Digital signal processing method and system for optical fiber communication system

Abstract

The application provides a digital signal processing method and system for an optical fiber communication system, comprising: performing signal preprocessing on an electrical signal received by the optical fiber communication system to obtain a block digital signal; processing the block digital signal by using a block digital signal processing algorithm to obtain a batch digital signal; processing the batch digital signal by using a batch signal processing algorithm to obtain an output batch signal; calculating an error according to the batch signal, calculating a gradient to obtain a gradient of a block digital signal processing algorithm parameter and a batch digital signal processing algorithm parameter; iteratively processing the batch digital signal; iteratively processing the block digital signal; performing parallel-serial conversion to obtain a processed digital signal of the optical fiber communication system. The application can effectively improve the digital signal processing performance, reduce the calculation complexity of the digital signal processing, and improve the energy efficiency ratio of the digital signal processing system.

Description

Technical Field

[0001] The present invention relates to the field of optical fiber communication technology, and in particular to a digital signal processing method and system for an optical fiber communication system. Background Art

[0002] In the field of fiber-optic communication systems, digital signal processing methods and systems are used to receive and process transmitted signals. Traditional digital signal processing methods and systems are designed on a module-by-module basis, effectively addressing specific communication system effects. However, existing digital signal processing methods and systems limit the integration of advanced compensation algorithms, such as nonlinear compensation through deep learning. Integrating new algorithms often presents the challenge of redesigning the digital signal processing system. Furthermore, the block-by-block design of existing digital signal processing methods and systems can lead to local minima.

[0003] Patent document CN112865869A (Application Number: CN202110114209.0) discloses a multi-symbol joint digital signal processing method based on nonlinear discrete spectrum signals. The method comprises: receiving an NFDM signal from optical fiber communication and performing an NFT transform on the NFDM signal to obtain a nonlinear discrete spectrum signal; performing multi-symbol LMMSE calculation and noise compensation on the nonlinear discrete spectrum signal to obtain noise-compensated discrete spectrum coefficients; and performing a multi-symbol joint probability distribution decision on the noise-compensated discrete spectrum coefficients to obtain the decided discrete spectrum coefficients of the transmission symbol sequence, thereby improving the quality of the NFDM signal. However, this patent fails to address existing technical issues.

[0004] In response to the above-mentioned defects, the present invention proposes a digital signal processing method and system for optical fiber communication systems, and proposes to use deep learning to optimize the modules of the digital signal processing system on a global scale, thereby achieving a higher performance and lower complexity digital signal processing method and system, greatly improving the energy efficiency ratio of the digital signal processing method and system. Summary of the Invention

[0005] In view of the defects in the prior art, the object of the present invention is to provide a digital signal processing method and system for an optical fiber communication system.

[0006] The digital signal processing method for an optical fiber communication system provided by the present invention comprises:

[0007] Step S1: performing signal preprocessing on the electrical signal received by the optical fiber communication system to obtain a block digital signal;

[0008] Step S2: Processing the block digital signal using a block digital signal processing algorithm to obtain batch digital signals;

[0009] Step S3: Processing the batch digital signal using a batch signal processing algorithm to obtain an output batch signal;

[0010] Step S4: Calculate the error based on the batch signal and calculate the gradient to obtain the gradient of the block digital signal processing algorithm parameters and the batch digital signal processing algorithm parameters;

[0011] Step S5: iteratively processing batch digital signals;

[0012] Step S6: iteratively processing the block digital signal;

[0013] Step S7: Perform parallel-to-serial conversion to obtain a processed digital signal of the optical fiber communication system.

[0014] Preferably, the step S1 includes:

[0015] Step S1.1: performing analog-to-digital conversion and downsampling on the received signal of the optical fiber communication to obtain a digital signal;

[0016] Step S1.2: Filter the digital signal to remove out-of-band noise;

[0017] Step S1.3: Perform serial-to-parallel conversion on the digital signal obtained in step S1.2, and divide it into N blocks of signals according to the overlap-preserve method.

[0018] Preferably, the step S2 includes:

[0019] Step S2.1: using an adaptive gradient descent method, updating the parameters of the block digital signal processing algorithm and setting the parameter gradients of the block digital signal processing algorithm to zero;

[0020] Step S2.2: Take the i-th block signal and perform a first linear compensation on the current signal block using a block digital signal processing algorithm to obtain a block-processed signal; where i is the index of the selected block signal, and its value range is greater than 0 and less than or equal to the number of block signals N;

[0021] Step S2.3: convert the entire block of digital signals into serial-to-parallel signals and divide them into M batches of digital signals according to the overlap-preserve method.

[0022] Preferably, step S3 includes:

[0023] Step S3.1: using an adaptive gradient descent method to update the parameters of the batch digital signal processing algorithm and set the parameter gradients of the batch digital signal processing algorithm to zero;

[0024] Step S3.2: Take the j-th batch signal and use the batch digital signal processing algorithm to perform a second linear compensation on the current batch signal to obtain the batch processed signal; where j is the index of the selected batch signal and its value range is greater than 0 and less than or equal to the number of batch signals M.

[0025] Preferably, step S4 includes:

[0026] Step S4.1: Pre-training phase, using the error formula Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, r is the estimated radius of the symbol constellation point, and y is the symbol output by the batch signal in step S3.2;

[0027] Step S4.2: Tracking phase, using the error formula Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, y is the symbol output by the batch signal in step S3.2, and x is the corresponding label of the symbol output by the batch signal in step S3.2;

[0028] Step S4.3: Using the back propagation algorithm, calculate the gradients of the parameters of the batch digital signal processing algorithm and the block digital signal processing algorithm.

[0029] Preferably, the step S5 includes: judging whether the currently processed batch index j is equal to the total number of batches M; if it is equal to M, ending the iteration; if it is less than M, returning to step S3 for iteration.

[0030] Preferably, the step S6 includes: judging whether the currently processed block index i is equal to the total number of blocks N; if it is equal to N, ending the iteration; if it is less than N, returning to step S2 for iteration.

[0031] Preferably, the step S7 includes: performing parallel-to-serial conversion on the batch digital signal processed signals obtained in step S3.2 to obtain the final digital signal processed signals.

[0032] The digital signal processing system for an optical fiber communication system provided by the present invention comprises:

[0033] Module M1: performs signal preprocessing on the electrical signal received by the optical fiber communication system to obtain block digital signals;

[0034] Module M2: Processing the block digital signal using a block digital signal processing algorithm to obtain batch digital signals;

[0035] Module M3: Process the batch digital signal using the batch signal processing algorithm to obtain the output batch signal;

[0036] Module M4: Calculate the error based on the batch signal and calculate the gradient to obtain the gradient of the block digital signal processing algorithm parameters and the batch digital signal processing algorithm parameters;

[0037] Module M5: iteratively processes batch digital signals;

[0038] Module M6: iterative processing of block digital signals;

[0039] Module M7: performs parallel-to-serial conversion to obtain the processed digital signal of the optical fiber communication system.

[0040] Preferably, the module M1 includes:

[0041] Module M1.1: performs analog-to-digital conversion and down-sampling on the received signal of optical fiber communication to obtain a digital signal;

[0042] Module M1.2: Filters the digital signal to remove out-of-band noise;

[0043] Module M1.3: performs serial-to-parallel conversion on the digital signal obtained by module M1.2 and divides it into N blocks of signals according to the overlap-preserve method;

[0044] The module M2 includes:

[0045] Module M2.1: Update the parameters of the block digital signal processing algorithm using the adaptive gradient descent method and set the parameter gradients of the block digital signal processing algorithm to zero;

[0046] Module M2.2: Take the i-th block signal and perform the first linear compensation on the current signal block using the block digital signal processing algorithm to obtain the block processed signal; where i is the index of the selected block signal, and the value range is greater than 0 and less than or equal to the number of block signals N;

[0047] Module M2.3: convert the entire block of digital signals into serial-to-parallel signals and divide them into M batches of digital signals according to the overlap-preserve method;

[0048] The module M3 includes:

[0049] Module M3.1: Update the parameters of the batch digital signal processing algorithm using adaptive gradient descent and set the gradients of the parameters of the batch digital signal processing algorithm to zero;

[0050] Module M3.2: Take the jth batch signal and perform a second linear compensation on the current batch signal using the batch digital signal processing algorithm to obtain the batch processed signal; where j is the index of the selected batch signal and has a value range greater than 0 and less than or equal to the number of batch signals M;

[0051] The module M4 includes:

[0052] Module M4.1: Pre-training phase, using the error formula Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, r is the estimated radius of the symbol constellation point, and y is the symbol output by the batch signal in module M3.2;

[0053] Module M4.2: Tracking Phase, Using Error Formulas Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, y is the symbol output by the batch signal in module M3.2, and x is the corresponding label of the symbol output by the batch signal in module M3.2;

[0054] Module M4.3: Use the backpropagation algorithm to calculate the gradients of the parameters of batch digital signal processing algorithms and block digital signal processing algorithms;

[0055] The module M5 includes:

[0056] Module M5.1: Determine whether the currently processed batch index j is equal to the total number of batches M. If it is equal to M, the iteration ends. If it is less than M, it returns to module M3 for iteration;

[0057] The module M6 includes: determining whether the currently processed block index i is equal to the total number of blocks N, and ending the iteration if it is equal to N; if it is less than N, returning to the module M2 for iteration;

[0058] The module M7 includes: performing parallel-to-serial conversion on the batch digital signal processed signals obtained by the module M3.2 to obtain the final digital signal processed signals.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] 1. The present invention can improve the processing efficiency of digital signal processing methods and enhance the performance of processed signals;

[0061] 2. The present invention can reduce the complexity of digital signal processing methods;

[0062] 3. This system can be applied to optical fiber communication systems, including long-distance, medium-distance and short-distance transmission and short-distance access optical communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0064] Figure 1 Schematic diagram of the flow of a digital signal processing method for an optical fiber communication system according to the present invention;

[0065] Figure 2The figure is a schematic structural diagram of a digital signal processing system for an optical fiber communication system according to the present invention. DETAILED DESCRIPTION

[0066] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0067] Example 1

[0068] Figure 1 FIG. 1 is a flow chart of a digital signal processing method for an optical fiber communication system according to the present invention. Figure 1 As shown, the digital signal processing method for an optical fiber communication system of the present invention comprises the following steps:

[0069] Step S1: performing signal preprocessing on the electrical signal received by the optical fiber communication system to obtain a block digital signal;

[0070] Step S2: Processing the block digital signal using a block digital signal processing algorithm to obtain batch digital signals;

[0071] Step S3: Processing the batch digital signal using a batch signal processing algorithm to obtain an output batch signal;

[0072] Step S4: Calculate the error based on the batch signal and calculate the gradient to obtain the gradient of the block digital signal processing algorithm parameters and the batch digital signal processing algorithm parameters;

[0073] Step S5: iteratively processing the batch digital signals;

[0074] Step S6: iteratively processing the block digital signal;

[0075] Step S7: Perform parallel-to-serial conversion to obtain a processed digital signal of the optical fiber communication system.

[0076] Specifically, the step S1 adopts:

[0077] Step S1.1: performing analog-to-digital conversion and downsampling on the received signal of the optical fiber communication to obtain a digital signal;

[0078] Step S1.2: Filter the digital signal to remove out-of-band noise;

[0079] Step S1.3: Perform serial-to-parallel conversion on the digital signal obtained in step S1.2, and divide it into N blocks of signals according to the overlap-preserve method.

[0080] In particular, the step S2 employs:

[0081] Step S2.1: updating the parameters of the block digital signal processing algorithm using the adaptive gradient descent method, and setting the parameter gradient of the block digital signal processing algorithm to zero;

[0082] Step S2.2: taking the i-th block signal, performing first linear compensation on the current block signal using the block digital signal processing algorithm to obtain a block-processed signal; where i is the index of the selected block signal, and the value range is greater than 0 and less than or equal to the block signal number N;

[0083] Step S2.3: converting the entire block digital signal into a serial form, and dividing it into M batches of digital signals according to the overlap reservation method.

[0084] In particular, the step S3 employs:

[0085] Step S3.1: updating the parameters of the batch digital signal processing algorithm using the adaptive gradient descent method, and setting the parameter gradient of the batch digital signal processing algorithm to zero;

[0086] Step S3.2: taking the j-th batch signal, performing second linear compensation on the current batch signal using the batch digital signal processing algorithm to obtain a batch-processed signal; where j is the index of the selected batch signal, and the value range is greater than 0 and less than or equal to the batch signal number M.

[0087] In particular, the step S4 employs:

[0088] Step S4.1: in the pre-training stage, the error formula is used to calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, r is the estimated radius of the symbol constellation point, and y is the output symbol of the batch signal in step S3.2;

[0089] Step S4.2: in the tracking stage, the error formula is used to calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, y is the output symbol of the batch signal in step S3.2, and x is the corresponding label of the output symbol of the batch signal in step S3.2;

[0090] Step S4.3: calculating the gradient of the parameters of the batch digital signal processing algorithm and the block digital signal processing algorithm using the back propagation algorithm.

[0091] In particular, the step S5 employs: judging whether the current processed batch index j is equal to the total batch number M, if equal to M, ending the iteration, and if less than M, returning to step S3 for iteration.

[0092] Specifically, the step S6 is as follows: determining whether the currently processed block index i is equal to the total number of blocks N; if it is equal to N, ending the iteration; if it is less than N, returning to step S2 for iteration.

[0093] Specifically, in step S7, the batch digital signal processed signals obtained in step S3.2 are subjected to parallel-to-serial conversion to obtain the final digital signal processed signals.

[0094] Example 2

[0095] Figure 2 FIG. 1 is a schematic diagram of the structure of a digital signal processing system for an optical fiber communication system according to the present invention. Figure 2 As shown, the digital signal processing system for optical fiber communication system of the present invention includes the following structure:

[0096] Module M1: signal preprocessing module;

[0097] Module M2: digital signal processing module;

[0098] Module M3: Batch digital signal processing module;

[0099] Module M4: error and gradient calculation module;

[0100] Module M5: Batch digital signal processing iteration module;

[0101] Module M6: Block digital signal processing iteration module;

[0102] Module M7: Digital signal processing output module.

[0103] Specifically, the module M1 adopts:

[0104] Module M1.1: performs analog-to-digital conversion and down-sampling on the received signal of optical fiber communication to obtain a digital signal;

[0105] Module M1.2: Filters the digital signal to remove out-of-band noise;

[0106] Module M1.3: Perform serial-to-parallel conversion on the digital signal obtained by module M1.2 and divide it into N blocks of signals according to the overlap-preserve method.

[0107] Specifically, the module M2 adopts:

[0108] Module M2.1: Update the parameters of the block digital signal processing algorithm using the adaptive gradient descent method and set the parameter gradients of the block digital signal processing algorithm to zero;

[0109] Module M2.2: Take the ith block signal, and use the block digital signal processing algorithm to perform the first linear compensation on the current block signal to obtain the block processed signal; wherein i is the index of the selected block signal, and the value range is greater than 0 and less than or equal to the block signal number N;

[0110] Module M2.3: Convert the whole block digital signal into a serial-parallel form, and divide it into M batches of digital signals according to the overlap reservation method.

[0111] Specifically, the module M3 adopts:

[0112] Module M3.1: Update the parameters of the batch digital signal processing algorithm using the adaptive gradient descent method, and set the parameter gradient of the batch digital signal processing algorithm to zero;

[0113] Module M3.2: Take the jth batch signal, and use the batch digital signal processing algorithm to perform the second linear compensation on the current batch signal to obtain the batch processed signal; wherein j is the index of the selected batch signal, and the value range is greater than 0 and less than or equal to the batch signal number M.

[0114] Specifically, the module M4 adopts:

[0115] Module M4.1: In the pre-training stage, use the error formula to calculate the error; wherein Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, r is the estimated radius of the symbol constellation point, and y is the output symbol of the batch signal in module M3.2;

[0116] Module M4.2: In the tracking stage, use the error formula to calculate the error; wherein Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, y is the output symbol of the batch signal in module M3.2, and x is the corresponding label of the output symbol of the batch signal in module M3.2.

[0117] Module M4.3: Use the back propagation algorithm to calculate the gradient of the parameters of the batch digital signal processing algorithm and the block digital signal processing algorithm.

[0118] Specifically, the module M5 adopts: judge whether the current processed batch index j is equal to the total batch number M, if equal to M, end the iteration, and if less than M, return to module M3 for iteration.

[0119] Specifically, the module M6 adopts: judge whether the current processed block index i is equal to the total block number N, if equal to N, end the iteration, and if less than N, return to module M2 for iteration.

[0120] Specifically, the module M7 performs parallel-to-serial conversion on the batch digital signal processed signals obtained by the module M3.2 to obtain the final digital signal processed signals.

[0121] The technical problem to be solved by the present invention is to realize digital signal processing of an optical fiber communication system. The present invention adaptively realizes parameter optimization in a digital signal processing method and a system based on a gradient descent algorithm; the present invention designs a parameter update method for a block digital signal processing method and a parameter update method for a batch digital signal processing method, so that the parameters in the digital signal processing method and the system better match the effect of an optical fiber channel; the digital signal processing method and system of the present invention can be set to perform processing at 1 to 2 times the symbol sampling, and at 1 times the symbol sampling, the complexity is greatly reduced; the digital signal processing method and system proposed by the present invention can be applied to optical fiber communication systems in different scenarios, and have higher transmission performance of optical fiber communication systems.

[0122] Those skilled in the art will appreciate that, in addition to implementing the system, device, and various modules provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, and the like by logically programming the method steps. Therefore, the system, device, and various modules provided by the present invention can be considered a hardware component, and the modules included therein for implementing various programs can also be considered structures within the hardware component; the modules for implementing various functions can also be considered both software programs for implementing the method and structures within the hardware component.

[0123] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A digital signal processing method for an optical fiber communication system, characterized in that: include: Step S1: performing signal preprocessing on the electrical signal received by the optical fiber communication system to obtain a block digital signal; Step S2: Processing the block digital signal using a block digital signal processing algorithm to obtain batch digital signals; Step S3: Processing the batch digital signal using a batch signal processing algorithm to obtain an output batch signal; Step S4: Calculate the error based on the batch signal and calculate the gradient to obtain the gradient of the block digital signal processing algorithm parameters and the batch digital signal processing algorithm parameters; Step S5: iteratively processing the batch digital signals; Step S6: iteratively processing the block digital signal; Step S7: performing parallel-to-serial conversion to obtain a processed digital signal of the optical fiber communication system; The step S1 comprises: Step S1.1: performing analog-to-digital conversion and downsampling on the received signal of the optical fiber communication to obtain a digital signal; Step S1.2: Filter the digital signal to remove out-of-band noise; Step S1.3: Perform serial-to-parallel conversion on the digital signal obtained in step S1.2 and divide it into N blocks of signals according to the overlap-preserve method; The step S2 comprises: Step S2.1: using an adaptive gradient descent method, updating the parameters of the block digital signal processing algorithm and setting the parameter gradients of the block digital signal processing algorithm to zero; Step S2.2: Take the i-th block signal and perform a first linear compensation on the current signal block using a block digital signal processing algorithm to obtain a block-processed signal; where i is the index of the selected block signal, and its value range is greater than 0 and less than or equal to the number of block signals N; Step S2.3: convert the entire block of digital signals into serial-to-parallel signals and divide them into M batches of digital signals according to the overlap-preserve method; The step S3 comprises: Step S3.1: using an adaptive gradient descent method to update the parameters of the batch digital signal processing algorithm and set the parameter gradients of the batch digital signal processing algorithm to zero; Step S3.2: Take the j-th batch signal and use the batch digital signal processing algorithm to perform a second linear compensation on the current batch signal to obtain the batch processed signal; where j is the index of the selected batch signal and its value range is greater than 0 and less than or equal to the number of batch signals M.

2. The digital signal processing method for an optical fiber communication system according to claim 1, wherein: The step S4 comprises: Step S4.1: Pre-training phase, using the error formula Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, r is the estimated radius of the symbol constellation point, and y is the symbol output by the batch signal in step S3.2; Step S4.2: Tracking phase, using the error formula Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, y is the symbol output by the batch signal in step S3.2, and x is the corresponding label of the symbol output by the batch signal in step S3.2; Step S4.3: Using the back propagation algorithm, calculate the gradients of the parameters of the batch digital signal processing algorithm and the block digital signal processing algorithm.

3. The digital signal processing method for an optical fiber communication system according to claim 2, wherein: The step S5 includes: determining whether the currently processed batch index j is equal to the total batch number M; if it is equal to M, ending the iteration; if it is less than M, returning to step S3 for iteration.

4. The digital signal processing method for an optical fiber communication system according to claim 3, wherein: The step S6 includes: judging whether the currently processed block index i is equal to the total number of blocks N; if it is equal to N, ending the iteration; if it is less than N, returning to step S2 for iteration.

5. The digital signal processing method for an optical fiber communication system according to claim 4, wherein: The step S7 includes: performing parallel-to-serial conversion on the batch digital signal processed signals obtained in step S3.2 to obtain the final digital signal processed signals.

6. A digital signal processing system for an optical fiber communication system, characterized in that: include: Module M1: performs signal preprocessing on the electrical signal received by the optical fiber communication system to obtain block digital signals; Module M2: Processing the block digital signal using a block digital signal processing algorithm to obtain batch digital signals; Module M3: Process the batch digital signal using the batch signal processing algorithm to obtain the output batch signal; Module M4: Calculate the error based on the batch signal and calculate the gradient to obtain the gradient of the block digital signal processing algorithm parameters and the batch digital signal processing algorithm parameters; Module M5: iteratively processes batch digital signals; Module M6: iterative processing of block digital signals; Module M7: performs parallel-to-serial conversion to obtain the processed digital signal of the optical fiber communication system; The module M1 includes: Module M1.1: performs analog-to-digital conversion and down-sampling on the received signal of optical fiber communication to obtain a digital signal; Module M1.2: Filters the digital signal to remove out-of-band noise; Module M1.3: performs serial-to-parallel conversion on the digital signal obtained by module M1.2 and divides it into N blocks of signals according to the overlap-preserve method; The module M2 includes: Module M2.1: Update the parameters of the block digital signal processing algorithm using the adaptive gradient descent method and set the parameter gradients of the block digital signal processing algorithm to zero; Module M2.2: Take the i-th block signal and perform the first linear compensation on the current signal block using the block digital signal processing algorithm to obtain the block processed signal; where i is the index of the selected block signal, and the value range is greater than 0 and less than or equal to the number of block signals N; Module M2.3: convert the entire block of digital signals into serial-to-parallel signals and divide them into M batches of digital signals according to the overlap-preserve method; The module M3 includes: Module M3.1: Update the parameters of the batch digital signal processing algorithm using adaptive gradient descent and set the gradients of the parameters of the batch digital signal processing algorithm to zero; Module M3.2: Take the j-th batch signal and use the batch digital signal processing algorithm to perform a second linear compensation on the current batch signal to obtain the batch processed signal; where j is the index of the selected batch signal and its value range is greater than 0 and less than or equal to the number of batch signals M.

7. The digital signal processing system for optical fiber communication system according to claim 6, characterized in that: The module M4 includes: Module M4.1: Pre-training phase, using the error formula Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, r is the estimated radius of the symbol constellation point, and y is the symbol output by the batch signal in module M3.2; Module M4.2: Tracking Phase, Using Error Formulas Calculate the error; where Loss is the error, m is the number of symbols contained in the batch digital signal, k is the index of the symbol in the batch digital signal, y is the symbol output by the batch signal in module M3.2, and x is the corresponding label of the symbol output by the batch signal in module M3.2; Module M4.3: Use the backpropagation algorithm to calculate the gradients of the parameters of batch digital signal processing algorithms and block digital signal processing algorithms; The module M5 includes: Module M5.1: Determine whether the currently processed batch index j is equal to the total number of batches M. If it is equal to M, the iteration ends. If it is less than M, it returns to module M3 for iteration; The module M6 includes: determining whether the currently processed block index i is equal to the total number of blocks N, and ending the iteration if it is equal to N; if it is less than N, returning to the module M2 for iteration; The module M7 includes: performing parallel-to-serial conversion on the batch digital signal processed signals obtained by the module M3.2 to obtain the final digital signal processed signals.

Images