Dynamic multiband optical network transmission capacity optimization method based on GSNR perception

Abstract

The invention provides a dynamic multiband optical network transmission capacity optimization method based on GSNR perception, and the method comprises the steps: constructing a nonlinear mapping relation model between the GSNR characteristics of an input signal and the pumping configuration parameters of a Raman amplifier and the GSNR characteristics of an output signal, and enabling the model to sense the GSNR characteristics of the input signal in real time, and the GSNR distribution change of the output signal is predicted, so that the GSNR distribution in the dynamic input scene is accurately predicted. The pumping configuration parameters of the Raman amplifier are dynamically adjusted in combination with an intelligent optimization algorithm, and real-time sensing and accurate regulation and control of the GSNR of a transmission system are realized, so that the GSNR average value and flatness of a channel in an optical network are improved, and the transmission capacity is remarkably improved. According to the method, the transmission capacity of the dynamic multiband optical network is improved, the problem of adaptability of a traditional method in a dynamic scene is effectively solved, and the performance and stability of the optical network are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of optical communication, specifically relates to the optimization of optical network transmission performance, and particularly to a transmission capacity optimization method based on generalized signal-to-noise ratio (GSNR) perception in a multi-band optical network. Background Art

[0002] With the development of new technologies such as the Internet, cloud computing, and 5G / 6G, the demand for data traffic has shown an explosive growth, and the existing optical communication networks are facing huge capacity expansion pressures. The traditional single-band optical network is limited in capacity improvement, while the multi-band optical network has become an important solution for future transmission capacity expansion. By introducing new bands, the optical network can utilize the entire communication window of standard single-mode fiber (SSMF), significantly expanding the transmission bandwidth. However, in a multi-band optical network, the transmission performance of signal channels is affected by complex factors such as nonlinear effects, amplified spontaneous emission noise (ASE), and stimulated Raman scattering (SRS), and the generalized signal-to-noise ratio shows significant unevenness, thus limiting the further improvement of the overall transmission capacity.

[0003] As an optical amplifier based on nonlinear optical effects, the Raman amplifier has become an important technical option in multi-band optical networks due to its wide gain bandwidth and low noise figure. Compared with erbium-doped fiber amplifiers (EDFAs), Raman amplifiers have higher flexibility, and by adjusting the wavelength and power of the pump light, arbitrary gain profile configurations can be achieved in the multi-band range. However, traditional Raman amplifier configuration methods usually rely on fixed signal input conditions and do not consider the real-time changes in signal transmission in dynamic multi-band optical networks. Existing research mostly uses numerical modeling and simulation methods to optimize the gain profile of Raman amplifiers, but these methods mainly focus on improving the amplifier gain performance and ignore the importance of GSNR optimization for transmission capacity improvement.

[0004] In recent years, with the development of machine learning technology, modeling methods based on artificial neural networks (ANNs) have gradually been applied to the pump design of Raman amplifiers. By training the model, researchers can establish the mapping relationship between pump parameters and the Raman gain profile, providing support for amplifier optimization. However, these methods still have limitations: on the one hand, they do not directly optimize the transmission capacity for multi-band optical networks; on the other hand, traditional models are usually trained for static input signals and lack adaptability to dynamic signal scenarios. When the signal conditions change, the prediction ability and optimization effect of the model will significantly decline.

[0005] In a multi-band optical network, the flatness and average value of the GSNR are important factors affecting the transmission capacity. The uneven distribution of the GSNR among channels will lead to the bottleneck effect of the system performance, making it difficult to improve the transmission rate of some channels, thus restricting the growth of the overall capacity. Therefore, how to optimize the GSNR distribution of the multi-band optical network under dynamic input conditions to improve the transmission capacity of the system is an important challenge in current optical communication research.

[0006] To address the above problems, there is an urgent need for a method that can sense the GSNR distribution and dynamically optimize the transmission performance of the multi-band optical network. This method should comprehensively consider the synergistic effect of the characteristics of the dynamically changing input signal and the pump power configuration, and use an efficient optimization mechanism to achieve real-time adjustment of the Raman amplifier configuration, thereby enhancing the overall transmission capacity of the multi-band optical network. Such a method should break through the application limitations of traditional technologies in static scenarios and have good adaptability and scalability to meet the actual needs of the efficient operation and capacity expansion of optical networks in various dynamic transmission environments. Summary of the Invention

[0007] Object of the Invention: The present invention aims to solve the problem that the overall transmission capacity is restricted due to the uneven distribution and insufficient optimization of the generalized signal-to-noise ratio (GSNR) in the transmission of multi-band optical networks. By proposing a method for optimizing the transmission capacity of a dynamic multi-band optical network based on GSNR sensing, the present invention hopes to optimize the configuration of Raman amplifiers in a dynamic transmission scenario, achieve real-time sensing and precise control of the GSNR, thereby enhancing the average value and flatness of the GSNR of each channel in the optical network and significantly improving the transmission capacity. The present invention constructs a GSNR sensing model to accurately predict the change in the GSNR distribution under dynamic input signal conditions, and combines an intelligent optimization algorithm to dynamically adjust the pump configuration of the Raman amplifier to ensure global optimality of the transmission performance in various dynamic scenarios. The present invention aims to reduce the complexity of the optimization process while improving the optimization efficiency, providing efficient support for the rapid deployment and operation of multi-band optical networks.

[0008] Technical Solution: The present invention provides a method for optimizing the transmission capacity of a dynamic multi-band optical network based on GSNR sensing, and the method includes:

[0009] By constructing a dynamic signal transmission model for a multi-band optical network, the input signal GSNR curves and their corresponding pump configuration parameters in different transmission scenarios are collected to generate a dataset for modeling and optimization. The dataset includes dynamically changing input GSNR curves, the pump power configurations of Raman amplifiers, and output GSNR curves, which are used for subsequent model training and optimization algorithm processes. To improve the efficiency of modeling and optimization, this data generation process groups channels through feature dimensionality reduction methods, maps the multi-band GSNR data into several key channel feature sets, ensures reasonable data dimensions, and simultaneously covers the typical characteristics of GSNR data in multi-band dynamic scenarios.

[0010] Construct an artificial neural network model for dynamic GSNR perception, which realizes the non-linear mapping relationship from the input signal GSNR curve and pump configuration parameters to the output GSNR curve. The modeling process includes the optimization of the model architecture and the training of parameters. First, in terms of the model architecture, by scanning different numbers of hidden layers and node scales, combined with different activation functions, a model structure that can accurately predict the GSNR distribution in multi-band signal transmission is designed. Second, by optimizing training parameters (such as learning rate, optimizer, etc.), the model is trained and verified using the dataset generated by the experimental simulation of the multi-band optical network. Finally, a GSNR perception model suitable for dynamic input scenarios is obtained. The training process of this model calculates the loss function based on the predicted output GSNR curve and the true value, and the training goal is to minimize the error between the two to ensure that the model has good generalization ability in the dynamic multi-band optical network.

[0011] Based on the constructed GSNR perception model, the present invention proposes a pump configuration optimization method in dynamic scenarios with the goal of improving transmission capacity. Specifically, an optimization objective function is designed to maximize the average value and distribution flatness of GSNR, and the pump power configuration of the Raman amplifier is globally searched and parameter optimized through intelligent optimization algorithms. The optimization process includes the following steps: in the initialization stage, multiple candidate pump configurations are generated based on the multi-dimensional parameter space as the initial solution set; in the iterative optimization process, the search direction and step size of the candidate solutions are dynamically adjusted to guide the solution set to approach the optimal region; in each iteration, the objective function value is calculated in real time based on the GSNR perception model, and the fitness of the candidate solutions is evaluated. Finally, the optimal pump configuration that maximizes the objective function value is selected. Through the adaptive search mechanism of this algorithm, the coordinated optimization of pump parameters and GSNR distribution in dynamic scenarios is realized.

[0012] Apply the optimization method to the specific transmission scenario of a dynamic multi - band optical network. Analyze the GSNR distribution of the real - time input signal through the GSNR perception model, and dynamically adjust the pump power configuration in combination with the particle swarm optimization algorithm, so as to effectively improve the generalized signal - to - noise ratio. During the optimization process, the model provides real - time feedback, enabling the pump configuration to dynamically adapt to the changes in the input signal, achieving the optimization and flattening of the GSNR distribution among the channels of the multi - band optical network, thereby enhancing the overall transmission capacity of the system.

[0013] A method for optimizing the transmission capacity of a dynamic multi - band optical network based on GSNR perception provided by the present invention is applicable to various optical network scenarios, including but not limited to systems such as ultra - long - distance optical transmission. Due to the large number of transmission channels and strong spectral non - flatness, there is a very high demand for optimizing the transmission capacity in a dynamic multi - band optical network. The present invention provides an efficient and highly adaptable optimization method for the above - mentioned transmission scenarios by combining GSNR perception and intelligent optimization algorithms.

[0014] In some embodiments, the optimization algorithm adopts the particle swarm optimization method (PSO). Each particle in the population corresponds to a set of pump power configurations, and the fitness function is the value of the objective function. Through the dynamic update of the particle position and velocity, the population gradually converges to the optimal solution. During the optimization process, the optimization algorithm utilizes the real - time feedback provided by the GSNR perception model, dynamically combines the adjustment of the pump power with the distribution characteristics of the GSNR among the channels of the multi - band optical network, and realizes the global optimization of the transmission performance among the channels.

[0015] Beneficial effects: Compared with the prior art, the present invention has the following advantages: (1) Dynamic optimization and high - transmission - capacity improvement: By combining the GSNR perception model and the intelligent optimization algorithm, the present invention can adapt to the signal change conditions of the dynamic multi - band optical network in real time and optimize the GSNR distribution among the channels. Under the condition of limited resources, it significantly improves the average value and flatness of the GSNR among the channels, overcomes the problem of insufficient adaptability of the traditional method in the dynamic transmission scenario, and realizes a higher transmission capacity for the multi - band optical network. (2) Real - time feedback and efficient configuration: The method of the present invention analyzes the dynamic input conditions in real time through the GSNR perception model, and uses the feedback results to guide the optimization algorithm to adjust the pump configuration parameters. Users can quickly adjust the configuration according to the optimization results, achieve efficient optimization in the dynamic transmission scenario, and provide guarantee for the stable operation of the multi - band optical network.

[0016] A method for optimizing the transmission capacity of a dynamic multi-band optical network based on GSNR perception realizes the global optimization of the transmission performance between channels in the dynamic multi-band optical network by constructing an accurate GSNR perception model and an optimization algorithm, effectively improving the transmission capacity of the system. Compared with the prior art, this method breaks through the limitation only applicable to static transmission scenarios, can not only optimize the pump configuration of Raman amplifiers in real time under dynamic input conditions, but also significantly improve the flatness and average value of the GSNR distribution, providing strong technical support for the efficient deployment and operation of multi-band optical networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of the method for optimizing the transmission capacity of a dynamic multi-band optical network based on GSNR perception of the present invention;

[0018] Figure 2 It is a schematic diagram of the framework for optimizing the transmission capacity of a dynamic multi-band optical network based on GSNR perception of the present invention;

[0019] Figure 3 It is a comparison chart of GSNR prediction results of the GSNR perception model of the present invention and its comparison model in a dynamic transmission scenario.

[0020] Figure 4 It is a comparison chart of capacity optimization results of the GSNR perception-based optimization method of the present invention and a comparison scheme in a dynamic transmission scenario. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0022] In the context of the rapid development of information technology today, the wide application of cloud computing, the Internet of Things, and 5G / 6G communications has led to an exponential growth in data traffic. As the core bearing platform of the information superhighway, optical networks are facing unprecedented demands for expanding transmission capacity. Traditional single-band optical networks have gradually been unable to meet this trend, prompting multi-band optical networks to gradually become the mainstream direction. To further improve the transmission efficiency and capacity of optical networks, it is necessary to break through the limitations brought by the unbalanced transmission performance between channels. The generalized signal-to-noise ratio (GSNR), as an important indicator for measuring the transmission performance of optical channels, the balance of its distribution directly affects the upper limit of the overall capacity of the optical network system. Therefore, achieving accurate perception and dynamic optimization of GSNR in multi-band optical networks has become the key to improving transmission capacity and system stability.

[0023] The main content of the present invention focuses on the dynamic perception and optimization control problem of the generalized signal-to-noise ratio (GSNR) distribution in a multi-band optical network, that is, by adjusting the pump configuration parameters of the Raman amplifier, the GSNR balance of multi-band signals during fiber transmission is achieved, thereby effectively improving the overall transmission capacity and stability of the system. In a multi-band fiber optic transmission system, affected by nonlinear effects, stimulated Raman scattering (SRS), amplified spontaneous emission noise (ASE), etc., the transmission performance of different band channels shows significant differences, resulting in uneven GSNR distribution, which becomes an important bottleneck restricting the optimization of system performance. As an important means to achieve multi-band transmission gain compensation, the Raman amplifier has high flexibility and programmability, and its pump power and wavelength configuration determine the system gain profile. However, due to the dynamic changes in signal conditions, traditional optimization methods based on fixed configuration or static modeling are difficult to quickly respond to the changes in the actual transmission environment, and there are problems such as poor adaptability and unstable optimization effects. Therefore, the present invention provides a method for optimizing the transmission capacity of a dynamic multi-band optical network based on GSNR perception. By constructing a high-precision perception model, the characteristics of the input signal are perceived in real time and the change of the output GSNR is predicted. Combining with an intelligent optimization algorithm, the pump configuration parameters are dynamically adjusted to achieve the optimal control of the transmission performance. The present invention effectively improves the optimization efficiency and adaptability of the multi-band optical network in a dynamic transmission scenario, and the relevant technical processes are as Figure 1 shown, and the specific steps include S100 to S400.

[0024] Step S100: Construct a data set for GSNR perception modeling; the data set is used to train an artificial neural network model to achieve high-precision prediction of the transmission performance of a dynamic multi-band optical network channel.

[0025] In the specific implementation process, the construction of the data set includes the following steps: Build a multi-band optical network simulation transmission platform to simulate different dynamic transmission scenarios; randomly set the GSNR distribution of the input signal and the pump power configuration of the Raman amplifier, and successively collect the corresponding output GSNR distribution curves to form a training data set and a test data set containing input GSNR characteristics, pump configuration parameters, and the corresponding output GSNR distribution.

[0026] In the specific implementation process, the multi-band optical network transmission scenarios built by the simulation platform specifically include, but are not limited to, the following scenarios, such as: multiplexing transmission scenarios with different numbers of bands, transmission scenarios with different fiber link lengths, and scenarios with different input signal power levels.

[0027] In the specific implementation process of the data set construction step, in order to reduce the model training complexity and improve the generalization ability of the model, the multi-band GSNR data is processed by the methods of channel grouping and feature dimensionality reduction, and a typical GSNR distribution feature set is extracted to retain the key transmission characteristics of the data.

[0028] In this embodiment, the simulation link is a hybrid amplification link structure covering the C+L band. In this link structure, a pump deployment scenario composed of Raman amplifiers is set. The pump configuration mode of the Raman amplifier is the backward pumping mode, and the length of the optical fiber link connected to the Raman amplifier is 80 km. To fully cover the typical characteristics of dynamic multi-band signal transmission, the signal covers 80 channels, the channel interval is 150 GHz, the signal baud rate is 128 GBaud, and the transmitted power spectrum is a flat spectrum.

[0029] In this embodiment, the dynamic transmission scenario is constructed by setting 1 to 10 unequal segments of EDFA (erbium-doped fiber amplifier) as amplification units to simulate the optical signal amplification process under various actual transmission conditions. The setting of the simulation parameters includes that the length of the optical fiber link is 40 km to 120 km, the interval is 40 km, the optical fiber attenuation coefficient is 0.2 dB / km, the dispersion coefficient is 16.7 ps / (nm·km), and the nonlinear coefficient is 1.265 (nm·km). -1 The dataset collection is based on the open-source optical network simulation platform GNPy. Through the above simulation configuration, a total of 25,100 groups of data samples are generated, 80% of which are used for the training of the neural network model, and 20% are used for model verification.

[0030] In this embodiment, during the data dimensionality reduction process, in order to reduce the dimension of the input features to meet the modeling requirements of the artificial neural network, the GSNR features of 80 channels are divided into groups of 5 adjacent channels, so as to obtain 16 groups of GSNR dimensionality reduction features, which constitute the input of the neural network model.

[0031] Step S200: Construct a GSNR perception model and perform model training; the GSNR perception model is implemented by an artificial neural network (ANN), aiming to establish a non-linear mapping relationship between the dynamic input GSNR distribution, Raman pump configuration parameters and output GSNR distribution.

[0032] In the specific implementation process, the artificial neural network model structure includes an input layer, multiple hidden layers and an output layer. The input layer mainly receives the GSNR feature data generated in step S100, including information such as the GSNR mean value, standard deviation of the dynamic signal and Raman pump configuration parameters. The number and node scale of the hidden layers will be optimized and adjusted according to the prediction error and generalization ability during the training process. Each hidden layer processes the input signal through an activation function (such as ReLU or Sigmoid) to help the model capture the complex non-linear relationship in the data. The output layer generates the output GSNR distribution predicted by the model, aiming to optimize the performance of the optical network through this distribution.

[0033] In this embodiment, the training process uses the standard backpropagation (BP) algorithm to iteratively optimize the model parameters. In each training step, the error value is calculated according to the loss function, and the loss function is the mean square error between the predicted value and the true value. The weights and biases in the model are adjusted through the gradient descent algorithm. During the training process, the optimization goal is to minimize the value of the loss function, so that the prediction ability of the model becomes more and more accurate, achieving a better fitting effect. To further improve the model's performance, the optimizer selects the Adam optimizer, which combines the advantages of the gradient descent method and automatically adjusts the learning rate of each parameter, thereby improving the training efficiency and accuracy.

[0034] In this embodiment, the error evaluation criterion uses the root mean square error (RMSE), but other error metrics such as the mean absolute error (MAE) or the coefficient of determination (R 2 ) can also be used for model performance evaluation. MAE is suitable for datasets with relatively uniform biases, while R 2 is used to measure the model fitting effect. According to specific application requirements, technicians can select suitable error metric criteria to optimize the prediction accuracy of the model.

[0035] In the specific implementation process, to avoid the model overfitting and improve the generalization ability of the model, a regularization strategy and a Dropout strategy are also introduced during the training process. The regularization strategy constrains the weights of the model to avoid overfitting on the training dataset, thereby improving the adaptability of the model to unknown data. The Dropout strategy further prevents the overfitting problem by randomly discarding some neurons during training and effectively enhances the robustness of the model. To ensure the training stability and prediction accuracy of the model, hyperparameter optimization is also performed during the training process. Parameters such as the learning rate and the optimizer type are determined through cross-validation and grid search.

[0036] Step S300: Optimize the dynamic pump configuration based on the optimization algorithm; the optimization algorithm adopts the particle swarm optimization algorithm (PSO), and each particle in the particle swarm corresponds to a set of pump power configurations. The fitness function is defined as the optimization objective function of the GSNR distribution.

[0037] The optimization process includes the following steps: First, evaluate the configuration of the current particle through the GSNR perception model, calculate the fitness value according to the model feedback, and evaluate the configuration effect; then, dynamically adjust the position and velocity of the particle in each iteration to guide it towards the global optimal solution; finally, evaluate the optimal particle of the current population according to the optimization objective function, and output the pump configuration corresponding to the particle with the highest fitness value as the optimal solution. Set the optimization threshold and termination condition during the iteration process. When the fitness value reaches the preset threshold or the set number of iterations is completed, stop the optimization and output the current optimal solution.

[0038] The optimized objective function is specifically as follows:

[0039]

[0040] where i and j are the indices of the channels and Raman amplifier pumps considered respectively. Gin,i is the GSNR of the input signal, and Pj is the pump power of the Raman amplifier. and are the average value and standard deviation of the output GSNR respectively, with the unit of dB.

[0041] Step S400: Apply the optimal pump power configuration parameters obtained by the particle swarm optimization algorithm to the actual dynamic transmission scenario of the multi-band optical network.

[0042] In the specific implementation process, by real-time monitoring the dynamic GSNR distribution characteristics of the input signal, using the GSNR perception model to predict the output GSNR distribution under the current input conditions in real time, and combining the feedback results of the optimization algorithm, dynamically adjust the pump configuration parameters of the Raman amplifier to achieve real-time optimization control of the GSNR between channels. In the actual transmission verification process, compare the optimized pump configuration obtained by this method with the traditional fixed configuration method. The experimental results show that the method of the present invention can significantly improve the average value and distribution flatness of the GSNR of the dynamic multi-band optical network transmission system, thereby effectively improving the overall transmission capacity and stability of the system.

[0043] In this embodiment, 100 groups of representative dynamic input signals are randomly generated. The dynamic input signals cover multiple channels within the C+L band range and include different signal power levels and fiber link length parameters to simulate the typical characteristics of the actual multi-band optical network under different transmission scenarios. Through the above signals, the optimization methods based on the GSNR perception model and the traditional static model of the present invention are respectively used to predict the output GSNR distribution, and the corresponding root mean square error (RMSE) is calculated to evaluate the prediction performance of the two models under dynamic conditions. The experimental results show that under the conditions of these 100 groups of dynamic input signals, the predicted RMSE of the GSNR perception model of the present invention is 0.58 dB, while the predicted RMSE of the traditional static model is 6.50 dB.

[0044] In an optimization case of a dynamic transmission scenario, based on the constructed multi-band optical network simulation platform, 200 groups of different dynamic input signals are randomly generated. Each group of input signals includes the multi-channel GSNR characteristics covering the C+L band range, which is used to comprehensively evaluate the adaptability and effect of the optimization method of the present invention in variable scenarios. For each group of input signals, a comparative experiment is respectively carried out by using the optimization method based on the GSNR perception model of the present invention and the traditional static configuration method.

[0045] During the experiment, for each input signal, first, the output GSNR distribution is predicted according to the GSNR perception model, and the corresponding optimal pump power configuration is determined through the particle swarm optimization algorithm (PSO). The optimized configuration is applied to the simulation transmission platform to obtain the actual output GSNR distribution, which is compared and verified with the prediction result of the perception model to ensure the accuracy of the model prediction. At the same time, the pump configuration is carried out by the traditional method, and the output GSNR distribution is recorded.

[0046] To quantify the performance difference between the two methods in terms of GSNR optimization, statistical analysis is respectively performed on the mean value and standard deviation of the output GSNR distribution. The average GSNR improvement amount and flatness improvement rate are used as evaluation indicators. In this embodiment, the GSNR flatness is characterized by the difference between the maximum value and the minimum value of GSNR. The experimental results show that the optimization method based on the GSNR perception model shows better optimization effects than the traditional method in all 200 groups of input signals.

[0047] Specifically, in all test scenarios, the average value of the output GSNR of the optimization method of the present invention has been significantly improved, and the average improvement amplitude is 1.08 dB. At the same time, the flatness of the GSNR between channels has also been significantly improved, and the average flatness improvement amplitude corresponding to the reduction of the standard deviation reaches 40.02%. Figure 3 shows the prediction effect of the GSNR perception model in the dynamic scenario, which has higher fitting accuracy and generalization ability than the traditional model; Figure 4 shows the comparison of the optimization results corresponding to 200 groups of input signals, clearly showing the advantages of the method of the present invention in improving the GSNR performance.

[0048] Combined with the above experimental results, it can be seen that the dynamic multi-band optical network transmission capacity optimization method based on GSNR perception proposed by the present invention can effectively meet the adaptive adjustment requirements of pump configuration under complex dynamic input conditions, not only significantly improving the transmission performance of the system, but also having good generalization ability and adaptability, and is applicable to various actual optical network deployment scenarios.

[0049] Those skilled in the art can implement the various exemplary components and methods described in combination with the embodiments disclosed in the present invention in the form of hardware, software, or a combination of both, and can also adjust and optimize the technical solutions according to specific application requirements. However, these adjustments and optimizations should not be considered to exceed the protection scope of the present invention. The technical features defined in the claims of the present invention can be applied alone or in any combination.

[0050] It should be noted that the term "comprising" or "including" used in the specification does not exclude the existence of other unenumerated elements or steps, and its scope of coverage should have the same meaning as "including but not limited to". Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

[0051] The above are only the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. For those skilled in the art, various changes and modifications can be made to the embodiments of the present invention. Any modification, equivalent replacement, and improvement made without departing from the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception, characterized in that: By taking the GSNR distribution of the input signal as the input dimension of the data-driven prediction model and combining the pump power configuration parameters of the Raman amplifier, the signal transmission performance in the dynamic transmission scenario is optimized, thereby improving the transmission capacity of the entire optical network. The method includes the following steps: (1) Construct a data set containing the GSNR distribution of dynamic signals and the pump configuration parameters of the Raman amplifier, extract the GSNR characteristics of the specified channel, and generate the feature data set required to build the prediction model; (2) Establish a GSNR perception prediction model, which outputs the predicted GSNR distribution based on the dynamic signal GSNR distribution and pump power configuration parameters in the input feature data, and provides real-time feedback for the optimization process through this model; (3) Based on the real-time feedback of the GSNR perception prediction model, an intelligent optimization algorithm is used to perform a global search in the multi-dimensional pump power parameter space to generate the optimal configuration scheme for the Raman pump in the current transmission scenario; (4) The optimized pump configuration is applied to actual transmission scenarios, and the GSNR balance between channels is improved through closed-loop feedback control, thereby optimizing the transmission capacity.

2. According to a method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, it is characterized in that: The designated channel refers to the target channel that needs to be optimized first in a dynamic transmission scenario, ensuring that the optimization process focuses on the key channel set that has the greatest impact on the overall capacity in the current transmission scenario.

3. The method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, characterized in that: The characteristic data set includes multi-band signal GSNR distribution data, Raman amplifier pump power configuration vector and corresponding output GSNR distribution data in a dynamic transmission scenario, which is obtained from an actual transmission scenario or a simulated transmission scenario.

4. The method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, characterized in that: The GSNR perception prediction model is constructed based on a data-driven modeling method. The input dimension expands the GSNR distribution information of the signal to enhance the model's perception and prediction capability of generalized signal-to-noise ratio changes in multi-band optical networks, thereby ensuring the optimization effect of transmission capacity in dynamic scenarios.

5. The method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, characterized in that: The GSNR perception prediction model constructs a mapping relationship between the GSNR distribution of the system input and output signals and the pump configuration parameters, providing a high-precision prediction basis for subsequent optimization.

6. The method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, characterized in that: In the step of calculating the loss function based on the predicted output GSNR curve and the true value of the GSNR perception prediction model, the loss function adopts mean square error, root mean square error, mean absolute error or maximum error.

7. The method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, characterized in that: The intelligent optimization algorithm dynamically updates the search strategy in the parameter space, combines the real-time feedback of the GSNR perception model, and uses the objective function as an evaluation criterion to achieve the optimal configuration of the pump power that improves the transmission capacity.

8. The intelligent optimization algorithm according to claim 1, characterized in that: The optimization objective function is defined as a combination of the average value of the GSNR between channels and the flatness, where the flatness is expressed in the form of the GSNR standard deviation, specifically in the form of: where i and j are the indices of the channel under consideration and the pump of the Raman amplifier, respectively. Gin,i is the GSNR of the input signal and Pj is the pump power of the Raman amplifier. and They are the mean and standard deviation of the output GSNR, in dB.

9. The method for optimizing transmission capacity of a dynamic multi-band optical network based on GSNR perception according to claim 1, characterized in that: The method constructs a closed-loop feedback control system to monitor the GSNR distribution changes of the input signal in dynamic transmission scenarios in real time, and periodically triggers the optimization algorithm to update the pump configuration parameters, ensuring that the multi-band optical network maintains high transmission capacity and stability during long-term operation.

10. The closed-loop feedback control system according to claim 9, characterized in that: The system dynamically combines the prediction results of the GSNR perception model with the output of the intelligent optimization algorithm through iterative optimization and real-time configuration adjustment to form an adaptive transmission capacity optimization closed loop.

Images