Optical Transmission Network Performance Prediction Method, System, Electronic Device, and Storage Medium

Abstract

The embodiments of the present application relate to the field of communication and disclose an optical transmission network performance prediction method, a system, an electronic device, and a storage medium. In the present application, the optical transmission network performance prediction method includes the steps of: acquiring a performance prediction model; acquiring model update information of the performance prediction model by a horizontal federated learning technique, the model update information being obtained after training the performance prediction model based on performance data of each network element in a single domain, and being used to update the performance prediction model; reporting the model update information and the single domain topology information to a multi-domain management server, and having the multi-domain management server generate an optical transmission network digital twin model for making performance prediction for the optical transmission network according to the model update information and the single domain topology information.

Description

Technical Field

[0001] (Cross - reference to related applications) This application claims the priority of a Chinese patent application filed on March 10, 2022, with the application number 202210234032.2.

[0002] Embodiments of this application relate to the field of communications, and in particular, to a method, system, electronic device, and storage medium for predicting the performance of an optical transmission network.

Background Art

[0003] With the development of communication technologies, the global communication industry is moving from the Internet era and the cloud era towards the smart era. New opportunities and challenges drive the acceleration of the overall network model change and upgrade. To realize the digital model change of the operator network and provide services and user experiences with zero waiting time, zero contact, and zero failures for vertical industries and consumer users, autonomous network technology has been proposed to build network capabilities of self - configuration, self - repair, and self - optimization throughout the entire life cycle of operator network operation. To realize the digital model change of the network, it is necessary to establish a digital mirroring of the internal and external environments of the network and integrate capabilities such as simulation and preventive prediction. How to build such a simulation network has become an urgent problem to be solved.

Summary of the Invention

Problems to be Solved by the Invention

[0004] The main objective of the embodiments of this application is to propose a method, system, electronic device, and storage medium for predicting the performance of an optical transmission network, which can realize the construction of a simulation network in which an optical transmission network is built.

Means for Solving the Problems

[0005] To achieve the above object, an embodiment of the present application provides an optical transmission network performance prediction method used for a single-domain management server. This method includes the steps of obtaining a performance prediction model, and obtaining model update information of the performance prediction model by means of horizontal federated learning technology. The model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain, and is used to update the performance prediction model. Reporting the model update information and the single-domain topology information to a multi-domain management server, and generating an optical transmission network digital twin model for performing performance prediction on the optical transmission network based on the model update information and the single-domain topology information by the multi-domain management server.

[0006] To achieve the above object, an embodiment of the present application further provides an optical transmission network performance prediction method used for a multi-domain management server. This method includes the steps of obtaining model update information and single-domain topology information reported by a single-domain management server. The model update information is obtained after training the performance prediction model of the single-domain management server based on the performance data of each network element in the single domain by means of horizontal federated learning technology, and is used to update the performance prediction model. Generating an optical transmission network digital twin model for performing performance prediction on the optical transmission network based on the model update information and the single-domain topology information.

[0007] To achieve the above object, an embodiment of the present application further provides an optical transmission network performance prediction system, which includes a single-domain management server and a multi-domain management server. The single-domain management server is communicatively connected to the multi-domain management server. Here, the single-domain management server is configured to obtain a performance prediction model and obtain model update information of the performance prediction model by means of horizontal federated learning technology. The model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain. Then, the model update information and the single-domain topology information are reported to the multi-domain management server. The multi-domain management server is used to generate an optical transmission network digital twin model. The multi-domain management server is configured to obtain the model update information and the single-domain topology information reported by the single-domain management server. The model update information is obtained after training the performance prediction model of the single-domain management server based on the performance data of each network element in the single domain by means of horizontal federated learning technology. Based on the model update information and the single-domain topology information, it is used to generate an optical transmission network digital twin model for performing performance prediction on the optical transmission network.

[0008] An embodiment of the present application further provides an electronic device, which includes at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions that can be executed by the at least one processor. By executing the instructions by the at least one processor, the at least one processor can be caused to execute the above optical transmission network performance prediction method.

[0009] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, which realizes the above optical transmission network performance prediction method when the computer program is executed by a processor.

[0010] The optical transmission network performance prediction method proposed in this application is such that a single-domain management server acquires model update information, reports it to the multi-domain management server, and the multi-domain management server finally generates an optical transmission network digital twin model. That is, by distributing data to each single domain in horizontal federated learning, the problem of difficult training is overcome. Since the model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain, based on the model update information, the update of the performance prediction model can be more consistent with the actual processing ability of the actual network element. Therefore, the single-domain management server reports the model update information and the single-domain topology information to the multi-domain management server, and the multi-domain management server can generate an optical transmission network digital twin model based on the updated performance prediction model. This optical transmission network digital twin model is also a model that matches the actual processing and transmission ability in the optical transmission network, thereby realizing the simulation network construction for building the optical transmission network.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Embodiments for Carrying Out the Invention

[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will describe each embodiment of the present application in detail while combining the accompanying drawings. However, as can be understood by those skilled in the art, in each embodiment of the present application, many technical details have been provided to enable readers to better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solutions required to be protected in the present application can be realized. The division of the following embodiments is for the convenience of description and does not constitute any arbitrary limitation on the specific implementation modes of the present application. Each embodiment can be combined with each other and cited from each other on the premise of non-contradiction.

[0013] The embodiments of the present application relate to an optical transmission network performance prediction method. As shown in FIG. 1, the method includes the following steps.

[0014] Step 101, obtain a performance prediction model, Step 102: Obtain the model update information of the performance prediction model by means of horizontal federated learning technology. Here, the model update information is obtained after training the performance prediction model based on the performance data of each network element in a single domain, and is used to update the performance prediction model. Step 103: Report the model update information and the single-domain topology information to the multi-domain management server, and cause the multi-domain management server to generate an optical transmission network digital twin model for performing performance prediction on the optical transmission network based on the model update information.

[0015] The optical transmission network performance prediction method of this embodiment is used in a single-domain management server. The single-domain management server is a management server that manages each network element in one domain of the optical transmission network. Each single-domain management server can run a single-domain management control system of the manufacturer corresponding to this single domain. The optical transmission network (referred to as "OTN") can be divided into different management domains in the horizontal direction. Among them, a single management domain may be composed of OTN devices of a single equipment manufacturer, or may be composed of a certain network or sub-network of an operator. A single management domain is a single domain. Each single domain has one single-domain management server for communicating with the multi-domain management server. The multi-domain composition system can be run on the multi-single-domain management server to generate an optical transmission network digital twin model. The optical transmission network digital twin model can perform performance prediction on the optical transmission network or execute other data simulation requirements.

[0016] The global communication industry is moving from the Internet era and the cloud era towards the smart era. New opportunities and new challenges drive the acceleration of a comprehensive network model change and upgrade. Against this background, in 2019, the concept of Autonomous Networks (AN) was proposed. The AN concept enables the digital model change of operators' networks through the integration of network technology and digital technology, providing a service and user experience with zero waiting time, zero contact, and zero failures for vertical industries and consumer users, and aims to build network capabilities of self-configuration, self-healing, and self-optimization throughout the entire life cycle of operators' network operations. To achieve AN, Digital Twins (DT) technology has been constructed and is considered an important support and component for realizing the digital model change foundation, AN architecture, and technology in the industry. Network digital twins play an important enabling technical role in cases such as "low-cost trial and error", "smart decision-making", and "predictive maintenance" through their abilities to accurately perceive their own and external environmental states, establish digital mirroring of the internal and external environments, and integrate simulation and preventive prediction.

[0017] Digital twin, i.e., DT technology, can be understood as constructing a digital (virtual) model of a physical object in the digital space, continuously modifying this model using data from the physical object, updating the state of the model, matching the state of the model with the physical object throughout its entire life cycle, and being able to faithfully mirror the real-time operating state of the physical object. This model has become the digital twin body of the physical object (referred to as the digital twin). Based on the digital twin, by monitoring, analyzing, predicting, diagnosing, training, simulating, and feeding back the simulation results to the physical object, it is possible to assist in the optimization and decision-making for the physical object. Technologies related to the digital twin model structure, real-time updates regarding the digital model state, simulation analysis, and control decision-making by the digital twin can also be collectively referred to as digital twin DT technology. From this, it can be seen that how to build a DT model by associating the structural characteristics, state characteristics of the physical object mirrored in the twin, and the application scenarios of DT is an important technology of DT.

[0018] In the autonomous network system architecture, in the complex network environment of the OTN multi-domain, it is necessary to build a DT network model, generate the OTN DT network layer, and realize the analysis ability of the entire network OTN by DT technology. The DT network layer is an abstraction of the model of the entire physical network between domains. The communication between DT network elements is not restricted by the limitations of the physical network space, and the visibility and operability of DT network elements are not restricted by spatial limitations such as physical network domain division and management control limitations. For this, it is necessary for the OTN DT model to have the general abstract description ability for the physical network elements and the network element device function mechanisms of each domain.

[0019] Taking the modeling of the OTN performance prediction algorithm belonging to the DT perception-based algorithm model within a single-domain OTN network domain as an example, network element devices come from different device manufacturers. Each network element device has the same network element operation function in the same network domain environment, but attributes such as each network element's own device slot, model number, and service life are not the same. The optical performance of device devices, optical fibers, etc. has different laws of change in actual operation and use. The DT performance prediction model obtained by sampling and training only some network elements or a single network element device only reflects the performance change characteristics of this part of the network element or this network element, and for other network element devices within the single domain or network element devices in other inter-domain networks, it does not have universality, has poor generalization ability, and problems such as overfitting are likely to occur during inference. As a result, the accuracy of OTN network performance prediction analysis is low and the authority is insufficient. However, if the performance data of each network element in each domain is reported to the multi-domain compilation system for unified training, there is a problem of data privacy leakage: since each OTN physical network domain is under the single-domain management and control system of different device manufacturers, in the construction and training of the inter-domain digital twin layer model, each physical single domain has a claim for data privacy, and it is not easy to directly report the OTN network performance data within each domain to the multi-domain compilation system for unified training. How to solve the problem of overfitting of the performance prediction function model caused by local training while solving the problem of sample data privacy leakage caused by centralized training has become a bottleneck in the modeling of the OTN DT case function model and needs to be solved urgently. The emergence of federated learning technology provides an effective route to solve the above problems.

[0020] The problems of modeling that federated learning can overcome are as follows: 1. Data does not come from local: Due to data security and privacy protection, data does not come from local, and the problem of non-convergence is prominent. 2. Model generalization ability: Since the represented training dataset is small, the problem of poor model generalization ability. 3. Model training efficiency: The problem of model training efficiency due to overemphasis on centralized cloud computing training and ignoring the computing power of edge devices.

[0021] The advantages of federated learning are as follows: 1. Only utilize local data training, without exchanging the data itself, and replace the encrypted updated model parameters. 2. Use data in different environments (time, location) of devices to perform model training. After the common model is updated, it is sent to the devices to improve the model generalization ability. 3. Use the computing power of edge devices to perform parallel training to improve the model training efficiency.

[0022] The flow of horizontal federated learning includes the following steps: Step 1, each participant calculates the model gradient locally, uses encryption technologies such as homomorphic encryption, differential privacy or secret sharing to camouflage the gradient information, and sends the result after camouflage (referred to as encrypted gradient) to the aggregation server. Step 2, the server performs a secure aggregation operation, for example, uses weighted average based on homomorphic encryption. Step 3, the server sends the result after aggregation to each participant. Step 4, each participant decrypts the received gradient and updates its respective model parameters using the decrypted gradient result. As described above, this application constructs a method mechanism for constructing an OTN DT network function model by adopting horizontal federated learning technology.

[0023] The optical transmission network performance prediction method of this application is as follows: The single-domain management server acquires model update information, reports it to the multi-domain management server, and the multi-domain management server finally generates an optical transmission network digital twin model, that is, overcomes the training difficulty problem caused by data being distributed in each single domain in horizontal federated learning. Since the model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain, based on the model update information, the update of the performance prediction model can be more consistent with the actual processing capacity of the actual network element. Therefore, the single-domain management server reports the model update information and the single-domain topology information to the multi-domain management server, and the multi-domain management server can generate an optical transmission network digital twin model based on the updated performance prediction model. This optical transmission network digital twin model is also a model that matches the actual processing and transmission capacity in the optical transmission network, thereby realizing the simulation network construction for building the optical transmission network.

[0024] Hereinafter, the implementation details of the optical transmission network performance prediction method of this embodiment will be specifically described. The following content is only for facilitating the understanding of the provided implementation details and is not essential for implementing this solution.

[0025] In step 101, the single-domain management server acquires a performance prediction model. Here, the performance prediction model may be stored in the single-domain management server in advance, or may be acquired by the single-domain management server from the multi-domain management server or other electronic devices.

[0026] In one example, as shown in FIG. 2, the optical transmission network performance prediction method may be used in a scenario where each network element in a single domain has training ability (hereinafter referred to as Solution A). In Solution A, the performance prediction model may be a single-domain performance prediction model, and the model update information may be a single-domain digital twin model obtained based on the updated single-domain performance prediction model. The single-domain management server realizes obtaining the model update information of the performance prediction model by the following method, sending the single-domain performance prediction model to each network element in this single domain, obtaining the parameter update amount of the single-domain performance prediction model reported by each network element. Here, each network element trains the single-domain performance prediction model respectively based on the local performance data, calculates the parameter update amount, updates the parameters of the single-domain performance prediction model based on the parameter update amount reported by each network element, obtains the updated single-domain performance prediction model, and obtains the single-domain digital twin model based on the updated single-domain performance prediction model and the single-domain topology information of this single domain.

[0027] In this embodiment, by using the single-domain performance prediction model as the performance prediction model and the single-domain digital twin model obtained based on the updated single-domain performance prediction model as the model update information, the single-domain management server sends the single-domain performance prediction model to each network element in this single domain, and each network element can adapt the updated single-domain performance prediction model to the performance characteristics of each network element by training the single-domain performance prediction model respectively. Since the attributes of each network element itself, such as the device slot, model number, and service life, are different, problems such as poor generalization ability of performance prediction model training and easy occurrence of overfitting during inference can be solved, and effects such as strong generalization ability of the performance prediction model and high prediction accuracy can be realized.

[0028] In another example, the optical transmission network performance prediction method may be used in a scenario where only the single-domain management server has training capabilities in a single domain, while other network elements do not have training capabilities (hereinafter referred to as Solution B). In Solution B, the performance prediction model may be a network performance prediction model, the model update information may be the parameter update amount of the network performance prediction model, and the single-domain management server can realize obtaining the model update information of the performance prediction model in the following way. Based on the performance data of each network element, train a network performance prediction model to obtain the parameter update amount of the network performance prediction model. Here, the parameter update amount of the network performance prediction model is reported to the multi-domain management server, and the multi-domain management server is used to update the network performance prediction model based on the parameter update amount of the network performance prediction model.

[0029] In this embodiment, by using the network performance prediction model as the performance prediction model and the parameter update amount of the network performance prediction model as the model update information, the single-domain management server can train the network performance prediction model based on the performance data of each network element by itself, obtain the parameter update amount of the network performance prediction model, and the multi-domain management server integrates the parameter update amounts of the network performance prediction models of each single domain to update the network performance prediction model. By having the single-domain management server perform model training, the number of devices required for training can be reduced, computing resources can be saved, and the optical transmission network performance prediction method can be realized even in a scenario with relatively scarce computing resources. At the same time, since the attributes of each network element, such as its own device slot, model number, and service life, are different, problems such as poor generalization ability of performance prediction model training and easy occurrence of overfitting during inference can be solved, and effects such as strong generalization ability of the performance prediction model and high prediction accuracy can be realized.

[0030] In step 102, the single-domain management server obtains the model update information of the performance prediction model. Here, the model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain, and is used to update the performance prediction model. The model update information may be calculated by the single-domain management server, or may be calculated by other network elements in the single domain where the single-domain management server is located.

[0031] JPEG0007706026000001.jpg140170

[0032] JPEG0007706026000002.jpg236170

[0033] In step 103, the single-domain management server reports the model update information and the single-domain topology information to the multi-domain management server, and the multi-domain management server generates an optical transmission network digital twin model for performing performance prediction on the optical transmission network based on the model update information and the single-domain topology information. Here, after receiving the model update information and the single-domain topology information reported by multiple single-domain management servers, the multi-domain management server generates an optical transmission network digital twin model. The optical transmission network digital twin model is a digital twin model of the entire optical transmission network, that is, a digital twin model including the multiple reported single domains, and digitally simulates the optical transmission network.

[0034] In one example, based on the parameter update amounts reported by each network element, the parameters of the single-domain performance prediction model are updated to obtain the updated single-domain performance prediction model. Specifically, the single-domain management server needs to repeatedly update the single-domain performance prediction model. Here, the repeated update involves transmitting the updated single-domain performance prediction model to each network element within the single domain, obtaining the parameter update amounts reported again by each network element, updating the single-domain performance prediction model based on the parameter update amounts reported again, and starting the next repeated update until the repeated stop condition is satisfied. Here, the repeated stop condition includes the convergence of the parameter update amount of the performance prediction model.

[0035] In this embodiment, by repeatedly updating the single-domain performance prediction model, the prediction effect of the single-domain performance prediction model is made to conform to the actual performance of the device, realizing the update of the single-domain performance prediction model. The optical transmission network digital twin model generated by the multi-domain management server has a better simulation effect.

[0036] JPEG0007706026000003.jpg103170

[0037] JPEG0007706026000004.jpg187170

[0038] JPEG0007706026000005.jpg138170

[0039] JPEG0007706026000006.jpg81170

[0040] JPEG0007706026000007.jpg131170

[0041] JPEG0007706026000008.jpg89170

[0042] JPEG0007706026000009.jpg53170

[0043] JPEG0007706026000010.jpg142170

[0044] JPEG0007706026000011.jpg40170

[0045] Step 2: Each single domain constructs its DT network layer of this domain based on functional models such as the network topology model of this domain (i.e., single domain topology information), the basic model of each network element node, and the network performance prediction of this domain, and reports according to the requirements of the multi-domain composition system.

[0046] Step 3: The multi-domain composition system connects the DT network layers of each domain to further construct a DT network layer model that generates the entire network (i.e., the digital twin model of the optical transmission network).

[0047] JPEG0007706026000012.jpg80170

[0048] JPEG0007706026000013.jpg94170

[0049] The features of Solution A are as follows: 1. Each network element node within a single domain itself has the ability of AI training, and through performance sampling and model training for this node, it can timely sense and predict the amount of change in the performance of this network element.

[0050] 2. The single-domain management control system also has AI training capabilities and, from the perspective of federated learning technology, plays the role of an edge server in this solution. Each network element node in the single domain reports the performance prediction model parameter gradient (or parameter update amount) trained by its own AI algorithm to this domain management control system. This domain management control system aggregates the model parameter gradients (or parameter update amounts) of all reported network elements and, at the same time, updates the parameters of the common domain performance prediction model (i.e., the single-domain performance prediction model) constructed by this domain management control system.

[0051] 3. The single-domain management control system broadcasts the updated common domain performance prediction model parameters to each network element node in this domain.

[0052] 4. Each network element node updates the performance prediction model parameters of this network element using the common model parameters and, based on this iteration, starts the next model training and interaction with the single-domain management control system.

[0053] 5. If a communication failure occurs between a certain network element node and the single-domain management control system, or a failure occurs in the node itself, resulting in the inability to upload the model parameter gradient (or parameter update amount) of this network element node, such a situation will not affect the update of the common model parameters of the single-domain management control system itself and the interaction with other network element nodes.

[0054] 6. Such a solution can solve problems such as poor generalization ability in performance prediction model training and overfitting that are likely to occur in model prediction inference due to differences in device slots, model numbers, service lives, etc. of each network element node in this domain.

[0055] 7. Each single-domain management and control system generates the OTN DT network layer of this domain constructed by the OTN physical network topology connection relationship of each network element DT model (including this network element DT basic structure model, DT performance prediction function model, etc.) of this domain based on the common performance prediction model of this domain obtained by the final training and the network topology information of this domain, and sends it to the multi-domain orchestration system. The multi-domain orchestration system connects the OTN DT network layers reported by each single domain to obtain the entire network OTN DT network layer. Since it is the entire network DT network layer obtained by connecting them, the performance prediction models between the single-domain parts included are different.

[0056] 8. When the single-domain management and control system and each network element inside this domain are constructed by the network of the same equipment manufacturer, before reporting the network element model parameter gradient (or parameter update amount) of this network element, it is not necessary to encrypt the performance prediction model parameter gradient of this network element.

[0057] 9. It is necessary to ensure that the RNN training model of each network element node distributed within a single domain that adopts the horizontally collaborative learning and training common performance prediction model (taking RNN as an example here) has the same structure as the RNN training model of the single-domain management and control system that plays the role of the edge server in training, including all vector element attributes input from the RNN model, the number of vector elements, the number of layers of the RNN model, the number of neurons in each layer, the activation function between layers, the connection relationship, the output vector attributes, the number of output vector elements, etc., to ensure the unified training and synchronous refresh of the RNN model parameters.

[0058] The technical effects of this solution are reflected in solving problems such as the poor generalization ability of the DT performance prediction model training due to different attributes of each network element itself, such as device slot, model number, and service life, and the easy occurrence of overfitting during inference, and realizing effects such as strong generalization ability and high prediction accuracy of the OTN DT performance prediction model.

[0059] In one example, the optical transmission network performance prediction system is applied to Solution B, and the specific steps include the following: Step 1: The multi-domain composition and each single domain are trained by the horizontal collaborative learning technology to obtain a multi-domain OTN network performance prediction function model (i.e., the network performance prediction model).

[0060] Step 2: The multi-domain composition system predicts the common function model and the encrypted polo information reported by each single domain based on the multi-domain OTN network performance, and generates a network OTN DT network layer overall model (i.e., the optical transmission network digital twin model).

[0061] JPEG0007706026000014.jpg66170

[0062] JPEG0007706026000015.jpg71170

[0063] JPEG0007706026000016.jpg98170

[0064] The features of Solution B are as follows: 1. Each network element node within a single domain does not have the ability to model AI training. Each single-domain management and control system has the ability to model AI training, and the multi-domain composition system between manufacturers also has the ability to model AI training.

[0065] 2. From the perspective of federated learning technology, in this solution, the multi-domain composition system serves as an edge server. Each single-domain management and control system reports the performance prediction model parameter gradients (or parameter update amounts) trained by its own AI algorithm to the multi-domain composition system using homomorphic encryption technology. The multi-domain composition system performs homomorphic encryption aggregation processing on the model parameter gradients (or parameter update amounts) of all the reported single-domain management and control systems, and at the same time updates the multi-domain performance prediction common model parameters constructed by the multi-domain composition system.

[0066] 3. The multi-domain composition system broadcasts the updated multi-domain performance prediction common model parameters to each single-domain management and control system.

[0067] 4. Each single-domain management and control system refreshes the domain performance prediction model parameters using the common model parameters, and based on this iteration, starts the next model training and interaction with the multi-domain composition system.

[0068] 5. If a communication failure occurs between a certain single-domain management and control system and the multi-domain composition system, or a failure occurs in the node itself, resulting in the inability to upload the model parameter gradients (or parameter update amounts) of this domain management and control system, such a situation will not affect the update of the multi-domain composition system's own common model parameters and the interaction with other single-domain management and control systems.

[0069] 6. Such a solution can solve problems such as poor generalization ability in performance prediction model training caused by differences in each network element node device slot, model number, service life, etc. between domains, and overfitting that is likely to occur in model prediction inference. The performance prediction model is universal between domains and has stronger generalization ability.

[0070] 7. The multi-domain composition system generates the inter-domain OTN DT network layer of the entire network based on the multi-domain common performance prediction model obtained in the final training and the encrypted network topology information reported from each domain. The performance prediction models of all OTN domains are the same.

[0071] 8. Since each domain is constructed by the networks of different equipment manufacturers, each domain management control system needs to encrypt the performance prediction model parameter gradient (or parameter update amount) of this domain before reporting it, and at the same time, the network topology information of this domain reported by each domain to the multi-domain management control system is encrypted as needed.

[0072] 9. It is necessary to adopt the horizontal federated learning training common performance prediction model to ensure that the RNN training model (taking RNN as an example) of each single-domain management control system has the same structure as the RNN training model of the multi-domain composition system that plays the role of the edge server in training, including all vector element attributes input from the RNN model, the number of vector elements, the number of layers of the RNN model, the number of neurons in each layer, the activation function between layers, the connection relationship, the output vector attributes, the number of output vector elements, etc., to ensure the unified training and synchronous refresh of the RNN model parameters.

[0073] The technical effects of this solution are reflected in solving problems such as poor generalization ability of the inter-domain OTN DT performance prediction model training and easy occurrence of overfitting during inference due to different attributes such as the device slot, model number, and service life of each manufacturer's network element, and realizing effects such as strong generalization ability and high prediction accuracy of the inter-domain OTN DT performance prediction model.

[0074] The step divisions of the various methods described above are only for the purpose of clarifying the description, and can be integrated as one step during implementation or divided into any steps. As long as the same logical relationship is included, all are within the protection scope of this patent. Although minor modifications to the algorithm or flow are added, or minor designs are introduced, it does not change the fact that the core designs of both the algorithm and the flow are within the protection scope of this patent.

[0075] The embodiments of this application further relate to a method for predicting the performance of an optical transmission network. As shown in FIG. 11, the multi-domain management server includes the following steps: Step 1101: Obtain the model update information and single-domain topology information reported by the single-domain management server. Here, the model update information is obtained after training the performance prediction model of the single-domain management server based on the performance data of each network element in the single domain by means of horizontal federated learning technology, and is used to update the performance prediction model. Step 1102: Generate an optical transmission network digital twin model for predicting the performance of the optical transmission network based on the model update information and the single-domain topology information.

[0076] The multi-domain management server of this embodiment is used in the method for predicting the performance of an optical transmission network. The multi-domain management server is a server that manages all single-domain management servers in the optical transmission network. The multi-domain management server runs a multi-domain orchestration system to implement the method for predicting the performance of the optical transmission network.

[0077] In one example, the performance prediction model is a single-domain performance prediction model, the model update information is a single-domain digital twin model obtained based on the updated single-domain performance prediction model, and the multi-domain management server is implemented in the following manner. Based on the model update information and the single-domain topology information, an optical transmission network digital twin model is generated, and based on the single-domain topology information, the single-domain digital twin models reported by the single-domain management servers of each single-domain are combined to obtain an optical transmission network digital twin model.

[0078] In another example, the performance prediction model is a network performance prediction model, the model update information is the parameter update amount of the network performance prediction model, and the multi-domain management server is implemented in the following manner. Based on the model update information and the single-domain topology information, an optical transmission network digital twin model is generated, based on the parameter update amount of the network performance prediction model, the network performance prediction model is updated, and based on the updated network performance prediction model and the single-domain topology information, an optical transmission network digital twin model is generated.

[0079] JPEG0007706026000017.jpg159170

[0080] This embodiment is the corresponding embodiment to the above-mentioned embodiment, and it is obvious that this embodiment can be implemented in cooperation with the above-mentioned embodiment. The details of the related technologies mentioned in the above embodiment are still valid in this embodiment, and for the sake of reducing duplication, they will not be described further here. Correspondingly, the details of the related technologies mentioned in this embodiment can be applied to the above embodiment.

[0081] The embodiments of the present application further relate to an optical transmission network performance prediction system. As shown in FIG. 12, It includes a single-domain management server 1201 and a multi-domain management server 1202. The single-domain management server 1201 is communicatively connected to the multi-domain management server 1202. Here, the single-domain management server 1201 is to obtain a performance prediction model and obtain model update information of the performance prediction model by means of horizontal federated learning technology. The model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain. Then, the model update information and the single-domain topology information are reported to the multi-domain management server 1202, and are used for the multi-domain management server 1202 to generate an optical transmission network digital twin model. The multi-domain management server 1202 is to obtain the model update information and the single-domain topology information reported by the single-domain management server 1201. The model update information is obtained after training the performance prediction model of the single-domain management server 1201 based on the performance data of each network element in the single domain by means of horizontal federated learning technology. Based on the model update information and the single-domain topology information, it is used to generate an optical transmission network digital twin model for performing performance prediction on the optical transmission network.

[0082] This embodiment is a system embodiment corresponding to the above-mentioned embodiment, and it is obvious that this embodiment can be implemented in cooperation with the above-mentioned embodiment. The details of the related technology mentioned in the above embodiment are still valid in this embodiment. To avoid duplication, no further explanation is given here. Correspondingly, the details of the related technology mentioned in this embodiment may also be applied to the above embodiment.

[0083] An embodiment of the present application further relates to an electronic device. As shown in FIG. 13, it includes at least one processor 1301 and a memory 1302 communicatively connected to the at least one processor. Here, the memory 1302 stores instructions that can be executed by the at least one processor 1301, and the instructions cause the method of any of the above embodiments to be executed by the at least one processor 1301.

[0084] Here, the memory 1302 and the processor 1301 are connected in a bus manner. The bus may include any number of interconnected buses and bridges, and the bus connects various circuits of one or more processors 1301 and the memory 1302. The bus can further connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, etc. Since these are all well-known in the art, they will not be further described in this specification. The bus interface provides an interface between the bus and the transceiver. The transceiver may be a single element or multiple elements. For example, it provides a unit for communicating with various other devices via a plurality of receivers, transmitters, and a transmission medium. The information processed by the processor 1301 is transmitted over a wireless medium via an antenna. Further, the antenna receives the information and transmits the information to the processor 1301.

[0085] The processor 1301 is responsible for the management of the bus and general processing, and can provide various functions including timing, peripheral interface, voltage regulation, power management, and other control functions. However, the memory 1302 may be used to store information used when the processor executes operations.

[0086] An embodiment of the present application relates to a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the embodiments of the above method are realized.

[0087] That is, as can be understood by those skilled in the art, all or some of the steps for implementing the method of the above embodiments may be instructed by a program to be completed by related hardware. This program is stored in a storage medium and includes several instructions for causing a device (which may be a one-chip microcomputer, a chip, etc.) or a processor to execute all or some of the steps of the method of each embodiment of the present application. However, the aforementioned storage medium includes various media capable of storing program codes such as a USB disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

Claims

1. An optical transmission network performance prediction method used for a single domain management server, comprising: obtaining a performance prediction model; obtaining model update information of the performance prediction model by means of horizontal federated learning technology, where the model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain, and is used to update the performance prediction model; reporting the model update information and the single domain topology information to a multi-domain management server, and causing the multi-domain management server to generate an optical transmission network digital twin model for performing performance prediction on the optical transmission network based on the model update information and the single domain topology information. An optical transmission network performance prediction method.

2. The performance prediction model includes a single domain performance prediction model. The model update information includes obtaining a single domain digital twin model based on the updated single domain performance prediction model. The obtaining of the model update information of the performance prediction model includes: transmitting the single domain performance prediction model to each network element in the single domain; obtaining the parameter update amount of the single domain performance prediction model reported by each network element, where each network element trains the single domain performance prediction model respectively based on local performance data and calculates the parameter update amount; updating the parameters of the single domain performance prediction model based on the parameter update amount reported by each network element to obtain the updated single domain performance prediction model; and obtaining the single domain digital twin model based on the updated single domain performance prediction model and the single domain topology information of the single domain. The optical transmission network performance prediction method according to Claim 1.

3. The updating of the parameters of the single domain performance prediction model based on the parameter update amount reported by each network element to obtain the updated single domain performance prediction model includes: iteratively updating the single domain performance prediction model. Here, the iterative update includes transmitting the updated single-domain performance prediction model to each network element in the single domain, obtaining the parameter update amount reported again by each network element, updating the single-domain performance prediction model based on the parameter update amount reported again, and starting the next iterative update until the iteration stop condition is satisfied. The optical transmission network performance prediction method according to claim 2.

4. Here, the performance prediction model includes a network performance prediction model. The model update information includes the parameter update amount of the network performance prediction model. The obtaining of the model update information of the performance prediction model as described above includes training the network performance prediction model based on the performance data of each network element and obtaining the parameter update amount of the network performance prediction model. Here, the parameter update amount of the network performance prediction model is reported to the multi-domain management server, and the multi-domain management server is used to update the network performance prediction model based on the parameter update amount of the network performance prediction model. The optical transmission network performance prediction method according to claim 1.

5.

6.

7. An optical transmission network performance prediction method used in a multi-domain management server, comprising: obtaining model update information and single-domain topology information reported by a single-domain management server, where the model update information is obtained after training the performance prediction model of the single-domain management server based on the performance data of each network element in the single domain by using cross-domain collaborative learning technology and is used to update the performance prediction model; generating an optical transmission network digital twin model for performing performance prediction on the optical transmission network based on the model update information and the single-domain topology information. Optical transmission network performance prediction method.

8. The performance prediction model includes a single-domain performance prediction model. The model update information includes obtaining a single-domain digital twin model based on the updated single-domain performance prediction model. Generating an optical transmission network digital twin model based on the model update information and the single-domain topology information described above includes: Based on the single-domain topology information, connecting the single-domain digital twin models reported by the single-domain management servers of each single domain to obtain the optical transmission network digital twin model. The optical transmission network performance prediction method according to claim 7.

9. The performance prediction model includes a network performance prediction model. The model update information includes the parameter update amount of the network performance prediction model. Generating an optical transmission network digital twin model based on the model update information and the single-domain topology information described above includes: Updating the network performance prediction model based on the parameter update amount of the network performance prediction model; Generating an optical transmission network digital twin model based on the updated network performance prediction model and the single-domain topology information. The optical transmission network performance prediction method according to claim 7.

10. An optical transmission network performance prediction system, including: A single-domain management server and a multi-domain management server, where the single-domain management server is communicatively connected to the multi-domain management server; Here, the single-domain management server obtains a performance prediction model and obtains model update information of the performance prediction model by means of horizontal federated learning technology. The model update information is obtained after training the performance prediction model based on the performance data of each network element in the single domain. The model update information and the single-domain topology information are reported to the multi-domain management server and used by the multi-domain management server to generate an optical transmission network digital twin model. The multi-domain management server obtains the model update information and the single-domain topology information reported by the single-domain management server. The model update information is obtained after training the performance prediction model of the single-domain management server based on the performance data of each network element in the single-domain by the horizontal federated learning technology. Based on the model update information and the single-domain topology information, it is used to generate the optical transmission network digital twin model for performing performance prediction on the optical transmission network. An optical transmission network performance prediction system.

11. An electronic device, including at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor can execute the optical transmission network performance prediction method according to any one of claims 1 to 6, or the optical transmission network performance prediction method according to any one of claims 7 to 9. An electronic device.

12. A computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, it realizes the optical transmission network performance prediction method according to any one of claims 1 to 6, or the optical transmission network performance prediction method according to any one of claims 7 to 9. A computer-readable storage medium.

Images