OCh link, and port location, and prompt an alarm or early warning.

For example, you can directly click to view the optical fiber/OCH being warned, as shown in Figure 26, which is a schematic diagram of healthy optical fiber link warning in the optical network in the embodiment of the present application. The optical network health creation prediction task displays sub-healthy and abnormal optical fiber link and OCh link information in a list, and gives processing suggestions.

As shown in Figure 27, in some implementations, the optical network health detection method also includes:

S1600, obtain the original performance data of the optical network from each network element of the optical network;

S1700: Store the original performance data in the database.

As shown in Figure 28, in some embodiments, the predicted task parameters include task space data parameters, and the method further includes:

S1800: Extract original performance data related to the optical network health performance prediction task from the database to synchronize the original performance data in the task space data.

To sum up, in the software-defined network, the embodiment of this application realizes optical network health monitoring and assurance from business PRBS test results, span-segment optical fiber and OCh historical data, and establishes a set of prediction management tasks by collecting these historical performance data. The mechanism imports data into AI algorithm analysis and big data analysis to achieve early and timely prediction of optical network performance sub-health and fault information, and obtain the fiber health statistics of the entire network, the OCh health statistics of the entire network, and the health score, and obtain The object optical network health monitoring curves include historical curves and predicted curves, which can quickly locate fault points, provide early warning for possible faulty links, give out which OCh and customer services are affected by sub-healthy optical fiber, and explain the cause of the fault. Provides a basis for engineers to automatically identify fiber attenuation degradation, wavelength optical performance degradation and resolve these degradations.

The following uses an application example to further introduce the embodiment of the present application.

Example 1

The optical network health monitoring methods in this example include:

Step 1: Initialize the module. When the system is initialized, the optical network performance value is obtained from the database, including but not limited to performance values based on OTS/OCh services, such as optical power, bit error rate, PRBS test information of the service, test start time, and end Time, test results, etc., read various resource information, and display this information on the interface display module.

Step 2: The interface display module displays service PRBS information, including but not limited to service label, service type, whether to select, service A endpoint, service Z endpoint, PRBS type, PRBS reception status, test start time, test end time, and test results. Select one or more services, select the PRBS type, and issue the service PRBS test. All network elements and link connection status diagrams are displayed in the form of a GIS map. Links with problems are reminded in red, and the display can be zoomed in and out. The health history and prediction curve of the selected link can be viewed.

Step 3: The service PRBS sending module delivers the code stream, starts the PRBS test on the board card of the network element where the first node of the service is located, encodes the PRBS according to the selected PRBS type, and sends the PRBS code stream to the receiving module, which is the end of the service The board on which the node is located.

Step 4: The service PRBS receiving module receives the code stream. After the board of the single board where the service tail node is located receives the PRBS information code stream, it can compare the received PRBS code with the theoretically calculated PRBS code that should be received. Determine whether the equipment or transmission line is normal.

Step 5: The service PRBS management module analyzes the received information and obtains the optical network health status data of all services. It can be multiple types of services, such as ODU, OAC, OCH, etc. These data are pushed to the optical network health analysis module for use. .

Step 6: The task management module creates the optical network health performance prediction task. Set the task name, pass the prediction interval and prediction duration, select the prediction resource file, store the task information, and build the task data space; after activating the task, parse the original performance asset data, save the original performance data of the task, and convert the original OTS/OCh business performance The data is converted into performance files and resource files required by the algorithm module and saved into the database.

Step 7: The task scheduling module triggers the optical network task, activates the task, and regularly retrieves and monitors the task status. After the task calculation is completed, it monitors that the task status changes from waiting to completed, and pushes the task result to the task calculation. module.

Step 8: The task calculation module analyzes and calculates the health of the optical network based on the results output by the task scheduling module and the operator calculation module, and stores the analyzed data into the database.

Step 9: The operator calculation module, also called the AI algorithm module, implements conversion and analysis of resource file formats, and generates future performance data by combining historical performance data with the underlying ARMA+LSTM algorithm. After calculation, a prediction file is generated for the task. Compute module. Because the optical power will change due to the influence of the external environment and is random and uncertain, the optical fiber optical power is a time series data with characteristics of nonlinearity, time variability and complexity. In view of these characteristics of optical fiber optical power, first use business PRBS to test the optical fiber passing by the business link, record the results of multiple test data, then use ARMA to predict small changes in optical power and bit error rate data, and use LSTM to predict optical power. For the changing trend of power and bit error rate data, the optical network health prediction data can be obtained by superimposing other prediction data through the algorithm model. For algorithm implementation, please refer to Embodiment 13.

Step 10: The optical network health analysis module analyzes whether the optical network is normal based on the service PRBS test results, and then obtains the fiber health statistics of the entire network, the OCh health statistics of the entire network, and the health score ( The health score calculates performance data including but not limited to business PRBS test data, optical fiber prediction data, OCh prediction data, OSNR and other performance data with certain weights to obtain the optical network health score. Different weights can be selected according to different scenarios. ), and obtain the curve of the object's optical network health monitoring, including the historical curve and the predicted curve (the abscissa uses time span, the ordinate uses optical power or bit error rate value, form a curve lattice, and then connect it into a curve line), quickly locate the fault point, provide early warning for possible faulty links, provide OCh and customer services affected by sub-healthy optical fiber, explain the cause of the fault, and give repair suggestions.

Step 11: The data synchronization module regularly updates the performance data of the original database to the task management module, and updates it to the optical network health database and locally to maintain data consistency. Synchronization can also be triggered manually.

Step 12: The interface display module displays the data of the optical network health analysis module in BS/CS mode, including but not limited to cylinders, pie charts, curves, etc. to display various data on optical network health, and displays it in special colors. There may be faulty links in the future. Display the test results of service PRBS, including but not limited to service label, service type, whether to select, service A endpoint, service Z endpoint, PRBS type, PRBS reception status, test start time, test end time, and test results. All network elements and link connection status diagrams are displayed in the form of a GIS map. Links with problems are reminded in red, and the display can be zoomed in and out. The health history and prediction curve of the selected link can be viewed.

Embodiments of the present application provide an optical network health monitoring method, which includes: obtaining an optical network health performance prediction instruction; obtaining optical network historical performance data in a first time period according to the optical network health performance prediction instruction; and obtaining optical network historical performance data according to the optical network history. Performance data and optical network prediction model to obtain optical network prediction performance data. The embodiments of the present application realize the monitoring of the performance of the optical network by acquiring the historical performance data of the optical network, and use the optical network prediction model to predict the future performance data of the optical network based on the historical performance data of the optical network, thereby realizing the monitoring of the optical network performance. Performance monitoring and prediction will facilitate efficient testing and fault location of each optical network service link, and provide early warning and prompts for possible future failures.

In addition, an optical network management unit includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements claims 1 to 21 The optical network health monitoring method described in any one of the above.

As a non-transitory computer-readable storage medium, memory can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory may include high-speed random access memory and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may include memory located remotely from the processor, and the remote memory may be connected to the processor through a network. Examples of the above-mentioned networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks and combinations thereof.

It should be noted that the optical network management unit in this embodiment can be applied as the optical network management unit in the system architecture of the embodiment as shown in Figure 1 or Figure 2; in addition, the optical network management unit in this embodiment, The optical network health monitoring method in the embodiment shown in Figure 3 can be executed. That is, the optical network management unit in this embodiment, the optical network management unit in the system architecture of the embodiment shown in Figure 1 or Figure 2, and the optical network health monitoring method in the embodiment shown in Figure 3 all belong to The concept is the same, so these embodiments have the same implementation principles and technical effects, which will not be described in detail here.

The non-transient software programs and instructions required to implement the optical network health monitoring method in the above embodiment are stored in the memory. When executed by the processor, the optical network health monitoring method in the above embodiment is executed.

In addition, embodiments of the present application also provide an optical network system, including:

Optical network management unit as described previously;

A source endpoint device, connected to the optical network management unit, for sending or receiving optical network service flows;

A plurality of intermediate node devices connected to the optical network management unit for forwarding optical network business flows;

A sink endpoint device is connected to the optical network management unit and is used to receive or send optical network service flows.

In some embodiments, the optical network system is an optical network system having the system architecture of the embodiment shown in Figure 1 or Figure 2. For related descriptions, please refer to the foregoing description and will not be described again here.

In addition, embodiments of the present application also provide a computer-readable storage medium storing computer-executable instructions, and the computer-executable instructions are used for:

Implement the aforementioned optical network health monitoring method.

In some embodiments, the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by a processor or controller, for example, by a processor in the above-mentioned electronic device embodiment, which can cause The above processor executes the optical network health monitoring method in the above embodiment.

The first aspect of the embodiments of the present application provides an optical network health monitoring method, which includes: obtaining an optical network health performance prediction instruction; obtaining optical network historical performance data within a first time period according to the optical network health performance prediction instruction; Optical network historical performance data and optical network prediction model are used to obtain optical network prediction performance data. The embodiments of the present application realize the monitoring of the performance of the optical network by acquiring the historical performance data of the optical network, and use the optical network prediction model to predict the future performance data of the optical network based on the historical performance data of the optical network, thereby realizing the monitoring of the optical network performance. Performance monitoring and prediction will facilitate efficient testing and fault location of each optical network service link, and provide early warning and prompts for possible future failures.

Those of ordinary skill in the art can understand that all or some steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit . Such software may be distributed on computer-readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As is well known to those of ordinary skill in the art, the term computer storage medium includes all media used to store information, such as volatile and non-volatile, removable and non-removable media implemented in any method or technology (computer readable instructions, data structures, program modules or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disk (DVD) or other optical disk storage, magnetic cassettes, tapes, disk storage or other magnetic storage devices, or may Any other medium used to store the desired information and that can be accessed by a computer. Additionally, it is known to those of ordinary skill in the art that communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media .

The above describes some embodiments of the embodiments of the present application, but the embodiments of the present application are not limited to the above-mentioned embodiments. Those skilled in the art can also make various equivalents without violating the spirit of the embodiments of the present application. Modifications or substitutions, these equivalent modifications or substitutions are included in the scope defined by the claims of the embodiments of this application.

Claims (24)

An optical network health monitoring method, including: Obtain optical network health performance prediction instructions; According to the optical network health performance prediction instruction, obtain historical performance data of the optical network within the first time period; According to the optical network historical performance data and the optical network prediction model, the optical network prediction performance data is obtained. An optical network health monitoring method according to claim 1, wherein the optical network prediction performance data is obtained based on the optical network historical performance data and the optical network prediction model, including: Perform formatting processing on the optical network historical performance data to obtain formatted time series data; The formatted time series data is input into the optical network prediction model to obtain optical network prediction performance data. An optical network health monitoring method according to claim 2, wherein said formatting of said optical network historical performance data to obtain formatted time series data includes: Extract performance original files and resource original files from the optical network historical performance data; Associating the performance original file with the performance structure file to form a performance data structure file; Associating the resource original file with the resource structure file to form a resource data structure file; Merge the performance data structure file and the resource data structure file to obtain the formatted time series data. An optical network health monitoring method according to claim 2 or 3, wherein the optical network prediction model includes an LSTM-ARIMA model; The input of the formatted time series data into the optical network prediction model to obtain the optical network prediction performance data includes: Perform wavelet decomposition on the formatted time series data to obtain high-frequency component data and low-frequency component data; Input the high-frequency component data into the ARIMA model for prediction, and obtain high-frequency prediction results; Input the low-frequency component data into the LSTM model for prediction, and obtain low-frequency prediction results; The high-frequency prediction results and the low-frequency prediction results are superimposed to obtain the optical network prediction performance data. An optical network health monitoring method according to claim 1, wherein the method further includes: Obtain prediction task parameters, wherein the task parameters include parameters corresponding to the first time period; Create an optical network health performance prediction task according to the prediction task parameters; The optical network health performance prediction instruction is generated according to the status of the optical network health performance prediction task. An optical network health monitoring method according to claim 5, wherein the prediction task parameters further include at least one of the following: Task name, prediction type, task type, prediction interval, prediction duration, resource file, task information, task space data parameters; Wherein, the resource file is used to determine the resource file of the first time period, and the task data space data parameter is used to create a task data space for storing task-related data. An optical network health monitoring method according to claim 5, wherein after creating an optical network health performance prediction task based on the prediction task parameters, the method further includes: Obtain original performance data from the database according to the prediction task parameters; Convert the original performance data into optical network historical performance data with original performance files and original resource files. An optical network health monitoring method according to claim 5, wherein the state of the optical network health performance prediction task includes a starting state, a suspended state, and an execution completed state; Generating the optical network health performance prediction instruction according to the status of the optical network health performance prediction task includes: When the optical network health performance prediction task is in a starting state, the optical network health performance prediction instruction is generated. An optical network health monitoring method according to claim 1, further comprising: Get PRBS test instructions; According to the PRBS test instruction, PRBS test is performed on the optical network service link to obtain the PRBS test result of each service link. An optical network health monitoring method according to claim 9, wherein the PRBS test is performed on the optical network service links to obtain the PRBS test results of each service link, including: Obtain the optical network service link to be tested; Determine the source endpoint device and the sink endpoint device according to the optical network service link; Send PRBS test instructions to the source endpoint device; Receive PRBS test results from the sink endpoint device. An optical network health monitoring method according to claim 10, wherein the optical network service link includes at least one of the following: ODU service link, OCA service link, OCH service link. An optical network health monitoring method according to claim 10, wherein the PRBS test on the optical network service link according to the PRBS test instruction includes: According to the PRBS test instructions, obtain PRBS test parameters; Conduct PRBS testing on the optical network service link according to the PRBS test parameters; Wherein, the PRBS test parameters include at least one of the following: PRBS service ID, PRBS type, PRBS receiving status, service source endpoint device, service sink endpoint device, test start time, test end time. An optical network health monitoring method according to claim 1, further comprising: Obtain multiple optical network predicted performance data of the optical network service link; According to multiple optical network predicted performance data, the changing trend of the optical network service link is obtained. An optical network health monitoring method according to claim 1, further comprising: Obtain PRBS test results of multiple optical network service links; Obtain historical optical network performance data of multiple optical network service links; After obtaining the optical network prediction performance data based on the optical network historical performance data and the optical network prediction model, it also includes: Obtain the optical network health level of the optical network service link based on the optical network historical performance data and optical network predicted performance data of the optical network service link; An optical network health score is calculated based on the PRBS test results of each optical network service link, the optical network health level and the number of optical network service links. An optical network health monitoring method according to claim 14, wherein the optical network health level includes healthy, sub-healthy and abnormal. An optical network health monitoring method according to claim 14, wherein the optical network health monitoring method is calculated based on the PRBS test results of each optical network service link, the optical network health level and the number of optical network service links. Network health score, including: Configure a first weight for the PRBS test results, configure a second weight for the optical network health level, perform a weighted calculation on the PRBS test results and the optical network health level of each optical network service link, and obtain the calculation results; Divide the calculation result by the number of optical network service links to obtain the optical network health score. An optical network health monitoring method according to any one of claims 14 to 16, further comprising: Display the optical network health status in BS/CS mode according to at least one of the following: PRBS test results of multiple optical network service links, optical network historical performance data, optical network predicted performance data, optical network health levels, and optical network health scores; Wherein, the display method of the optical network health status includes at least one of the following: GIS map, cylinder statistical chart, pie chart, and curve statistical chart. An optical network health monitoring method according to claim 17, wherein the optical network health status is displayed in a BS/CS manner according to at least one of the following: PRBS test results of a plurality of optical network service links , optical network historical performance data, optical network predicted performance data, optical network health level, optical network health score, including: Display the health status of each optical network service link in a GIS map according to the optical network health level of each optical network service link; and / or, According to the optical network health level of each optical network service link, display the optical fiber health statistics chart of the entire network and the OCH health statistics chart of the entire network; and / or, Display the PRBS test results of each optical network service link; and / or, Optical network health monitoring curves are displayed according to optical network historical performance data and/or optical network predicted performance data of multiple optical network service links, where the optical network health monitoring curves include historical curves and/or predicted curves. An optical network health monitoring method according to claim 7, further comprising: Obtain the original performance data of the optical network from each network element of the optical network; The raw performance data is stored in the database. An optical network health monitoring method according to claim 19, wherein the predicted task parameters include task space data parameters, and the method further includes: Extract original performance data related to the optical network health performance prediction task from the database to synchronize the original performance data in the task space data. According to any one of claims 1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20 The optical network health monitoring method, wherein the optical network historical performance data includes at least one of the following: optical signal power, bit error rate, optical signal-to-noise ratio, and PRBS test data. An optical network management unit, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any one of claims 1 to 21. The optical network health monitoring method described above. An optical network system including: The optical network management unit as claimed in claim 22; A source endpoint device, connected to the optical network management unit, for sending or receiving optical network service flows; A plurality of intermediate node devices connected to the optical network management unit for forwarding optical network business flows; A sink endpoint device is connected to the optical network management unit and is used to receive or send optical network service flows. Computer-readable storage media stores computer-executable instructions for: Implement the optical network health monitoring method described in any one of claims 1 to 21.