Power pre-equalization method and device for optical transmission system
By optimizing the channel of the WDM optical communication system through optical power adjustment and covariance adaptive evolution strategy, the problem of large signal quality differences was solved, and the uniformity of the receiver signal and the improvement of system performance were achieved.
Patent Information
- Application Number
- CN202511084195.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-04
- Publication Date
- 2025-11-11
AI Technical Summary
In existing WDM optical communication systems, there are significant differences in signal quality during signal transmission between channels, leading to excessively high bit error rates at the receiving end.
By adjusting the optical power of each channel of the wavelength division multiplexing optical signal, multiple sets of optical power distribution combinations are generated. The optimal optical power weight is calculated by using a covariance adaptive evolution strategy for iterative optimization, thus achieving pre-equalization.
The overall performance of the receiver has been optimized, resulting in uniform signal quality and improved overall system transmission performance.
Smart Images
Figure CN120934675A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of optical communication systems, and specifically to a power pre-equalization method and apparatus for an optical transmission system. Background Technology
[0002] Optical communication networks, with their advantages of high capacity, low loss, and low latency, have become one of the main technologies in current transmission networks. Furthermore, the successive adoption of wavelength division multiplexing (WDM), erbium-doped fiber amplifiers (EDFA), and higher-order modulation techniques has significantly improved the transmission capacity and relay distance of optical communication networks. The continuously improving transmission capabilities and enhanced advantages of optical communication networks have led to their widespread application in high-speed, long-distance communication scenarios such as backbone networks, metropolitan area networks, and transoceanic communications. However, while the research and application of new technologies have promoted the development of optical communication technology, they have also introduced new problems. For example, dense wavelength division multiplexing (DWDM) technology can improve communication capacity, but it also increases the incident light power in the optical fiber, leading to enhanced nonlinear effects when the signal light is transmitted in the optical fiber, which in turn leads to enhanced nonlinear noise introduced during signal transmission. EDFA does not require optical-to-electrical-to-optical conversion and directly amplifies the signal light in the optical domain, but it does not perform 3R regeneration of the signal, resulting in the inability to optimize signal noise and waveform distortion. In addition, spontaneous emission (ASE) noise is introduced during EDFA amplification, and this type of noise accumulates continuously as the optical signal is transmitted through the optical fiber. High-order modulation technology can increase the number of information bits carried by each symbol, but its received signal has a high optical signal-to-noise ratio (OSNR).
[0003] Furthermore, for wavelength division multiplexing (WDM) optical communication network systems, complex interaction processes exist between signals. These include four-wave mixing (FWM), cross-phase modulation (XPM), channel-specific self-phase modulation (SPM), gain competition between channels due to EDFA amplification gain unevenness, and channel power migration caused by Raman effects in the fiber. The combined effect of these factors results in uneven optical power and inconsistent performance degradation across different channels at the receiver. In some cases, certain channels may still have significant performance margins compared to their thresholds, while others may be severely degraded and unusable.
[0004] For WDM optical transmission systems, when the signal light leveled at the transmitting end is transmitted over a long distance through optical fiber to the end of the system, there are significant differences in the signal quality of different channels. In extreme cases, the difference in signal quality may cause some signal light to have an excessively high bit error rate at the receiving end of the system. Summary of the Invention
[0005] This application provides a power pre-equalization method and apparatus for an optical transmission system to optimize the performance of the received signal at the receiving end.
[0006] In a first aspect, embodiments of this application provide a power pre-equalization method for an optical transmission system, the method comprising: Based on randomly set optical power weights, the optical power of each channel of the wavelength division multiplexing optical signal is adjusted to generate multiple sets of optical power distribution combinations; Numerical modeling of the optical communication transmission system is performed, and transmission performance indicators are calculated based on the combination of multiple sets of optical power distributions to generate the original dataset; Based on the original dataset, a sample dataset is obtained by sampling. A covariance adaptive evolution strategy is used to iteratively optimize the sample dataset to obtain the optimal optical power weight for each channel. The optical power of each channel at the transmitting end is pre-equalized using the optimal optical power weight.
[0007] In conjunction with the first aspect, in one implementation method, the wavelength division multiplexing optical signal is generated as follows: Multiple optical signals are generated, each carrying a modulation format, rate, and bandwidth; the optical signals from the multiple channels are multiplexed into a wavelength division multiplexed optical signal using wavelength division multiplexing (WDM) technology.
[0008] In conjunction with the first aspect, in one implementation, numerical modeling of the optical communication transmission system is performed, and transmission performance indicators are calculated based on the combination of the multiple sets of optical power distributions to generate an original dataset, including: An optical communication performance evaluation algorithm is used to numerically model an optical communication transmission system. The optical power distribution combination is used as input to calculate the transmission performance index of the optical communication transmission system. The optical power distribution combination is associated with the corresponding transmission performance index to generate samples. The original dataset is generated based on multiple samples.
[0009] In conjunction with the first aspect, in one embodiment, the optical communication performance evaluation algorithm includes a Gaussian noise transmission performance evaluation algorithm; the transmission performance index is the generalized signal-to-noise ratio.
[0010] In conjunction with the first aspect, in one implementation, a sample dataset is obtained by sampling from the original dataset, including: The original dataset is treated as a population that follows a Gaussian distribution, and a sample dataset is obtained by sampling from it. The distribution parameters of the population include the mean, standard deviation, and covariance matrix.
[0011] In conjunction with the first aspect, in one implementation, a covariance adaptive evolution strategy is employed to iteratively optimize the sample dataset to obtain the optimal optical power weight for each channel, including: Calculate the fitness of the sample dataset and select an elite subset based on the fitness ranking; Based on the elite subset, the covariance matrix is updated by combining both rank min and rank 1 methods; The mean and standard deviation are updated based on the updated covariance matrix to generate a new population. Based on the new population sampling, a sample dataset is obtained. The sample dataset is iteratively optimized until the iteration reaches the preset termination condition, and the optimal optical power weight of each channel is obtained.
[0012] In conjunction with the first aspect, in one implementation, the fitness of the sample dataset is calculated, and an elite subset is selected based on the fitness ranking, including: Calculate and sort the fitness values of each sample in the sample dataset, and select the samples that account for k% of the population to form an elite subset. For the original population, the samples with the highest fitness ranking (1-k%) in the sample dataset are also included in the elite set.
[0013] In conjunction with the first aspect, in one implementation, obtaining the optimal optical power weight for each channel includes: acquiring the sample dataset at the end of the iteration, filtering out the elite subset, obtaining the optical power distribution combination corresponding to the parameter with the highest fitness value in the elite subset, and obtaining a set of optical power weights for each channel corresponding to the optical power distribution combination, as the optimal optical power weight.
[0014] In conjunction with the first aspect, in one implementation, after pre-equalizing the optical power of each channel at the transmitting end using optimal optical power weights, the method further includes: The combined pre-equalized optical power distribution is input into the model of the optical communication transmission system to calculate the corresponding output of the optical communication transmission system; at the same time, the output of the optical communication transmission system is calculated when the input optical power of all channels is 0dBm, and the two outputs are compared to evaluate the performance.
[0015] Secondly, embodiments of this application provide a power pre-equalization apparatus based on any one of the power pre-equalization methods for optical transmission systems, comprising: The pre-equalization module is used to adjust the optical power of each channel of the wavelength division multiplexing optical signal according to the randomly set optical power weights, and generate multiple sets of optical power distribution combinations; it is also used to pre-equalize the optical power of each channel at the transmitting end using the optimal optical power weights. The optical communication system evaluation model is obtained by numerical modeling of the optical communication transmission system. It is used to calculate the transmission performance index based on the combination of multiple sets of optical power distributions and generate the original dataset. The pre-equalization algorithm module is used to obtain a sample dataset based on the original dataset, and to iteratively optimize the sample dataset using a covariance adaptive evolution strategy to obtain the optimal optical power weight for each channel.
[0016] The beneficial effects of the technical solutions provided in this application include: This application performs numerical modeling of an optical communication transmission system. Based on multiple combinations of optical power distributions, it calculates transmission performance indicators to generate an original dataset. A sample dataset is obtained by sampling from the original dataset. A covariance adaptive evolution strategy is used to iteratively optimize the sample dataset to obtain the optimal optical power weight for each channel. The optimal optical power weight is then used to pre-equalize the optical power of each channel at the transmitting end. Using the transmission performance indicators of each channel at the receiving end as the optimization objective, numerical simulation is used to simulate the overall performance of the system output signal under various combinations of input optical power for different channels. Then, the covariance adaptive evolution strategy is used to calculate the optimal optical power weight for each channel. The optimal optical power weight is used to pre-equalize the signal power at the transmitting end, resulting in optimal overall performance at the receiving end and achieving overall performance tuning of the entire communication system. Furthermore, it supports the function of allowing users to formulate different optimization strategies according to their own needs. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating an embodiment of the power pre-equalization method for the optical transmission system of this application; Figure 2 A flowchart illustrating an embodiment of generating the original dataset for this application; Figure 3 This is a flowchart illustrating the process of iteratively optimizing the sample dataset using a covariance adaptive evolution strategy employed in this application. Figure 4 This is a schematic diagram of an embodiment of the power pre-equalization device for the optical transmission system of this application; Figure 5 A schematic diagram of the optimized optical power pre-configuration and flat power distribution of this application; Figure 6 This is a schematic diagram of the output GSNR under the optimized power distribution and flat power distribution configuration of this application. Detailed Implementation
[0018] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.
[0019] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.
[0020] In a first aspect, embodiments of this application provide a power pre-equalization method for an optical transmission system.
[0021] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the power pre-equalization method of this application. Figure 1 As shown, the power pre-equalization method includes the following steps: S1: Adjust the optical power of each channel of the wavelength division multiplexing optical signal according to the randomly set optical power weights to generate multiple sets of optical power distribution combinations.
[0022] S2: Perform numerical modeling of the optical communication transmission system, calculate transmission performance indicators based on multiple sets of optical power distribution combinations, and generate the original dataset.
[0023] S3: Based on the original dataset, a sample dataset is obtained by sampling. The Covariance Adaptive Evolutionary Strategy (CMA-ES) is used to iteratively optimize the sample dataset to obtain the optimal optical power weight for each channel.
[0024] S4: Use the optimal optical power weight to pre-equalize the optical power of each channel at the transmitting end.
[0025] In this embodiment, by numerically modeling the optical communication transmission system, the overall performance of the system output signal is simulated under various combinations of input optical power in different channels. Then, using the covariance adaptive evolution strategy, the optimal optical power weight of each channel is calculated. The optimal optical power weight is used to pre-equalize the signal power at the transmitting end in advance, so that the overall performance of the receiving end is optimal, and the overall performance of the entire communication system is optimized. This solves the technical problem of large differences in signal quality at the receiving end of WDM systems in the prior art.
[0026] Furthermore, in step S1 above, the wavelength division multiplexing optical signal is generated as follows: the WDM system transmitter generates optical signals of multiple channels, each channel optical signal carrying modulation format, rate and bandwidth, etc. The modulation format, rate and bandwidth can be the same or different; the optical signals of multiple channels are multiplexed into a wavelength division multiplexing optical signal through wavelength division multiplexing technology.
[0027] In step S1 above, weight parameters are set for each channel optical signal. When the optical power of all wavelengths is set with the same weight, the output optical power of the pre-equalization module is flat. The optical power weight of each channel is randomly set, and the optical power of each channel of the wavelength division multiplexing optical signal is adjusted to generate an optical power distribution combination. Based on multiple sets of randomly set optical power weights, multiple sets of optical power distribution combinations are obtained.
[0028] Further, the generation of the original dataset in step S2 above includes: pre-modeling the optical communication transmission system numerically using an optical communication performance evaluation algorithm to obtain an optical communication system evaluation model. The optical communication system evaluation model uses combinations of optical power distributions as input to calculate the transmission performance indicators of the optical communication transmission system. Each combination of optical power distributions input into the optical communication system evaluation model is associated with the corresponding output transmission performance indicator to generate a sample. The number of samples in the original dataset is pre-set, and the original dataset is composed of multiple samples.
[0029] An optical communication transmission system is the transmission carrier of output signals, realizing the transmission of signals from the transmitting end to the receiving end. In this embodiment, the optical communication transmission includes N transmission segments composed of optical fibers and EDFAs.
[0030] Furthermore, in one embodiment, the optical communication performance evaluation algorithm employs a Gaussian noise (GN) transmission performance evaluation algorithm, and the transmission performance metric is the generalized signal-to-noise ratio (GSNR). In other embodiments, the performance metric can also be the optical signal-to-noise ratio (OSNR).
[0031] Furthermore, in one embodiment, steps S1 and S2 can be integrated into one step. The optical communication transmission system can be numerically modeled beforehand using an optical communication performance evaluation algorithm to obtain an optical communication system evaluation model. For example... Figure 2 As shown, the original dataset can be generated by following these steps: a1: Acquire wavelength division multiplexed optical signals.
[0032] a2: Randomly set the optical power weight for each channel, adjust the wavelength division multiplexed optical signal, and generate an optical power distribution combination.
[0033] a3: Input the combined optical power distribution into the optical communication system evaluation model and output the calculated GSNR. Simultaneously with outputting the GSNR, the optical power spectrum of the combined optical power distribution can also be acquired.
[0034] a4: Combine the input optical power distribution and GSNR as samples and add them to the original dataset.
[0035] a5: Determine whether the number of samples in the original dataset has reached the preset initial number threshold. If yes, it means that the initial dataset has been generated and the process ends; otherwise, proceed to step a2.
[0036] Furthermore, in one embodiment, such as Figure 3 As shown, step S3 above specifically includes the following steps: S31: Treat the original dataset as a population that follows a Gaussian distribution, and sample it to obtain a sample dataset. The distribution parameters of this population include the mean, standard deviation, and covariance matrix.
[0037] S32: Calculate the fitness of the sample dataset and select an elite subset based on fitness ranking.
[0038] S33: Based on elite subsets, update the covariance matrix by combining rank min and rank 1 methods.
[0039] S34: Update the mean and standard deviation based on the updated covariance matrix to generate a new population.
[0040] S35: Obtain a sample dataset based on the new population sampling, calculate the fitness of the sample dataset, and select a new elite subset based on fitness ranking.
[0041] S36: Determine whether the preset number of iterations has been reached. If yes, proceed to S37; otherwise, proceed to S33.
[0042] S37: Obtain the optical power distribution combination corresponding to the parameter with the highest fitness value in the elite subset selected by S35, and obtain a set of optical power weights for each channel corresponding to the optical power distribution combination, which is used as the optimal optical power weight.
[0043] Furthermore, in one embodiment, in step S32 above, the fitness value of each sample in the sample dataset is calculated and sorted, and samples accounting for a proportion k of the population are selected to form an elite subset. The k of the original population can be preset. For the original population, samples with the highest fitness ranking (1-k) in the sample dataset are also included in the elite set to prevent the original population from having too few samples.
[0044] Based on the power pre-equalization method for the aforementioned optical transmission system, after pre-equalizing the optical power of each channel at the transmitting end using optimal optical power weights, a verification step can also be included. This verification step includes: The combined pre-equalized optical power distribution is input into the optical communication transmission system, and the corresponding output of the optical communication transmission system is calculated. At the same time, the output of the optical communication transmission system is calculated when the input optical power of all channels is 0dBm. The two outputs are compared to evaluate the performance and verify whether the optical signal quality is improved after using the optimal optical power weight for pre-equalization.
[0045] This application employs a covariance adaptive evolutionary strategy (CMA-ES) to calculate the optimal optical power weights for each channel and pre-equalize the optical power of each channel at the transmitting end, thereby achieving the highest overall performance at the receiving end. The covariance adaptive evolutionary strategy optimizes the results by adjusting the distribution parameters based on the relationship between the distribution parameters and the outcome. Evolution refers to finding the genotype with the highest fitness through natural selection.
[0046] The following section explains the principle of the covariance adaptive evolution strategy, which can be summarized as follows: Assume the target... For the function value to be optimized, It is its function parameter, because and The relationships between them cannot be represented by explicit relational expressions, therefore they cannot be solved using gradient descent, but for specific inputs The following can be calculated The function value. Therefore, assume probability distribution yes One solution, the goal of the evolutionary algorithm is to find ,make maximum.
[0047] Therefore, the solution to the evolutionary strategy is to find the optimal one. Furthermore, this application specifies the power of each channel. It can be viewed as an n-dimensional Gaussian distribution, where n is the number of channels in the WDM system. This represents the distribution parameters, including the mean in a Gaussian distribution. μ and standard deviation σ ,Right now: Formula 1; in, I Given an n-dimensional identity matrix, the evolutionary strategy steps are as follows: (1) To Perform initialization. = Set the counter to t=0 to record the number of iterations.
[0048] (2) From the population that follows a Gaussian distribution, a sample is taken from the population, and the sample size is... : Formula 2; in, t represents the number of iterations. Let represent the variance of the (t+1)th iteration, whose distribution follows a normal distribution; This represents the step size. The initial population is a sample dataset obtained by sampling the original dataset. The proportion of the elite subset to the population is pre-set, k0. To avoid getting trapped in local optima, the initial value of k0 is as small as possible, such as 5% of the population size.
[0049] (3) Select the one that makes optimal The elite subset is formed by selecting samples that constitute a proportion of the population k. If it is the original population, then k = k0. To avoid the elite subset of the original population having too few species, for the original population, it is usually... The samples with the highest fitness ranking (1-k) were also included in the elite set and labeled as : Formula 3; (4) By forming a new elite subset, the fitness of the population is calculated, and the maximum difference in fitness, FitnessDiff, is calculated to update the number of elite subsets to be selected in the next iteration. As the number of iterations increases, the number of samples with fitness approaching the optimal level increases, and the proportion of elites to be retained gradually increases. Because FitnessDiff gradually decreases as fitness approaches the optimal level, the proportion of k also gradually increases. Furthermore, the value of k has little correlation with the initial value. If the value of k0 is large, many samples with poor fitness will be selected into the elite set, that is, many samples deviate from the optimal fitness value. In this case, the calculated FitnessDiff value will be large, and the calculated k value will become smaller, thus reducing the number of elite samples selected in the next iteration, thereby achieving adaptive selection of the number of samples.
[0050] (5) Estimate the new mean and standard deviation of the next generation of samples using the existing elite sample set: Formula 4; Repeat (2) to (5) until the result is stable. The result is stable if it is less than the preset threshold.
[0051] In the evolutionary process described above, standard deviation σ The standard deviation determines the exploration degree of the evolutionary algorithm. If the standard deviation is too small, local optima may occur; if the standard deviation is too large, sampling of the offspring population can be performed in a larger search space, but slow convergence or failure to converge may occur. To address these issues, this application utilizes the covariance matrix to monitor the positive dependencies between samples in the distribution environment, thus avoiding excessively large or small standard deviations and finding a suitable standard deviation. The new distribution parameters become: Formula 5; At this point, the variance distribution follows Instead of the original normal distribution, we use covariance. In Equation 5, standard deviation... σ That is, the step size, and C is the covariance matrix. The distribution of the covariance adaptive evolution strategy transforms the identity matrix into the covariance matrix, thereby strengthening the dependency between samples.
[0052] The covariance matrix C is a symmetric matrix with the following desirable properties: C is always a diagonal matrix; C is always a positive semi-definite matrix; all eigenvalues are non-negative real numbers; all eigenvalues are orthogonal; and the eigenvectors of C can form an n-dimensional orthonormal basis.
[0053] Suppose matrix C has an eigenvector. The orthonormal basis formed has the following corresponding eigenvalues: ,make , Formula 6; The square root of C is: Formula 7; Leveraging the properties of the covariance matrix, the covariance adaptive evolutionary strategy can be applied to systems using elite samples. Estimating the covariance matrix from scratch : Formula 8; To ensure the accuracy of the above evaluation, the population size must be large enough. However, a large population inevitably consumes time in fitness calculations. Therefore, it is desirable to use a smaller sample size for rapid iteration to reduce optimization time. The covariance adaptive evolution strategy employs a more reliable method to update the covariance matrix C, which includes two independent evolutionary paths: rank min and rank 1. rank The update method is to use The history of a population is estimated from scratch in each generation. Assuming the use of average estimates, if historical experience across many generations is available, then: Alternatively, Polyak averaging can be used, and historical information can be utilized using the learning rate. Formula 9; The learning rate typically chosen is: .
[0054] The rank-1 update method is another update method, which mainly estimates the moving step size based on historical data. And updating symbol information, this method is to solve The problem of losing symbolic information can be addressed by using an evolutionary path to record symbolic information, similar to adjusting the step size. This path remains consistent before and after population updates. Conjugate of distribution: Formula 10; Formula 11; in and These are the learning rate for updating the path and the learning rate for updating the covariance, respectively. Rank In the previous method, estimation was performed from scratch in each generation, resulting in a comprehensive but time-consuming search process. The rank-1 update method utilizes the symbolic information of the moving step size and the correlation between consecutive steps, and this information can be passed down through generations as the population is updated, thereby improving update performance. This application combines the two update methods to improve the performance of the evolutionary strategy.
[0055] Formula 12 The above explains the principle of the covariance adaptive evolution strategy in this application. The updated covariance matrix is obtained through Equation 12. The distribution parameters are obtained by combining formula 5. θ , and then calculate .
[0056] Secondly, such as Figure 4 As shown, an embodiment of a power pre-equalization device for an optical transmission system is provided. In this embodiment, the power pre-equalization device includes a pre-equalization module, an optical communication system evaluation model, and a pre-equalization algorithm module.
[0057] The pre-equalization module adjusts the optical power of each channel of the wavelength division multiplexing (WDM) optical signal according to randomly set optical power weights, generating multiple combinations of optical power distributions. It also pre-equalizes the optical power of each channel at the transmitting end using optimal optical power weights. The WDM optical signal originates from the WDM transmit signal module. The pre-equalized optical signal, after passing through the optical communication system evaluation model, is transmitted to the WDM receive signal module.
[0058] The optical communication system evaluation model is obtained by numerical modeling of the optical communication transmission system. It is used to calculate transmission performance indicators based on multiple combinations of optical power distributions and generate the original dataset. Figure 4 In this embodiment, the optical communication system evaluation model obtains the optical power distribution combination from node A and the transmission performance index from node B. The transmission performance index is GSNR. Figure 4 (A1) shows the spectral diagram of node A; (B1) shows the spectral diagram of node B without pre-equalization. (A2) shows the spectral diagram of node A; (B2) shows the spectral diagram of node B after pre-equalization.
[0059] The pre-equalization algorithm module is used to obtain a sample dataset based on the original dataset, and to iteratively optimize the sample dataset using a covariance adaptive evolution strategy to obtain the optimal optical power weight for each channel.
[0060] The following is an embodiment involving a specific application scenario, illustrating the pre-equalization method for an optical transmission system in conjunction with a power pre-equalization device. This embodiment aims to improve the GSNR performance of the entire system. The optical communication transmission system is a submarine open optical cable communication system scenario, comprising 10 span systems, each span being 100km long, with an attenuation coefficient of 0.2dB / km, an fiber dispersion coefficient of 17ps / nm / km, a nonlinear coefficient of 1.3681 / W / m, a Raman slope of 0.028 / W / km / THz, a channel spacing of 50GHz, a baud rate of 28GHz, and 63 channels. All EDFAs in the optical communication transmission system operate in power-locked mode, with a constant output power of 20dBm. It is assumed that the input optical power of all channels at the transmitting end is 0dBm before optimization, and this configuration is used as a comparison after optimization to verify the effectiveness of the method in this embodiment.
[0061] First, for the pre-equalization module, the optical power weight of each channel is randomly assigned, or it can be set according to the priority of the service. The pre-equalization module adjusts the optical power of each channel of the wavelength division multiplexing (WDM) optical signal according to the optical power weight, generating an optical power distribution combination. This optical power distribution combination serves as the input to the optical communication system evaluation model, which calculates the GSNR of each channel. To more clearly distinguish the differences, the first... The GSNR of each channel is expressed as follows: In this embodiment, a fitness function is set based on the sum of the GSNR of all channels to calculate the fitness; at the same time, it can provide users with the option to set weights for each channel according to the priority level of the channel service and customize the fitness function.
[0062] In this embodiment, the average GSNR fitness function is set as follows: Formula 13; Assuming the 23rd channel has a higher priority than other channels, performance improvements for the 23rd channel will be prioritized during optimization. Therefore, J 23 =0.16667, and all other J values are 0.01344, where J represents the channel weight.
[0063] The pre-equalization algorithm module iterates continuously based on optical power and fitness function value using a covariance adaptive evolution strategy to find the optical power weight that maximizes the fitness function value. This optimal optical power weight is then used to set the pre-equalization module. The pre-equalization module adjusts the optical power of each channel of the wavelength division multiplexing optical signal to generate the optimal combination of optical power distributions.
[0064] The optimal optical power distribution combination is input into the optical communication system evaluation model to calculate the system's output performance. To verify the effectiveness of this method, the system's output performance is also calculated under the condition of flat input power (0 dBm input optical power for all channels), and the output performance of the two cases is compared.
[0065] The optimized input optical power pre-configuration and flat power distribution are shown in the attached figure. Figure 5 As shown in the attached figure, under the optimized power distribution and flat power distribution configurations, the corresponding output GSNR is as follows. Figure 6 As shown. From the appendix Figure 6 As can be seen, the GSNR performance of the higher-priority channel, namely channel 23, improved from 14.87dB to 16.06dB after optimization, an improvement of 1.19dB. The average GSNR of the entire system improved from 14.396dB to 15.1957dB, an improvement of 0.7997dB. While optimizing the overall system GSNR, the performance of important channels was prioritized for improvement. Therefore, the power pre-equalization method in this application is proven to be effective.
[0066] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0067] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.
[0068] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.
[0069] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.
[0070] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.
[0071] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.
[0072] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.
Claims
1. A power pre-equalization method for an optical transmission system, characterized in that, The method includes: Based on randomly set optical power weights, the optical power of each channel of the wavelength division multiplexing optical signal is adjusted to generate multiple sets of optical power distribution combinations; Numerical modeling of the optical communication transmission system is performed, and transmission performance indicators are calculated based on the combination of multiple sets of optical power distributions to generate the original dataset; Based on the original dataset, a sample dataset is obtained by sampling. A covariance adaptive evolution strategy is used to iteratively optimize the sample dataset to obtain the optimal optical power weight for each channel. The optical power of each channel at the transmitting end is pre-equalized using the optimal optical power weight.
2. The power pre-equalization method for an optical transmission system as described in claim 1, characterized in that, The wavelength division multiplexing optical signal is generated as follows: Multiple optical signals are generated, each carrying a modulation format, rate, and bandwidth; the optical signals from the multiple channels are multiplexed into a wavelength division multiplexed optical signal using wavelength division multiplexing (WDM) technology.
3. The power pre-equalization method for an optical transmission system as described in claim 1, characterized in that, Numerical modeling of the optical communication transmission system is performed, and transmission performance indicators are calculated based on the combination of multiple sets of optical power distributions to generate an original dataset, including: An optical communication performance evaluation algorithm is used to numerically model an optical communication transmission system. The optical power distribution combination is used as input to calculate the transmission performance index of the optical communication transmission system. The optical power distribution combination is associated with the corresponding transmission performance index to generate samples. The original dataset is generated based on multiple samples.
4. The power pre-equalization method for an optical transmission system as described in claim 3, characterized in that: The optical communication performance evaluation algorithm includes a Gaussian noise transmission performance evaluation algorithm; the transmission performance index is the generalized signal-to-noise ratio.
5. The power pre-equalization method for an optical transmission system as described in claim 1, characterized in that, The sample dataset was obtained by sampling from the original dataset, including: The original dataset is treated as a population that follows a Gaussian distribution, and a sample dataset is obtained by sampling from it. The distribution parameters of the population include the mean, standard deviation, and covariance matrix.
6. The power pre-equalization method for an optical transmission system as described in claim 5, characterized in that, A covariance adaptive evolution strategy is employed to iteratively optimize the sample dataset, yielding the optimal optical power weights for each channel, including: Calculate the fitness of the sample dataset and select an elite subset based on the fitness ranking; Based on the elite subset, the covariance matrix is updated by combining both rank min and rank 1 methods; The mean and standard deviation are updated based on the updated covariance matrix to generate a new population. Based on the new population sampling, a sample dataset is obtained. The sample dataset is iteratively optimized until the iteration reaches the preset termination condition, and the optimal optical power weight of each channel is obtained.
7. The power pre-equalization method for an optical transmission system as described in claim 6, characterized in that, Calculate the fitness of the sample dataset, and select an elite subset based on fitness ranking, including: Calculate and sort the fitness values of each sample in the sample dataset, and select the samples that account for k% of the population to form an elite subset. For the original population, the samples with the highest fitness ranking (1-k%) in the sample dataset are also included in the elite set.
8. The power pre-equalization method for an optical transmission system as described in claim 6, characterized in that, To obtain the optimal optical power weight for each channel, the following steps are taken: obtaining the sample dataset at the end of the iteration, selecting the elite subset, obtaining the optical power distribution combination corresponding to the parameter with the highest fitness value in the elite subset, and obtaining a set of optical power weights for each channel corresponding to the optical power distribution combination, which are then used as the optimal optical power weights.
9. The power pre-equalization method for an optical transmission system as described in claim 1, characterized in that, After pre-equalizing the optical power of each channel at the transmitting end using the optimal optical power weight, the process also includes: The combined pre-equalized optical power distribution is input into the model of the optical communication transmission system to calculate the corresponding output of the optical communication transmission system; at the same time, the output of the optical communication transmission system is calculated when the input optical power of all channels is 0dBm, and the two outputs are compared to evaluate the performance.
10. A power pre-equalization apparatus based on the power pre-equalization method of the optical transmission system according to any one of claims 1-9, characterized in that, include: The pre-equalization module is used to adjust the optical power of each channel of the wavelength division multiplexing optical signal according to the randomly set optical power weights, and generate multiple sets of optical power distribution combinations; it is also used to pre-equalize the optical power of each channel at the transmitting end using the optimal optical power weights. The optical communication system evaluation model is obtained by numerical modeling of the optical communication transmission system. It is used to calculate the transmission performance index based on the combination of multiple sets of optical power distributions and generate the original dataset. The pre-equalization algorithm module is used to obtain a sample dataset based on the original dataset, and to iteratively optimize the sample dataset using a covariance adaptive evolution strategy to obtain the optimal optical power weight for each channel.