A method for estimating fiber output power spectrum of C+L optical transmission system

Abstract

The application belongs to the field of optical evaluation in optical transmission, and particularly relates to a method for optical fiber output power spectrum estimation of a C+L optical transmission system, which comprises the following steps: S1, inputting a full-wave low power into a fiber, obtaining an input end channel optical power spectrum and an output end channel optical power spectrum, and calculating an optical fiber inherent loss; and S2, inputting a full-wave low power into a fiber, obtaining an input end channel optical power spectrum and an output end channel optical power spectrum, and fitting model parameters. The single-machine model precision of the application has higher precision in a C+L waveband optical transmission system compared with a C waveband M. Zirngibl Raman model; the model parameter scaling and calibration of the application are compatible with the original C waveband M. Zirngibl Raman model; the application is applied to C+L waveband transmission quality model evaluation, is a light layer basic model of a digital twin network, reduces single-machine error, finally reduces model iteration times, and improves operation and maintenance efficiency.

Description

Technical Field

[0001] This invention belongs to the field of optical evaluation in optical transmission, specifically a method for estimating the output power spectrum of optical fibers in C+L optical transmission systems. Background Technology

[0002] Optical system transmission quality (QoT) assessment combines various impairment models of the optical system to achieve a comprehensive measurement of the optical performance and reliability of services. On the optical path (i.e., the working path), the performance of optical network topology nodes (such as channel optical power, service code type, fiber type and length, optical amplifier type, gain setting, etc.) is collected by detection equipment. Combined with optical layer models such as optical amplifiers, optical fibers, and optical modules, it is possible to perceive and optimize the performance of any node in the optical network (such as OSNR, BER, etc.). However, optical performance monitors (OPMs) are generally deployed at the beginning and end of the multiplexing section, and the power of intermediate topology nodes is unknown. On the other hand, on the optical path (i.e., the backup path / recovery path), due to the limited performance data that can be collected, the prediction of OSNR on the backup path model is more difficult than that of the working path. If the underlying model can achieve high-precision OSNR prediction, it can guide the operation and maintenance tools to configure reasonable preset attenuation for each optical service, thereby improving the single recovery success rate of switching to the optical path when the optical path recovers.

[0003] Both the working path channel power / OSNR estimation and the backup path power / OSNR prediction rely heavily on the accuracy of the underlying model. Currently, there are three mainstream solutions in the industry: the first is the analytical model approach (essentially a white-box model), which is simple and computationally fast, but its application is limited by assumptions. For example, the C-band fiber model uses the model proposed by M. Zirngibl et al., which assumes that the Raman gain coefficient spectrum of the fiber is a linear function of the frequency interval, thus limiting its application in wider spectrum operations. The second is AI-based models (essentially black-box models), which rely heavily on a large amount of calibration data to ensure model accuracy, resulting in higher deployment costs and lower practicality. The third is gray-box models, which combine traditional analytical (white-box) and machine learning models (black-box), offering high accuracy and high generalization. A modular modeling approach is adopted for fiber optic links, modeling each unit separately, such as the optical transceiver and fiber, with each unit composed of a gray-box model. Furthermore, to meet the speed requirements of network management, the complexity of each unit will be further reduced. Therefore, in the coupling of various units, it is necessary to further combine the waveform simulation model (to calculate the signal waveform) and the performance calculation model (to calculate the signal-to-noise ratio) to achieve the optimal trade-off between accuracy and complexity.

[0004] As backbone networks accelerate to 400G, wavelength division multiplexing (WDM) systems are gradually upgrading from the C-band 4THz spectrum range to C4.8T and C6T. 400G QPSK ×80 wave applications occupy a spectral bandwidth of 12THz (C+L). Stimulated Raman scattering (SRS) is very strong in C+L systems, causing a power shift from short wavelengths to long wavelengths in the WDM signal transmitted through the fiber. Therefore, a higher-precision fiber model is needed to improve the end-to-end estimation capability of the QoT model in various scenarios. Summary of the Invention

[0005] This invention proposes a method for estimating the output power spectrum of optical fibers in C+L optical transmission systems, comprising the following steps:

[0006] S1. Full-wave low-power fiber input: Obtain the optical power spectrum of the input channel and the optical power spectrum of the output channel, and calculate the inherent loss of the optical fiber.

[0007] S2. Full-wave low-power fiber input, obtain the optical power spectrum of the input channel and the optical power spectrum of the output channel, and fit the model parameters.

[0008] Furthermore, in step S1, a full-wavelength division multiplexing signal with a power lower than the standard input power is input into the optical fiber to obtain the optical power spectrum of the input channel. Output channel optical power spectrum Calculate the inherent loss of optical fiber .

[0009] Furthermore, in step S2, a full-wavelength division multiplexing signal with a power higher than the standard input power is input into the optical fiber to obtain the optical power spectrum of the input channel. Output channel optical power spectrum After obtaining the above calibration data, the optical fiber model was fitted with parameters, and the model is as follows:

[0010] (1);

[0011] in For wavelength, The total fiber input power under calibration conditions is... Effective length (where α is the fiber attenuation coefficient). The center frequency within the band can be used as the basis for fitting the Raman gain slope by substituting the data from the calibration process. and ;

[0012] After model calibration is completed, under arbitrary channel input power combinations Then we can perform integration on equation (1), that is:

[0013] (2);

[0014] The Raman net transfer spectrum can be obtained. By combining this with the inherent loss of the optical fiber, the output power spectrum of the optical fiber can be calculated as follows: .

[0015] The beneficial effects of this invention are:

[0016] The single-machine model of this invention has higher accuracy in C+L band optical transmission systems compared to the C-band M. Zirngibl Raman model; the model parameter calibration and adjustment of this invention are compatible with the original C-band M. Zirngibl Raman model; this invention is applied to the evaluation of C+L band transmission quality models and is the basic optical layer model of digital twin networks. Reducing single-machine errors will ultimately reduce the number of model iterations and improve operation and maintenance efficiency. Attached Figure Description

[0017] Figure 1 A flowchart of a method for estimating the output power spectrum of optical fibers in C+L optical transmission systems;

[0018] Figure 2 This is a schematic diagram illustrating the application of a method for estimating the output power spectrum of optical fibers in a C+L optical transmission system.

[0019] Figure 3 This is a performance verification diagram of a single fiber optic model unit in Example 1;

[0020] Figure 4 This is a performance verification diagram of the end-to-end model of the C+L band 11THz spectral width in Example 2. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] The following is in conjunction with the appendix Figure 1-4 The present invention will be further described as follows:

[0023] like Figure 2As shown, to illustrate the application scenario of this technical solution, a simplified schematic diagram of the working path and backup path of a C+L band optical transmission system is provided. The working path represents the path from optical cross-connect station 1 through optical fiber, optical line amplifier unit 1, optical fiber, optical cross-connect station 2, optical fiber, optical line amplifier unit 2, optical fiber, and optical cross-connect station 3 in sequence; the backup path represents the path from optical cross-connect station 1 through optical fiber, optical cross-connect station 4, optical fiber, and optical cross-connect station 3 in sequence. Regardless of whether it is the working path or the backup path, in order to accurately calculate the channel-level optical power of any network node, after knowing the channel-level power / preset value within the optical cross-connect station, it is necessary to cascade the fiber model and optical amplifier model from front to back. To avoid error accumulation, the accuracy of the single-machine model is particularly important. This technical solution aims to improve the estimation accuracy of the power spectrum (channel optical power) of the fiber output after the wavelength division multiplexed signal is transmitted within the optical fiber.

[0024] This technical solution requires model calibration of the optical fiber model. The specific steps for model calibration are as follows: Figure 1 As shown:

[0025] A method for estimating the output power spectrum of optical fibers in C+L optical transmission systems includes the following steps:

[0026] S1. First, input a full-wavelength wavelength division multiplexing (WDM) signal with a power lower than the standard input power into the optical fiber. The single-wavelength power is a low value, for example, 30 dB lower than the standard input power. Obtain the optical power spectrum of the input channel. Output channel optical power spectrum Calculate the inherent loss of optical fiber .

[0027] S2. Input a full-wavelength division multiplexing (WDM) signal with a power higher than the standard input power into the optical fiber. The single-wavelength power is relatively high; for example, it can be 5 dB higher than the standard input power. Obtain the optical power spectra of the input and output channels as follows: , .

[0028] After obtaining the above calibration data, the optical fiber model proposed in this invention is subjected to parameter fitting, and the model is as follows:

[0029] (1)

[0030] The aforementioned model is the core of this invention and the point to be protected. It originates from a revision of the model previously proposed by M. Zirngibl, and is derived by fitting a quadratic term to the Raman gain spectrum based on that model. For wavelength, The total fiber input power under calibration conditions is... Effective length (where α is the fiber attenuation coefficient). This is the center frequency within the band. Substituting the data from the calibration process, the Raman gain slope can be obtained by fitting the data. , .

[0031] After model calibration is completed, under arbitrary channel input power combinations Then we can perform integration on equation (1), that is:

[0032] (2)

[0033] The Raman net transfer spectrum can be obtained. By combining this with the inherent loss of the optical fiber, the output power spectrum of the optical fiber can be calculated as follows: .

[0034] During the operation and maintenance phase: When applying the fiber model proposed in this invention end-to-end in an optical transmission system, whether for channel power estimation in the working path or the backup path, the above calibration process can be used in the commissioning phase to save fiber parameters to the optical transmission equipment. However, in the operation and maintenance phase, for the C+L band working path, only the optical power of the first and last channels of the multiplex section can be obtained, and for the C+L band backup path, only the optical power of the first and last channels of the filler wave multiplex section can be obtained. Therefore, maintaining the model accuracy after commissioning depends on end-to-end model calibration, i.e., obtaining the channel-level error by subtracting the channel power after model cascading from the channel power reported at the end of the multiplex section, and reducing this error by adjusting the model parameters. In this invention, this requires adjustment. , This ensures the long-term accuracy of the model.

[0035] Example 1:

[0036] like Figure 3 As shown, wavelength division multiplexing (WDM) signals with a C-band spectral width of 6 THz and an L-band spectral width of 5 THz are selected, with a grid configuration of 100 GHz. First, the fiber model needs to be calibrated. Its single-wavelength power is set to -30 dBm, and it is passed through a 75 km long G.654 fiber to be calibrated. The power spectra of the input and output fibers are obtained using OPM to obtain the inherent span loss of the fiber. Further, its single-wavelength power is set to +4 dBm, and it is passed through the fiber to be calibrated again. The power spectra of the input and output fibers are again obtained using OPM. Substituting these spectra into the model allows for the fitting of the Raman slope coefficient of the fiber in this model. , Furthermore, to verify the accuracy of this model, the input power was set to single-wavelength +1 / 2 / 3 / 4 dBm and configured with 100 waves within the C+L 11 THz spectral width. The error between the fiber output power spectrum estimated by this invention and the measured output power spectrum can be effectively controlled within 0.3 dB.

[0037] Example 2:

[0038] like Figure 4 As shown, the C6T+L5T band is selected: C6T wavelength range 1524.7nm~1571.9nm, L5T wavelength range 1575.6nm~1617.2nm, with a frequency spacing of 100GHz, totaling 110 wavelengths. The optical transmission system contains 2 multiplexing sections, 5+4 (9 spans in total), and uses G.654 optical fiber. Firstly, in the initial setup phase, the fiber Raman slope coefficient of each fiber span is determined using the method described in this invention. The data is stored in the device network management system. After a period of initial operation and maintenance, the output power spectrum of the multiplex section can be estimated through the fiber optic model and optical amplifier model in the end-to-end cascaded system. The model error is obtained by subtracting the channel-level power reported by the OPM at the end of the multiplex section. By adjusting the fiber Raman gain slope and other model parameters, the end-to-end model error is reduced. Finally, zero error in the channel-level power estimation at the end of the multiplex section is achieved under this system. Furthermore, by combining the amplifier gain / noise spectrum, the end-to-end OSNR estimation accuracy can be achieved to <1.5dB.

[0039] In summary, this invention is not limited to the specific embodiments described above. Those skilled in the art can make various modifications and alterations without departing from the spirit and scope of this invention. The scope of protection of this invention should be determined by the claims of this invention.

Claims

1. A method for estimating the output power spectrum of optical fibers in C+L optical transmission systems, characterized in that, Includes the following steps: S1. Full-wave low-power fiber input: Obtain the optical power spectrum of the input channel and the optical power spectrum of the output channel, and calculate the inherent loss of the optical fiber. S2. Full-wave low-power fiber input, obtain the optical power spectrum of the input channel and the optical power spectrum of the output channel, and fit the model parameters.

2. The method for estimating the output power spectrum of optical fibers in a C+L optical transmission system according to claim 1, characterized in that, In step S1, a full-wavelength division multiplexing signal with a power lower than the standard input power is input into the optical fiber to obtain the optical power spectrum of the input channel. Output channel optical power spectrum Calculate the inherent loss of optical fiber .

3. The method for estimating the output power spectrum of optical fibers in a C+L optical transmission system according to claim 2, characterized in that, In step S2, a full-wavelength division multiplexing signal with a power higher than the standard input power is input into the optical fiber to obtain the optical power spectrum of the input channel. Output channel optical power spectrum After obtaining the above calibration data, the optical fiber model was fitted with parameters, and the model is as follows: (1)； in For wavelength, The total fiber input power under calibration conditions is... Effective length (where α is the fiber attenuation coefficient). The center frequency within the band can be used as the basis for fitting the Raman gain slope by substituting the data from the calibration process. and ; After model calibration is completed, under arbitrary channel input power combinations Then we can perform integration on equation (1), that is: (2)； The Raman net transfer spectrum can be obtained. By combining this with the inherent loss of the optical fiber, the output power spectrum of the optical fiber can be calculated as follows: .

Images