A fiber optic nonlinear compensation method and system for digital subcarrier multiplexing

Abstract

This invention discloses an optical fiber nonlinear compensation method and system based on tunable virtual zero-dispersion point digital subcarrier multiplexing, relating to the field of optical fiber communication technology. The method includes: acquiring a received DSCM signal and obtaining multiple subcarrier signals; constructing a virtual reverse transmission model with virtual zero-dispersion points based on transmission parameters; and performing nonlinear compensation on at least a portion of the subcarrier signals based on a set of perturbation coefficients characterizing the coupling relationship between nonlinear effects and dispersion effects within and / or between subcarriers, to obtain compensated subcarrier signals. The perturbation coefficient set can be a full set of coefficients or a simplified set of coefficients processed by at least one of the following: screening, real-valued, quantized, threshold-trimmed, or grouped accumulation. The method can also perform dispersion compensation and / or time delay alignment processing before, after, or before and after compensation. This invention can reduce computational load and storage overhead while maintaining compensation performance, and is suitable for high-speed, long-distance coherent optical transmission systems.

Description

Technical Field

[0001] This invention relates to the field of optical fiber communication technology, and more particularly to an optical fiber nonlinearity compensation method and system for digital subcarrier multiplexing. Background Technology

[0002] As coherent optical communication systems evolve towards higher baud rates, larger capacities, and longer transmission distances, Digital Subcarrier Multiplexing (DSCM) has emerged as a key candidate for next-generation high-speed optical transmission systems. This is because it can decompose a broadband signal into multiple narrower-bandwidth digital subcarriers, exhibiting significant advantages in dispersion resistance, nonlinearity suppression, and improving the rate-distance tradeoff. However, under long-distance and high-power conditions, fiber Kerr nonlinearity remains a critical bottleneck limiting DSCM system performance. Conventional dispersion compensation at the receiver is insufficient to effectively suppress the nonlinear distortion caused by intra-subcarrier self-phase modulation (SPM) and inter-subcarrier cross-phase modulation (XPM) coupled with dispersion; therefore, compensation is necessary. Existing compensation schemes mainly include: Traditional DBP (Digital Backpropagation) method: Based on the step-by-step Fourier method, it solves the optical fiber transmission process in reverse and can compensate for dispersion and nonlinearity at the same time. However, in order to consider dispersion broadening and walk-off effects, it usually requires a lot of steps, and the computational complexity is very high, which makes it difficult to meet the power consumption and parallelism requirements of high-speed coherent DSP (Digital Signal Processor) chips.

[0003] Improved DBP / single-step approximation method: By optimizing the low-pass filter, the number of steps can be reduced, which can reduce the complexity to a certain extent. However, it is still insufficient for the analytical characterization of the coupling between multiple subcarriers in DSCM.

[0004] Neural network-assisted nonlinear compensation methods can improve compensation performance, but require training data, offline optimization, and high hardware resources, resulting in high deployment costs.

[0005] Perturbation nonlinear compensation method: The coupling effect of dispersion and nonlinearity is characterized by first-order perturbation theory, which can achieve single-step compensation. However, for DSCM signals, if full-scale perturbation coefficients are used directly, a large number of perturbation coefficients and triple multiplication are required, which is too costly in terms of hardware.

[0006] In summary, existing compensation schemes suffer from high computational complexity, insufficient utilization of the nonlinear coupling mechanism within / between subcarriers in DSCM scenarios, difficulty in balancing performance and complexity, and existing coefficient quantization schemes mostly remain at the level of additional simplification of existing full-scale compensation without further exploring the real-value-dominated characteristics of DSCM perturbation coefficients, and without optimizing the relationship between the zero-dispersion point location, real-value coefficient ratio, correlation length, and FFT (Fast Fourier Transform) size in the virtual reverse fiber as a unified design object.

[0007] Therefore, the core technical problem to be solved by this invention is how to establish a single-step, analytical, and easily hardware-implementable nonlinear compensation method suitable for DSCM signals while maintaining low implementation complexity. Summary of the Invention

[0008] To address the issues of high computational complexity, large coefficient size, and high implementation complexity in existing technologies for nonlinear compensation of DSCM signals, this invention provides a nonlinear compensation method and system for digital subcarrier multiplexed optical fibers based on a virtual reverse transmission model. This method introduces an adjustable virtual zero-dispersion point and utilizes a set of perturbation coefficients characterizing the coupling relationship between intra- and inter-subcarrier nonlinear effects and dispersion effects to perform nonlinear compensation on DSCM signals. Simultaneously, by filtering, realizing, quantizing, thresholding, and / or grouping and accumulating the perturbation coefficient set, the computational load and storage overhead are reduced while maintaining compensation capability.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect of the present invention, a fiber nonlinearity compensation method for digital subcarrier multiplexing is provided, comprising the following steps: S1: Obtain the received DSCM signal, and obtain multiple subcarrier signals based on the received DSCM signal; S2: Based on the transmission distance, dispersion parameters, nonlinear parameters and subcarrier configuration of the DSCM signal, construct a virtual reverse transmission model with virtual zero dispersion points; S3: Based on the set of perturbation coefficients corresponding to the virtual reverse transmission model, perform nonlinear compensation on at least a portion of the subcarrier signals to obtain the compensated subcarrier signals; The perturbation coefficient set is used to characterize the coupling relationship between nonlinear effects and dispersion effects within and / or between subcarriers, and the perturbation coefficient set is either a full set of coefficients or a simplified set of coefficients processed by at least one of screening, realization, quantization, threshold pruning, and group accumulation.

[0010] Preferably, the received DSCM signal is a time-domain signal containing X-polarization and Y-polarization, and multiple subcarrier signals are obtained from the received DSCM signal through time-domain decomposition or frequency-domain decomposition.

[0011] Preferably, when using the frequency domain decomposition method, the received DSCM signal is subjected to at least one of the following processing methods: block division, Fourier transform, and filtering, spectrum shifting, spectrum truncation, and downsampling based on a preset subcarrier frequency configuration, in order to obtain multiple subcarrier frequency domain signals.

[0012] Preferably, in the subcarrier extraction process, a joint implementation mechanism of frequency domain subcarrier demultiplexing and chromatic dispersion compensation is adopted. By pre-setting the composite weight of the integrated subcarrier filtering window and the dispersion compensation phase factor, a frequency domain multiplication is performed on the frequency domain received signal. In a single operation, the isolation and separation of M subcarriers are achieved using a frequency domain window function, and the corresponding distance L is achieved through phase factor multiplication. The chromatic dispersion pre-compensation of z0 ultimately outputs a pure, independent M-channel frequency domain subcarrier signal that has completed preliminary linear dispersion correction.

[0013] Preferably, the dispersion compensation includes pre-dispersion compensation and post-dispersion compensation, which are located before and after nonlinear compensation, respectively. The dispersion compensation includes three aspects: compensation for intra-subcarrier dispersion, alignment of fractional delay caused by dispersion between subcarriers, and unified correction of phase offset caused by transmission path differences between subcarriers.

[0014] Preferably, the single-step nonlinear compensation is used to compensate for intra-subcarrier self-phase modulation impairment and / or inter-subcarrier cross-phase modulation impairment.

[0015] Preferably, the single-step nonlinear compensation adopts any one of the following: an additive compensation model, a phase rotation compensation model, or a compensation model combining additive and phase rotation.

[0016] Preferably, the simplification processing of the perturbation coefficient set includes at least one of the following: retaining only real-valued perturbation coefficients; deleting low-contribution perturbation coefficients according to a threshold; quantizing the perturbation coefficients; and grouping and accumulating the perturbation coefficients according to subcarrier spacing, symmetry relationship, or coefficient value.

[0017] Preferably, the set of perturbation coefficients is determined based on at least some of the parameters in the following parameters: transmission distance, virtual zero dispersion point location, dispersion parameter, nonlinear parameter, power distribution, pulse waveform, subcarrier spacing, and sampling relationship, and is obtained through analytical calculation, time-domain convolution operation, frequency-domain equivalent operation, or a combination thereof.

[0018] Preferably, the virtual zero dispersion point is adjustable, and its preferred value is half of the total transmission distance of the signal in the optical fiber, forming a symmetrical dispersion compensation structure, which is used to increase the power ratio of the real-valued perturbation coefficient and shorten the correlation length.

[0019] In a second aspect of the present invention, an optical fiber nonlinearity compensation system for digital subcarrier multiplexing is provided, comprising: Signal acquisition module: used to acquire the received DSCM signal and obtain multiple subcarrier signals based on the received DSCM signal; Signal compensation module: used to perform compensation processing on multiple subcarrier signals respectively to obtain compensated subcarrier signals.

[0020] In a third aspect of the invention, a computer-readable storage medium is provided having a program stored thereon, which, when executed by a processor, implements the digital subcarrier multiplexing fiber nonlinear compensation method.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention establishes a first-order perturbation analytical model for single-step compensation of DSCM signals, which can explicitly characterize the coupling relationship between intra-subcarrier SPM, inter-subcarrier XPM and dispersion, and avoids the high-complexity inverse transmission process of traditional multi-step DBP.

[0022] (2) By introducing virtual reverse fiber and adjustable zero dispersion point z0, the processing flow of “front-end dispersion compensation - intermediate nonlinear compensation - back-end dispersion compensation” can be organized under a unified framework, providing additional degrees of freedom for performance optimization and implementation optimization.

[0023] (3) By taking advantage of the fact that real-valued coefficients in the DSCM perturbation coefficients account for the main power contribution, only real-valued coefficients are retained to retain most of the effective nonlinear compensation capability, while significantly reducing the amount of multiplication operations.

[0024] (4) By further quantizing and grouping the coefficients, the number of coefficients and storage overhead can be significantly reduced. The number of real-value coefficients can be reduced from 5852 to 96, a reduction of about 98.3%.

[0025] (5) A symmetrical CDC structure is preferred, i.e., z0=L / 2. This structure is beneficial to increasing the power ratio of real coefficients and shortening the correlation length, thereby allowing for a smaller FFT size and further reducing the complexity of hardware implementation.

[0026] (6) In the simulation of 5-channel WDM and dual-polarization 64-GBaud DSCM, the RPB-NLC with 8 subcarriers and 12 quantization levels can achieve performance comparable to 3 step / span DBP on optical links of 1600 km and 2400 km, while the computational complexity is only about 3.5% and 2.4% of that of DBP, which has a good performance-complexity trade-off. Attached Figure Description

[0027] Figure 1 This is a flowchart of an optical fiber nonlinearity compensation method for digital subcarrier multiplexing according to the present invention; Figure 2 This is a flowchart of the steps in the first embodiment of the present invention; Figure 3 This is a distribution diagram of one type of perturbation nonlinear action coefficient in the first embodiment of the present invention; Figure 4 This is a detailed flowchart of the second embodiment of the present invention; Figure 5 Is Figure 3 Under the condition of nonlinear compensation, the distribution of perturbation nonlinear action coefficients is shown. Figure 6 This is a distribution diagram of the perturbation nonlinearity coefficients after processing using uniform amplitude quantization. Figure 7 This is the first simulation result; Figure 8 This is the second simulation result; Figure 9 This is the third simulation result. Detailed Implementation

[0028] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0029] Please see Figure 1 As shown, in a first aspect of the present invention, an optical fiber nonlinearity compensation method for digital subcarrier multiplexing is provided, comprising the following steps: S1: Obtain the received DSCM signal, and obtain multiple subcarrier signals based on the received DSCM signal; The received DSCM signal is a time-domain signal containing X-polarization and Y-polarization. Multiple subcarrier signals are obtained from the received DSCM signal through time-domain decomposition or frequency-domain decomposition.

[0030] After coherent reception and analog-to-digital sampling, the DSCM time-domain signal obtained at the receiving end contains two independent time-domain sequences, X-polarized and Y-polarized. Each of these sequences is a broadband composite signal carrying M subcarriers. The subcarriers are arranged at fixed intervals in the frequency domain and superimposed on each other in the time domain. Therefore, it is necessary to separate the subcarriers through a decomposition operation. The decomposition process must simultaneously process the X and Y polarized signals in parallel to ensure that the timing and phase between polarizations are synchronized.

[0031] Time-domain decomposition (TDD) directly filters and separates composite signals in the time domain without frequency-domain transformation, making it suitable for scenarios with a small number of subcarriers and offering low latency. Specifically, based on the number of subcarriers M, their bandwidth, and center frequency, M matched time-domain bandpass filter banks are designed. Each filter corresponds to the passband of one subcarrier, ensuring high isolation between subcarriers. The X and Y polarized time-domain DSCM signals are filtered in parallel through these M time-domain bandpass filters. Each filter outputs a time-domain signal corresponding to one subcarrier. The filtered subcarrier signals are downsampled by a factor of M to reduce the sampling rate. Simultaneously, timing calibration ensures synchronization between the X and Y polarizations and between each subcarrier, ultimately resulting in 2×M independent time-domain subcarrier signals, which can be directly used for subsequent time-domain nonlinear compensation processing.

[0032] Frequency domain decomposition utilizes the frequency domain gating property of Fourier transform to convert time-domain composite signals to the frequency domain. It achieves accurate subcarrier separation through frequency domain filtering. Frequency domain decomposition can be combined with frequency domain dispersion compensation (CDC) to significantly reduce computational complexity.

[0033] When using frequency domain decomposition, the received DSCM signal is processed by at least one of the following methods: block division, Fourier transform, filtering based on a preset subcarrier frequency configuration, spectrum shifting, spectrum truncation, and downsampling, to obtain multiple subcarrier frequency domain signals. The steps include: The received DSCM signal is divided into blocks based on the overlap retention method or the overlap addition method to obtain multiple data blocks; Specifically, to avoid spectral leakage caused by data truncation during FFT transformation of continuous time-domain signals, and to eliminate dispersion interference at block boundaries, ensuring the accuracy of subsequent subcarrier demodulation and nonlinear compensation, this invention employs a block processing mechanism. First, a fixed data block length N is set, typically chosen to match the number of points in the subsequent FFT transformation, such as 512 or 1024 points. Then, either the overlap-and-save method or the overlap-and-add method is used to slide-divide the continuous DSCM time-domain signal into blocks. If the overlap-and-save method is used, 50% overlap is retained between data blocks, retaining only the middle valid data segments and discarding distorted data at the boundaries. If the overlap-and-add method is used, the overlapping portion of each data block is used to eliminate boundary filtering distortion. After the above processing, the continuous one-dimensional time-domain signal sequence is converted into several discrete data block sequences of length N.

[0034] Perform FFT transformation on each of the data blocks to obtain the frequency domain received signal; Based on the preset number of subcarriers M, subcarrier center frequency, and subcarrier bandwidth, the frequency domain received signal is divided into M independent sub-bands; Specifically, the system parameters are first pre-configured, setting the total number of subcarriers M, the center frequency fc of each subcarrier, and the subcarrier bandwidth Bs. Based on the Nyquist sampling theorem and the system symbol rate, the total bandwidth Btotal of the entire signal frequency band is calculated. In the frequency domain, the total bandwidth Btotal of the broadband frequency domain received signal is uniformly divided into M continuous and non-overlapping independent sub-frequency bands according to the number of subcarriers M. The center frequency of each sub-frequency band corresponds to the preset subcarrier center frequency, and the bandwidth corresponds to the subcarrier bandwidth, thus establishing a dedicated frequency domain search space for each subcarrier.

[0035] The M independent sub-bands are demultiplexed by frequency domain windowing filtering and filter bank processing to obtain M subcarrier frequency domain signals.

[0036] For each independently defined sub-frequency band, subcarrier demultiplexing is performed in the frequency domain. This invention employs a combined approach of frequency domain windowing filtering and polyphase filter banks. Specifically, a spectral window function (such as a raised cosine window) is applied at the boundary of each sub-frequency band to suppress spectral leakage and improve subcarrier isolation. The spectral components corresponding to each sub-frequency band are extracted from the broadband frequency domain signal through frequency domain complex multiplication or time domain convolution operations. Finally, the frequency domain data corresponding to the M sub-frequency bands are separated to obtain M subcarrier frequency domain signals. At this point, each frequency domain subcarrier signal has been freed from interference from adjacent subcarriers and can be independently subjected to subsequent dispersion compensation and nonlinear compensation operations.

[0037] S2: Based on the transmission distance, dispersion parameters, nonlinear parameters and subcarrier configuration of the DSCM signal, construct a virtual reverse transmission model with virtual zero dispersion points; In the process of extracting subcarriers, a joint implementation mechanism of frequency domain subcarrier demultiplexing and chromatic dispersion compensation is adopted. By pre-setting the composite weight of the integrated subcarrier filtering window and the dispersion compensation phase factor, the frequency domain received signal is multiplied once. In a single operation, the isolation and separation of M subcarriers are achieved by using the frequency domain window function, and the chromatic dispersion pre-compensation at the corresponding distance L-z0 is completed by the phase factor multiplication. Finally, a clean, independent M frequency domain subcarrier signal with preliminary linear dispersion correction is output, which significantly reduces the computational complexity and hardware overhead.

[0038] S3: Based on the set of perturbation coefficients corresponding to the virtual reverse transmission model, perform nonlinear compensation on at least a portion of the subcarrier signals to obtain the compensated subcarrier signals; The perturbation coefficient set is used to characterize the coupling relationship between nonlinear effects and dispersion effects within and / or between subcarriers, and the perturbation coefficient set is either a full set of coefficients or a simplified set of coefficients processed by at least one of screening, realization, quantization, threshold pruning, and group accumulation.

[0039] The dispersion compensation includes pre-dispersion compensation and post-dispersion compensation, which are located before and after nonlinear compensation, respectively, forming a symmetrical compensation method. The dispersion compensation includes three aspects: compensation for intra-subcarrier dispersion, alignment of fractional delay caused by dispersion between subcarriers, and unified correction of phase offset caused by transmission path differences between subcarriers.

[0040] The single-step nonlinear compensation is used to compensate for self-phase modulation impairment within subcarriers and / or cross-phase modulation impairment between subcarriers.

[0041] The single-step nonlinear compensation adopts any one of the following: additive compensation model, phase rotation compensation model, or a compensation model combining additive and phase rotation.

[0042] The set of perturbation coefficients is determined based on at least some of the parameters in the following parameters: transmission distance, virtual zero dispersion point location, dispersion parameter, nonlinear parameter, power distribution, pulse waveform, subcarrier spacing, and sampling relationship. It is obtained through analytical calculation, time-domain convolution operation, frequency-domain equivalent operation, or a combination thereof.

[0043] The simplification of the perturbation coefficient set includes at least one of the following: retaining only real-valued perturbation coefficients; deleting low-contribution perturbation coefficients according to a threshold; quantizing the perturbation coefficients; and grouping and accumulating the perturbation coefficients according to subcarrier spacing, symmetry, or coefficient values.

[0044] In some embodiments, the nonlinear compensation can be achieved using additive compensation, phase rotation compensation, or a combination of both. The corresponding compensation amount is calculated from a set of perturbation coefficients, which can be the full set of coefficients or a simplified set of coefficients after real-value filtering, quantization, threshold pruning, or group accumulation. The compensation amount can be obtained through analytical calculation, time-domain convolution, frequency-domain equivalent implementation, or a combination thereof. The specific integral expressions, compensation formulas, and parameter configurations given in this invention are merely exemplary embodiments and do not constitute a limitation on the scope of protection of this invention.

[0045] The nonlinear compensation adopts an additive-multiplicative compensation model, and the compensation formula is as follows: ; In the formula, This represents the signal after nonlinear compensation for the m-th subcarrier. This represents the signal before nonlinear compensation for the m-th subcarrier, where x / y represents the signal with x-polarization or y-polarization. This represents the summation of nonlinear compensation terms. This represents the nonlinear phase product compensation term, and n represents the sampling point index of the signal.

[0046] The virtual zero dispersion point is adjustable, and its preferred value is half of the total transmission distance of the signal in the optical fiber, forming a symmetrical dispersion compensation structure, which is used to increase the power ratio of the real-valued perturbation coefficient and shorten the correlation length.

[0047] For each subcarrier frequency domain signal, perform post-chromatic dispersion compensation with a compensation distance of z0 in the frequency domain, and compare it with L in step S2. The pre-dispersion compensation of z0 works in conjunction with the full-link linear dispersion correction of the total transmission distance L, completely eliminating intra-subcarrier dispersion, inter-subcarrier fractional delay, and phase offset introduced by fiber optic transmission. Next, an inverse fast Fourier transform (IFFT) is performed on each subcarrier frequency domain signal to convert it back to the time domain, forming a time-domain signal. Finally, according to the overlap-preservation or overlap-addition block processing rules used in step S1, overlapping redundant data from the block division is removed, and the time-domain waveform is spliced, time-series calibrated, and synchronized. Simultaneously, for the dual-polarization DSCM system, the compensated signals of x-polarization and y-polarization are synthesized separately, ultimately outputting a complete time-domain output signal that has undergone joint compensation for linear dispersion and nonlinear impairments.

[0048] In a second aspect of the present invention, an optical fiber nonlinearity compensation system for digital subcarrier multiplexing is provided, comprising: Signal acquisition module: used to acquire the received DSCM signal and obtain multiple subcarrier signals based on the received DSCM signal; Signal compensation module: used to perform compensation processing on multiple subcarrier signals respectively to obtain compensated subcarrier signals.

[0049] In a third aspect of the invention, a computer-readable storage medium is provided having a program stored thereon, which, when executed by a processor, implements the digital subcarrier multiplexing fiber nonlinear compensation method.

[0050] The method provided by this invention can be directly applied to dual-polarization DSCM systems, enabling simultaneous processing of x-polarized and y-polarized signals and compensation for coupling nonlinearity damage between polarization states. In long-distance transmission scenarios of 1600–2400 km, it can achieve compensation performance comparable to traditional multi-step digital backpropagation (DBP), while the computational complexity is only 2.4%–3.5% of that of traditional DBP. It has an excellent trade-off between performance and complexity and is very suitable for low-power implementation of high-speed coherent optical communication DSP chips.

[0051] The following specific embodiments illustrate the solution of the present invention in detail: Please see Figure 2 The diagram shown is a flowchart of the first embodiment of the present invention. After receiving the time-domain signal, subcarrier decomposition is performed on the time-domain signal. This decomposition can be implemented in ways including, but not limited to, time-domain decomposition and frequency-domain decomposition. After subcarrier decomposition, a first dispersion compensation (pre-dispersion compensation) is performed on the subcarriers. The compensation distance is L-z0, where L is the total transmission distance of the signal, z0 represents the nonlinear compensation point, and m, m′∈[ M / 2, M / 2 [1] represents the subcarrier index, and M represents the total number of subcarriers. The dispersion compensation formula for the m-th subcarrier is as follows: Formula 1 In the formula, Δω = ω – mΩ, where ω represents the frequency axis, β2 is the second-order dispersion coefficient, z is the transmission distance, and Ω is the spacing between subcarriers.

[0052] Then, nonlinear compensation is performed on the signal after the first dispersion compensation, and the compensation formula is as follows: Formula 2 in: Formula 3 Formula 4 In the formula, This represents the signal after nonlinear compensation. This represents the signal before nonlinear compensation for the m-th subcarrier, where x / y represents the signal with x-polarization or y-polarization. This represents the summation of nonlinear compensation terms, (·) * This indicates the conjugate operation. This represents the nonlinear phase product compensation term. The nonlinear coefficients are represented by n, p, q, and l, which represent the sampling point indices of the signal, respectively. C m’→m ( p , q ) represents the perturbation nonlinear effect coefficient of the m'-th subcarrier on the m-th subcarrier, and is calculated using the following formula: Formula 5 In the formula, Let C represent the perturbation nonlinearity coefficient, w(z) be the power change curve of the signal during fiber transmission related to the fiber attenuation coefficient α and amplifier gain, g(z,t) represent the pulse waveform of the subcarrier signal, g*(z,t) represent the conjugate of g(z,t), T be the sampling interval of the subcarrier signal, and t be the time axis. m’→m The distribution of one case of (p, q) is as follows Figure 3 As shown, the conditions are: subcarrier roll-off coefficient is 0.1Ω = 2π × 8.8 Grad / s, β2 = 20.407 ps. 2 / km, T=62.5 ps, L=1600 km, z0=800 km, α=0.2 dB / km.

[0053] After performing nonlinear compensation on the signal, a second dispersion compensation is performed on the signal according to Equation 1 with a compensation distance of z0, and finally the compensated subcarrier signal is obtained.

[0054] It should be noted that, in Figure 2 In addition to the compensation, other compensation steps such as polarization compensation and phase noise compensation can be added.

[0055] Please see Figure 4 The diagram shown is a detailed flowchart of the second embodiment of the present invention. After receiving the time-domain signal, the time-domain signal is first transformed by time-frequency to obtain multiple subcarrier signals in the frequency domain. The subcarrier signals are then subjected to spectrum extraction in the frequency domain, i.e., subcarrier decomposition, and dispersion impairment is compensated (pre-dispersion compensation). The compensation distance is L-z0, and the compensation formula is Equation 6: Formula 6 in: Formula 7 Formula 8 Formula 9 In the formula, This represents the dispersion compensation for the m-th subcarrier. This represents the fractional relative delay between subcarriers caused by dispersion. This represents the total inter-carrier delay caused by dispersion. This represents the number of discrete sampling points that represent the integer relative delay between subcarriers caused by dispersion.

[0056] After subcarrier signal extraction and dispersion compensation, the subcarrier signals are frequency-time transformed back to the time domain. Then, each of the M subcarriers is delayed according to Equation 9, with a corresponding transmission distance of L-z0, to compensate for the walk-off effect between subcarriers caused by dispersion. Next, nonlinear compensation is performed on the delayed signal, with the compensation formula shown in Equation 2. The nonlinearly compensated signal is then delayed again using Equation 9, with a corresponding transmission distance of L-z0. The signal after this second delay is then time-frequency transformed back to the frequency domain. Subsequently, the signal in the frequency domain undergoes post-dispersion compensation with a compensation distance of z0 using Equation 6, and the post-dispersion compensated subcarrier signal is then frequency-time transformed to obtain the final co-compensated time-domain subcarrier signal.

[0057] Furthermore, the present invention also provides a simplified embodiment, mainly simplifying the nonlinear compensation part in the first and second embodiments, while the remaining steps are the same as in the first and second embodiments. This simplified embodiment simplifies Formula 2, retaining only the phase term, and is expressed as: Formula 10 In the formula, It is calculated using Formula 4.

[0058] After simplification, the perturbation nonlinearity coefficients in Formula 5 retain only real terms. Figure 3 Under the given conditions, the distribution of the perturbation nonlinear action coefficients is as follows: Figure 5 As shown.

[0059] It should be noted that, for formula 4... The calculation includes, but is not limited to, time-domain calculations using convolution and frequency-domain multiplication. For Equations 2 and 10... and The computational complexity can be reduced using methods including, but not limited to, k-means clustering and uniform quantization. Figure 6 This indicates the distribution of the nonlinear action coefficients of the perturbation after processing using uniform amplitude quantization, which can... Figure 5 The number of coefficients has been reduced from 885 to 12.

[0060] The simulation results are provided below.

[0061] Please see Figure 7 The figure shows the simulation results of 5-channel WDM, dual-polarization 64-GBaud digital subcarrier modulation, with the number of subcarriers set to M=8 and z0=L / 2. Figure 7 The figure shows the relationship between the Q² factor and the transmit power per channel under different schemes. Initially, due to the improvement in signal-to-noise ratio (SNR), the Q² factor increases with the increase in input power per channel; however, with further increases in input power, the Q² factor decreases as the Kerr nonlinearity effect becomes dominant. For the case using only chromatic dispersion compensation (CDC), the optimal input power is 1 dBm / channel; while for the conventional digital backpropagation (DBP) with 3 steps per span, the method proposed in this invention, and the simplified method proposed in this invention with unquantized and quantized levels of 12, the optimal value is 2 dBm / channel. Compared with conventional DBP, the FPB-NLC of this invention shows a performance improvement of 0.1 dB; and compared with the case using only CDC, the performance improvement is 0.31 dB. In addition, the simplified method of this invention shows performance comparable to DBP.

[0062] Please see Figure 8 The figure shows the simulation results of 5-channel WDM, dual-polarization 64-GBaud digital subcarrier modulation, comparing the Q² factor gain of the simplified method of this invention (using 3 steps per segment) relative to the case of dispersion compensation only. Figure 8 As can be seen, within the test transmission distance range of 1200 km to 2400 km, the simplified method of the present invention with a quantization level of 12 exhibits performance comparable to conventional digital backpropagation using 3 steps per segment, and achieves a Q² factor gain of approximately 0.2 dB compared to the case with dispersion compensation only.

[0063] Please see Figure 9The results shown are simulation results of 5-channel WDM and dual-polarization 64-GBaud digital subcarrier modulation. By comparing the computational complexity of the traditional digital backpropagation method with the simplified method of the present invention with 12 quantization levels, the results show that the simplified method of the present invention exhibits performance comparable to that of the traditional digital backpropagation. Moreover, when the transmission distance is 1600 km and 2400 km, respectively, its computational complexity is significantly reduced to 3.5% and 2.4% of that of the traditional digital backpropagation.

[0064] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method of optical fiber nonlinear compensation by digital subcarrier multiplexing, characterized by, Includes the following steps: S1: Obtain the received DSCM signal, and obtain multiple subcarrier signals based on the received DSCM signal; S2: Based on the transmission distance, dispersion parameters, nonlinear parameters and subcarrier configuration of the DSCM signal, construct a virtual reverse transmission model with virtual zero dispersion points; S3: Based on the set of perturbation coefficients corresponding to the virtual reverse transmission model, perform nonlinear compensation on at least a portion of the subcarrier signals to obtain the compensated subcarrier signals; The perturbation coefficient set is used to characterize the coupling relationship between nonlinear effects and dispersion effects within and / or between subcarriers, and the perturbation coefficient set is either a full set of coefficients or a simplified set of coefficients processed by at least one of screening, realization, quantization, threshold pruning, and group accumulation.

2. The method of claim 1, wherein the digital subcarrier multiplexing is performed by a digital signal processor (DSP). The received DSCM signal is a time-domain signal containing X-polarization and Y-polarization. Multiple subcarrier signals are obtained from the received DSCM signal through time-domain decomposition or frequency-domain decomposition.

3. The method of claim 2, wherein the digital subcarrier multiplexing is performed by a digital signal processor (DSP). When using the frequency domain decomposition method, the received DSCM signal is subjected to at least one of the following processing methods: block division, Fourier transform, filtering, spectrum shifting, spectrum truncation, and downsampling based on a preset subcarrier frequency configuration, in order to obtain multiple subcarrier frequency domain signals.

4. The method of claim 1, wherein the digital subcarrier multiplexing is performed by a digital signal processor. In the subcarrier extraction process, a joint mechanism of frequency domain subcarrier demultiplexing and chromatic dispersion compensation is adopted. By pre-setting the composite weight of the integrated subcarrier filtering window and the dispersion compensation phase factor, a single frequency domain multiplication is performed on the received signal. This single operation simultaneously achieves: isolation and separation of M subcarriers using a frequency domain window function, and completion of the corresponding distance L using phase factor multiplication. The chromatic dispersion pre-compensation of z0 ultimately outputs a pure, independent M-channel frequency domain subcarrier signal that has completed preliminary linear dispersion correction.

5. The fiber nonlinearity compensation method for digital subcarrier multiplexing according to claim 1, characterized in that, The dispersion compensation includes pre-dispersion compensation and post-dispersion compensation, which are located before and after nonlinear compensation, respectively. The dispersion compensation includes three aspects: compensation for intra-subcarrier dispersion, alignment of fractional delay caused by dispersion between subcarriers, and unified correction of phase offset caused by transmission path differences between subcarriers.

6. The method of claim 1, wherein the digital subcarrier multiplexing is performed by a digital signal processor. The single-step nonlinear compensation is used to compensate for self-phase modulation impairment within subcarriers and / or cross-phase modulation impairment between subcarriers.

7. The method of claim 1 or 6, wherein, The single-step nonlinear compensation adopts any one of the following: additive compensation model, phase rotation compensation model, or a compensation model combining additive and phase rotation.

8. The method of claim 1, wherein, The simplification process of the perturbation coefficient set includes at least one of the following: retaining only real-valued perturbation coefficients; deleting low-contribution perturbation coefficients according to a threshold. The perturbation coefficients are quantized; the perturbation coefficients are grouped and accumulated according to subcarrier spacing, symmetry, or coefficient values.

9. The method of claim 1, wherein the digital subcarrier multiplexing is performed by a digital signal processor. The set of perturbation coefficients is determined based on at least some of the parameters in the following parameters: transmission distance, virtual zero dispersion point location, dispersion parameter, nonlinear parameter, power distribution, pulse waveform, subcarrier spacing, and sampling relationship. It is obtained through analytical calculation, time-domain convolution operation, frequency-domain equivalent operation, or a combination thereof.

10. The method of claim 1, wherein the digital subcarrier multiplexing is performed by a digital signal processor. The virtual zero dispersion point is adjustable, and its preferred value is half of the total transmission distance of the signal in the optical fiber, forming a symmetrical dispersion compensation structure, which is used to increase the power ratio of the real-valued perturbation coefficient and shorten the correlation length.

11. A fiber optic nonlinearity compensation system for digital subcarrier multiplexing, applied to the fiber optic nonlinearity compensation method for digital subcarrier multiplexing as described in any one of claims 1-10, characterized in that, include: Signal acquisition module: used to acquire the received DSCM signal and obtain multiple subcarrier signals based on the received DSCM signal; Signal compensation module: used to perform compensation processing on multiple subcarrier signals respectively to obtain compensated subcarrier signals.

12. A computer-readable storage medium having stored thereon a program, characterized in that, When the program is executed by the processor, it implements the digital subcarrier multiplexing fiber nonlinear compensation method according to any one of claims 1 to 10.

Images