A raman gain adaptive regulation method and system for a multi-band optical network

Abstract

The application provides a Raman gain adaptive regulation method and system for a multi-band optical network, and the steps of the method comprise: supplementing gain values of non-target channels, constructing a gain vector based on target gain values of target channels and the supplemented gain values of the non-target channels, and performing at least one round of iteration; in each iteration, inputting values in a population into a pre-trained inverse model, the inverse model outputting corresponding pump adjustment parameters; determining gain values of all channels based on the pump adjustment parameters, screening target gain values of target channels from the gain values of all channels, and calculating fitness based on target gain values of target channels and the screened target gain values of the target channels; adjusting the supplemented gain values of non-target channels based on target function values, re-constructing the gain vector, and performing the next round of iteration, and after completing the last iteration, screening final pump adjustment parameters from all pump adjustment parameters.

Description

Technical Field

[0001] This invention relates to the field of optical amplifier technology, and in particular to a Raman gain adaptive control method and system for multi-band optical networks. Background Technology

[0002] With the development of new technologies such as 5G / 6G mobile communication and the continuous evolution of next-generation applications, high-capacity access traffic, such as machine-to-machine communication, is constantly increasing. Data traffic in transmission networks is growing rapidly, and optical networks, as fundamental transmission infrastructure, face enormous traffic demands, urgently requiring further capacity expansion. Introducing advanced modulation formats and constellation shaping is an economical and effective solution for expanding single-channel capacity, but single-channel capacity is ultimately limited by the nonlinear Shannon limit, insufficient to support network expansion needs. Multi-band optical networks that introduce new bands for transmission are a promising solution. They can utilize the entire communication window of standard single-mode fiber (SSMF) to expand the capacity of deployed optical networks and have proven to be the preferred short- and medium-term solution for optical network expansion.

[0003] In multi-band optical networks, different signal channels are subject to non-uniform effects from the interaction of various effects, including Kerr nonlinearity, amplified spontaneous emission noise, and stimulated Raman scattering. In such systems, the power curves of signal channels may exhibit arbitrary shapes. To ensure transmission signal quality and optimize information rates across multiple bands, it is necessary to achieve ultrafast gain profile reconfiguration.

[0004] Fiber Raman amplifiers are optical amplifiers based on nonlinear optical effects. Compared with other types of optical amplifiers, they have a very low noise figure, ensuring high-quality signal transmission. More importantly, fiber Raman amplifiers allow for flexible gain profile design by adjusting pump power and wavelength, and provide multi-band gain availability when operating in multi-pump configurations. This makes them ideal for achieving arbitrary gain profiles in multi-band optical networks in a controllable manner.

[0005] However, in commercial Raman pump modules, the pump wavelength and number are pre-set, and only the pump power can be adjusted. Such limited pump adjustment parameters cannot support the Raman amplifier in achieving precise, arbitrary gain across all wavelength channels in the entire amplification band. In actual optical transmission, not all channels are used for service transmission; typically, only some wavelength channels of a link are used. Existing technology can only adjust pump parameters for the gain of all channels, failing to guarantee specificity for the wavelength channels of a given link. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a Raman gain adaptive control method and system for multi-band optical networks, in order to eliminate or improve one or more defects existing in the prior art.

[0007] One aspect of the present invention provides a Raman gain adaptive modulation method for multi-band optical networks, the method comprising the following steps:

[0008] Obtain the gain value of the target channel, supplement the gain values ​​of the non-target channels, construct a gain vector based on the gain value of the target channel and the supplemented gain values ​​of the non-target channels, and perform at least one round of iteration;

[0009] In each iteration, the values ​​in the gain vector are input into a pre-trained inverse model, which outputs the corresponding pump adjustment parameters.

[0010] The gain values ​​of all channels are determined based on the pump adjustment parameters. The gain value of the target channel is selected from the gain values ​​of all channels. The objective function value is calculated based on the target gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels.

[0011] The gain value of the non-target channel supplement is adjusted based on the objective function value, and the gain vector is reconstructed for the next iteration. After the last iteration, the final pump adjustment parameters are selected from all the pump adjustment parameters.

[0012] The above scheme first randomly supplements the gain values ​​of non-target channels, obtains the corresponding pump adjustment parameters through an inverse model, and then determines the gain values ​​of all channels based on the pump adjustment parameters until the fitness requirement is met. Since the pump adjustment parameters in this scheme are related to the gain values ​​of all channels, this scheme obtains the corresponding pump adjustment parameters by supplementation. However, the obtained pump adjustment parameters may not correspond to the gain values ​​of the target channels. Therefore, this scheme further calculates the objective function value based on the target gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels, processes the target channel, and determines the final pump adjustment parameters through multiple rounds of iteration to ensure the targeting of the specified wavelength channel of the link.

[0013] In some embodiments of the present invention, in the step of constructing a gain vector based on the gain value of the target channel and the supplementary gain value of the non-target channel, a random value is selected in a preset gain interval or the gain value of the non-target channel is supplemented based on the target gain value of two adjacent target channels of the non-target channel to obtain the supplementary gain value of the non-target channel.

[0014] In the step of supplementing the gain value of a non-target channel based on the target gain values ​​of two adjacent target channels, a line graph is constructed based on the target gain values ​​of all channels and the target channels. The horizontal axis of the line graph represents the channel number, and the vertical axis represents the gain value. The points corresponding to the target gain values ​​of two adjacent target channels of the non-target channel are connected by a line. The vertical axis height corresponding to the position of the line passing through the non-target channel is used as the gain value supplemented for the non-target channel.

[0015] In some embodiments of the present invention, in the step of determining the gain value of all channels based on the pump adjustment parameters, the pump adjustment parameters are input into a preset forward model, and the forward model outputs the gain value of all channels.

[0016] In some embodiments of the present invention, both the forward model and the reverse model employ a feedforward neural network, a random forest, or a convolutional neural network structure.

[0017] In the specific implementation process, the forward model and the inverse model of this solution can have multiple structures, including but not limited to feedforward neural network, random forest or convolutional neural network structure or combination structure of feedforward neural network, random forest or convolutional neural network structure.

[0018] In some embodiments of the present invention, in the step of calculating the objective function value based on the gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels:

[0019] The root mean square error is calculated based on the gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels.

[0020] The objective function value is calculated based on the root mean square error.

[0021] In some embodiments of the present invention, in the step of calculating the objective function value based on the root mean square error, if a genetic algorithm is used, the objective function value is the fitness value. In this step, the objective function value is calculated based on the following formula:

[0022]

[0023] Where S is the calculated objective function value. For the calculated root mean square error, C known Indicates the target channel.

[0024] In some embodiments of the present invention, in the step of selecting the final pump adjustment parameters from all pump adjustment parameters after the last iteration is completed, a first screening strategy or a second screening strategy is adopted. The first screening strategy sets a preset number of iterations, and after all iterations are completed, a set of pump adjustment parameters is selected as the final pump adjustment parameters. The second screening strategy determines the final pump adjustment parameters during the iteration process based on a preset screening threshold.

[0025] In some embodiments of the present invention, in the step of selecting one set of pump adjustment parameters as the final pump adjustment parameters, the pump adjustment parameter corresponding to the highest fitness value is selected as the final pump adjustment parameter.

[0026] In some embodiments of the present invention, in the step of determining the final pump adjustment parameters during the iteration process based on a preset screening threshold, after each iteration is completed, the calculated fitness value is compared with the preset fitness threshold. If the calculated fitness value is greater than the preset fitness threshold, the iteration of that round is taken as the last iteration, and the pump adjustment parameters of that round are output as the final pump adjustment parameters.

[0027] A second aspect of the present invention also provides a Raman gain adaptive modulation system for multi-band optical networks. The system includes a computer device, which includes a processor and a memory. The memory stores computer instructions, and the processor executes the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps of the method described above.

[0028] A third aspect of the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned Raman gain adaptive modulation method for multi-band optical networks.

[0029] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the text, or may be learned by practice of the invention. The objects and other advantages of the invention will become apparent from the description and the accompanying drawings.

[0030] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0031] The accompanying drawings, which are provided to further illustrate the invention and form part of this application, are not intended to limit the scope of the invention.

[0032] Figure 1 This is a schematic diagram of one implementation of the Raman gain adaptive modulation method for multi-band optical networks in this scheme;

[0033] Figure 2 This is a schematic diagram of the architecture of this solution;

[0034] Figure 3 This is a schematic diagram illustrating the system application of this solution;

[0035] Figure 4 This is one configuration curve for the experimental example;

[0036] Figure 5 This is another configuration curve for the experimental example. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0038] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0039] like Figure 1 , 2 As shown in Figure 3, this invention proposes a Raman gain adaptive control method and system for multi-band optical networks. The steps of this method include:

[0040] Step S100: Obtain the gain value of the target channel, supplement the gain value of the non-target channel, construct a gain vector based on the gain value of the target channel and the supplemented gain value of the non-target channel, and perform at least one round of iteration.

[0041] In the specific implementation process, by setting rules to fill the target gain of non-specified wavelength channels, which are channels without specific gain requirements, a target curve space containing a series of target gain values ​​passing through the specified channels is obtained.

[0042] Step S200: In each iteration, the value in the gain vector is input into the pre-trained inverse model, and the inverse model outputs the corresponding pump adjustment parameters.

[0043] Step S300: Determine the gain values ​​of all channels based on the pump adjustment parameters, select the gain value of the target channel from the gain values ​​of all channels, and calculate the target function value based on the target gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels.

[0044] In some embodiments of this scheme, in the step of determining the gain values ​​of all channels based on the pump adjustment parameters, a gain estimation model for a multi-band programmable Raman amplifier is constructed to provide real-time feedback. The evaluation of the quality of curves in the target curve space is based on this model. The modeling process can be implemented based on numerical modeling, simulation modeling, and machine learning pre-training methods, either individually or in combination, to simulate the pump prediction and amplification process of the programmable Raman amplifier. The desired effect should be to accurately simulate the pump prediction and amplification process of the programmable Raman amplifier; that is, with the same input gain curve, the gain curve generated by the real programmable Raman amplifier should be as consistent as possible with the gain curve generated by the simulation process. After all curves in the target curve space are processed by the above-mentioned gain estimation model of the programmable Raman amplifier, a generated curve space consisting of several one-to-one corresponding generated curves is obtained. The difference between the configured curve and its one-to-one corresponding generated curve will be used as a measure of the gain generation effect of the programmable Raman amplifier.

[0045] Specifically, based on the feedback provided by the constructed gain estimation model, Raman gain control optimization is performed in the target curve space according to the input target parameters. This involves filtering and optimizing curves in the target curve space that contain the target gain value for a specified channel until the curve with the optimal gain generation effect is found. The found curve is then used as the output of the optimization method. When this curve is used as the input to a programmable Raman amplifier, it achieves the best gain generation accuracy in the specified wavelength channel.

[0046] The optimal curve found through Raman gain control optimization is used as the input target gain curve of the programmable Raman amplifier. Based on this curve, the programmable Raman amplifier completes the gain control process after pump prediction and command issuance.

[0047] The propagation equation of the Raman amplifier needs to consider the distribution of forward and backward pump, signal, and ASE noise power. The nonlinear differential equation satisfied by its propagation is as follows:

[0048]

[0049] Among them, P + (z,v i ) and P - (z,v i ) Frequency v iNearby forward and reverse transmitted optical power; α, η, h, k, T are the fiber attenuation coefficient, Rayleigh scattering coefficient, Planck constant, Boltzmann constant, and absolute temperature, respectively; A eff It is the optical fiber at frequency v m Effective area; g R Γ is the Raman gain coefficient; Γ is the polarization factor.

[0050] In step S400, the gain value of the non-target channel supplement is adjusted based on the objective function value, and the gain vector is reconstructed for the next iteration. After the last iteration is completed, the final pump adjustment parameters are selected from all the pump adjustment parameters.

[0051] In the specific implementation process, this scheme adopts a genetic algorithm, in which an initial population is constructed based on the gain value of the target channel and the gain value of the non-target channel. In each round of iteration, the fitness is calculated and the population is updated to complete the iteration. Through continuous selection, crossover and mutation, an optimal individual is found iteratively and the individual is used as the output of the algorithm.

[0052] In some embodiments of this invention, selection, crossover, and mutation are performed cyclically, continuously evaluating individuals in the population until the iteration termination condition is met or the set number of iterations is completed. First, a selection operation is performed to select superior individuals from the current population, providing a basis for subsequent crossover and mutation operations. The selection process is typically based on the fitness of individuals; individuals with higher fitness have a higher probability of being selected. In this invention, fitness is calculated based on the reciprocal of RMSE; therefore, the selection process will favor individuals whose corresponding generated gain curves have smaller errors to the target gain on the specified wavelength channels. The crossover operation is performed on the gain curves of selected parent individuals, exchanging the gain values ​​of some wavelength channels to generate new gain curves, thereby introducing new genetic diversity while preserving the superior traits of the parent individuals. Finally, a mutation operation is used to make small-scale adjustments to the individual's gain curve, i.e., fine-tuning the gain value of specific wavelength channels to prevent the algorithm from converging prematurely to a local optimum and to maintain population diversity. It is worth noting that crossover and mutation operations are only performed on unspecified wavelength channels; for specified wavelength channels, the target gain value is always kept unchanged. Set the initial number of iterations, the iteration optimization threshold, and the iteration termination threshold. Once an individual's fitness exceeds the iteration optimization threshold, accelerate the iteration appropriately; when an individual's fitness exceeds the termination threshold, stop the iteration and output that individual as the result. Otherwise, continue iterating until the initial number of iterations is reached, and output the best individual found during the iteration process as the final result.

[0053] In practical implementation, this solution adaptively optimizes the target gain of the programmable Raman amplifier on some specified transmission channels to achieve high gain generation accuracy on specified wavelength channels, thereby making full use of limited pump resources and maximizing the gain generation accuracy of the multi-band programmable Raman amplifier on specific wavelength channels. This is used to overcome the problem of limited gain generation accuracy across all wavelength channels caused by the limited adjustable parameters of the programmable Raman amplifier, and promote the application and effective deployment of programmable Raman amplifiers.

[0054] This solution is applicable to various optical transmission scenarios, including but not limited to multi-band fixed grid optical network transmission, multi-band flexible grid optical network transmission, and multi-band ultra-long distance transmission systems. These are transmission scenarios that have an urgent need for adaptive gain control of multi-band optical network amplifiers due to their large transmission bandwidth and uneven transmission spectrum.

[0055] The above scheme first randomly supplements the gain values ​​of non-target channels, obtains the corresponding pump adjustment parameters through an inverse model, and then determines the gain values ​​of all channels based on the pump adjustment parameters until the fitness requirement is met. Since the pump adjustment parameters in this scheme are related to the gain values ​​of all channels, this scheme obtains the corresponding pump adjustment parameters by supplementation. However, the obtained pump adjustment parameters may not correspond to the gain values ​​of the target channels. Therefore, this scheme further calculates the objective function value based on the target gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels, processes the target channel, and determines the final pump adjustment parameters through multiple rounds of iteration to ensure the targeting of the specified wavelength channel of the link.

[0056] In some embodiments of this invention, the target gain value of a specified wavelength channel is kept constant. Gain values ​​of non-specified wavelength channels are filled according to a specified rule. Several gain curves are obtained through different filling values, forming an initial population and completing population initialization. The specified filling rule is to find the nearest channels with gain requirements on both sides of the specified channel. The value of the channel falling on the line connecting the two points is used as the reference gain, and the filling value will randomly oscillate within a set range above and below the reference gain. Each individual in the initial population is a complete gain curve, composed of the target gain value of the specified wavelength channel and the filling gain value of the non-specified wavelength channels. This process involves the creation of the initial population, which determines the initial diversity of the algorithm's solution space. In this step, we set the number of initial populations to be negatively correlated with the number of channels with gain requirements. This means that when there are many channels with gain requirements, the target gain can fluctuate in fewer locations, and the number of initial populations can be relatively small, thus reducing the complexity of the algorithm's search. Conversely, when there are few channels with gain requirements, the target gain curve's gain value can fluctuate across many wavelength channels, and the number of initial populations is relatively large, thus providing the algorithm with more candidate solutions to explore the optimal gain curve.

[0057] In some embodiments of the present invention, in the step of constructing a gain vector based on the gain value of the target channel and the supplementary gain value of the non-target channel, a random value is selected in a preset gain interval or the gain value of the non-target channel is supplemented based on the target gain value of two adjacent target channels of the non-target channel to obtain the supplementary gain value of the non-target channel.

[0058] In the step of supplementing the gain value of a non-target channel based on the target gain values ​​of two adjacent target channels, a line graph is constructed based on the target gain values ​​of all channels and the target channels. The horizontal axis of the line graph represents the channel number, and the vertical axis represents the gain value. The points corresponding to the target gain values ​​of two adjacent target channels of the non-target channel are connected by a line. The vertical axis height corresponding to the position of the line passing through the non-target channel is used as the gain value supplemented for the non-target channel.

[0059] In some embodiments of the present invention, in the step of determining the gain value of all channels based on the pump adjustment parameters, the pump adjustment parameters are input into a preset forward model, and the forward model outputs the gain value of all channels.

[0060] In some embodiments of the present invention, both the forward model and the reverse model employ a feedforward neural network, a random forest, or a convolutional neural network structure.

[0061] In the specific implementation process, a multi-band programmable Raman amplifier gain estimation model is constructed to provide real-time feedback. This step is crucial for subsequent control optimization. The modeling process can be implemented based on numerical modeling, simulation modeling, and machine learning pre-training, either individually or in combination, to simulate the pump prediction and amplification of the programmable Raman amplifier. In this embodiment, the scheme uses machine learning for modeling. By pre-training the machine learning model, a model capable of accurately predicting the pump and accurately simulating the amplification process of the programmable Raman amplifier is obtained, thus acquiring the generation curves of the programmable Raman amplifier corresponding to different target gain curves. When simulating this process based on the trained machine learning model, a large amount of data needs to be collected. This data includes the parameter settings of the programmable Raman amplifier under different pump configuration parameters and the corresponding Raman gain curves. This data can be generated through experimental measurements, simulation software, or by solving the propagation equation of the Raman amplifier. The quality and diversity of data acquisition are also crucial to the accuracy of the model. Simultaneously, a sufficiently suitable machine learning method needs to be selected to build the model, considering factors such as method complexity, training time, prediction accuracy, and generalization ability. In this embodiment, a feedforward neural network is selected to build the model.

[0062] In some embodiments of the present invention, in the step of calculating the objective function value based on the gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels:

[0063] The root mean square error is calculated based on the gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels.

[0064] The objective function value is calculated based on the root mean square error.

[0065] In the specific implementation process, this scheme uses genetic algorithm, reinforcement learning algorithm or particle swarm algorithm for iteration. Specifically, if a genetic algorithm is used, the objective function value is the fitness value and the gain vector is the population; if a reinforcement learning algorithm is used, the objective function value is the reward function value; if a particle swarm algorithm is used, the objective function value is the fitness value and the gain vector is the particle swarm.

[0066] In some embodiments of the present invention, if a genetic algorithm is used, the objective function value is the fitness value. In the step of calculating the objective function value based on the root mean square error, the objective function value is calculated based on the following formula:

[0067]

[0068] Where S is the calculated objective function value, RMSE Cknown For the calculated root mean square error, C knownIndicates the target channel.

[0069] In some embodiments of the present invention, in the step of selecting the final pump adjustment parameters from all pump adjustment parameters after the last iteration is completed, a first screening strategy or a second screening strategy is adopted. The first screening strategy sets a preset number of iterations, and after all iterations are completed, a set of pump adjustment parameters is selected as the final pump adjustment parameters. The second screening strategy determines the final pump adjustment parameters during the iteration process based on a preset screening threshold.

[0070] In some embodiments of the present invention, in the step of selecting one set of pump adjustment parameters as the final pump adjustment parameters, the pump adjustment parameter corresponding to the highest fitness value is selected as the final pump adjustment parameter.

[0071] In some embodiments of the present invention, the iteration is set to 1000 times, resulting in 1000 configuration curves. The curve with the best configuration effect is selected as a comparison of the control method proposed in this invention. The configuration curves obtained by the two methods are as follows: Figure 4 As shown. The two generated curves obtained after the two configuration curves are simulated by a programmable Raman amplifier are as follows. Figure 5 As shown in the figure. The results show that, compared with the optimal configuration result in 1000 random configurations, the gain value of the generated curve obtained by the control method proposed in this invention is closer to the target gain value in a specified portion of the channels, achieving higher gain generation accuracy.

[0072] In some embodiments of the present invention, in the step of determining the final pump adjustment parameters during the iteration process based on a preset screening threshold, after each iteration is completed, the calculated fitness value is compared with the preset fitness threshold. If the calculated fitness value is greater than the preset fitness threshold, the iteration of that round is taken as the last iteration, and the pump adjustment parameters of that round are output as the final pump adjustment parameters.

[0073] This solution breaks through the current application limitations of programmable Raman amplifiers, enabling them to meet the new requirements of multi-band optical networks for optical amplifiers. Specifically, by ensuring the gain generation accuracy on a specified wavelength channel, the programmable Raman amplifier can effectively play its role with limited pump resources, thus mitigating the signal transmission quality degradation caused by the introduction of new wavelengths and ensuring the capacity expansion effect of multi-band optical networks.

[0074] This solution achieves high gain generation accuracy in a specified wavelength channel by finding an optimal input curve for the programmable Raman amplifier, overcoming the application limitations of programmable Raman amplifiers caused by the limited number of pump adjustment parameters. Simultaneously, this method provides real-time feedback on the gain generation effect to the user, enabling appropriate adjustments to the target gain based on the generation results, thus ensuring stable operation of the amplification process.

[0075] The beneficial effects of this plan include:

[0076] 1. This solution enables dynamic optimization and high gain generation accuracy: By adaptively adjusting to the actual transmission state of the network in real time, it achieves high-precision gain configuration for specific wavelength channels with limited pump resources, solving the problem of limited gain generation accuracy for all wavelength channels due to the limited number of pump adjustment parameters, and promoting the widespread application of programmable Raman amplifiers.

[0077] 2. This solution provides real-time feedback: It can provide users with real-time feedback on the generation effect of the gain, enabling them to make appropriate adjustments to the target gain based on the generation effect, thus ensuring the stable operation of the amplification process.

[0078] This invention also provides a Raman gain adaptive control system for multi-band optical networks. The system includes a computer device, which includes a processor and a memory. The memory stores computer instructions, and the processor executes the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps of the method described above.

[0079] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the aforementioned Raman gain adaptive modulation method for multi-band optical networks. The computer-readable storage medium can be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, floppy disk, hard disk, removable storage disk, CD-ROM, or any other form of storage medium known in the art.

[0080] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.

[0081] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0082] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0083] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations of the embodiments of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A Raman gain adaptive modulation method for multi-band optical networks, characterized in that, The steps of this method include: Obtain the gain value of the target channel, supplement the gain value of the non-target channel, construct a gain vector based on the gain value of the target channel and the supplemented gain value of the non-target channel, randomly select a value in a preset gain interval or supplement the gain value of the non-target channel based on the target gain value of two adjacent target channels, obtain the supplemented gain value of the non-target channel, and perform at least one round of iteration. In each iteration, the values ​​in the gain vector are input into a pre-trained inverse model, which outputs the corresponding pump adjustment parameters. The gain values ​​of all channels are determined based on the pump adjustment parameters. The gain value of the target channel is selected from the gain values ​​of all channels. The objective function value is calculated based on the target gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels. The gain value of the non-target channel supplement is adjusted based on the objective function value, and the gain vector is reconstructed for the next iteration. After the last iteration, the final pump adjustment parameters are selected from all the pump adjustment parameters.

2. The Raman gain adaptive control method for multi-band optical networks according to claim 1, characterized in that, In the step of supplementing the gain value of a non-target channel based on the target gain values ​​of two adjacent target channels, a line graph is constructed based on the target gain values ​​of all channels and the target channels. The horizontal axis of the line graph represents the channel number, and the vertical axis represents the gain value. The points corresponding to the target gain values ​​of two adjacent target channels of the non-target channel are connected by a line. The vertical axis height corresponding to the position of the line passing through the non-target channel is used as the gain value supplemented for the non-target channel.

3. The Raman gain adaptive control method for multi-band optical networks according to claim 1, characterized in that, In the step of determining the gain value of all channels based on the pump adjustment parameters, the pump adjustment parameters are input into a preset forward model, and the forward model outputs the gain value of all channels.

4. The Raman gain adaptive control method for multi-band optical networks according to claim 3, characterized in that, Both the forward and reverse models employ the structures of feedforward neural networks, random forests, or convolutional neural networks.

5. The Raman gain adaptive control method for multi-band optical networks according to claim 1, characterized in that, In the step of calculating the objective function value based on the gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels: The root mean square error is calculated based on the gain value of the target channel and the gain value of the target channel selected from the gain values ​​of all channels. The objective function value is calculated based on the root mean square error.

6. The Raman gain adaptive control method for multi-band optical networks according to claim 5, characterized in that, If a genetic algorithm is used, the objective function value is the fitness value. In the step of calculating the objective function value based on the root mean square error, the objective function value is calculated based on the following formula: Where S is the calculated objective function value. The calculated root mean square error, Indicates the target channel.

7. The Raman gain adaptive control method for multi-band optical networks according to any one of claims 1 to 6, characterized in that, In the step of selecting the final pump control parameters from all pump control parameters after the last iteration, a first screening strategy or a second screening strategy is adopted. The first screening strategy sets a preset number of iterations. After all iterations are completed, one set of pump control parameters is selected as the final pump control parameters. The second screening strategy determines the final pump adjustment parameters during the iteration process based on a preset screening threshold.

8. The Raman gain adaptive control method for multi-band optical networks according to claim 7, characterized in that, In the step of selecting one set of pump control parameters as the final pump control parameters, if a genetic algorithm is used, the objective function value is the fitness value, and the pump control parameter corresponding to the highest fitness value is taken as the final pump control parameter.

9. The Raman gain adaptive control method for multi-band optical networks according to claim 7, characterized in that, If a genetic algorithm is used, the objective function value is the fitness value. In the step of determining the final pump adjustment parameters during the iteration process based on a preset screening threshold, after each iteration, the calculated fitness value is compared with the preset fitness threshold. If the calculated fitness value is greater than the preset fitness threshold, then the iteration of that round is taken as the last iteration, and the pump adjustment parameters of that round are output as the final pump adjustment parameters.

10. A Raman gain adaptive control system for multi-band optical networks, characterized in that, The system includes a computer device, which includes a processor and a memory. The memory stores computer instructions, and the processor executes the computer instructions stored in the memory. When the computer instructions are executed by the processor, the system implements the steps of the method as described in any one of claims 1 to 9.

Images