A pre-equalization based ultra-high order qam modulation transmission system and method

Abstract

The application provides a pre-equalization-based ultra-high-order QAM modulation transmission system and method, which comprises a transmitting end algorithm module, a data transmitting module, a data receiving module and a receiving end algorithm module.The transmitting end algorithm module comprises a pre-equalization processing layer, which processes a to-be-sent data matrix based on a weight matrix; the weight matrix is updated based on the number of 0 taps in the weight matrix after processing the to-be-sent data matrix each time; the data transmitting module comprises an arbitrary wave generator and an IQ modulator, and outputs data through the arbitrary wave generator and the IQ modulator; the data receiving module comprises a coherent receiver and an oscilloscope, the coherent receiver is connected with the IQ modulator through a transmission optical fiber, and incoming data is sequentially processed through the coherent receiver and the oscilloscope; the receiving end algorithm module comprises a decoding layer, receives data transmitted by the oscilloscope, and obtains a received data matrix through the decoding layer.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of optical fiber communication technology, and in particular to a pre-equalization-based ultra-high-order QAM modulation transmission system and method. BACKGROUND

[0002] In recent years, with the progress of information science and 5G network research technology, people's access to information is getting faster and faster, and the information exchange rate is getting higher and higher. How to more effectively use certain frequency spectrum resources to realize high-speed data transmission has become the focus of the communication industry. At present, the global main network data flow is carried by the optical fiber communication system, so the demand for optical fiber transmission capacity is accelerating, and meeting the rapid growth of network capacity has become a severe challenge for future optical communication technology.

[0003] According to Shannon's theorem, there is a theoretical limit to channel capacity. Research shows that when the signal satisfies the Gaussian distribution, it can infinitely approach this limit, so the probability shaping technology (PS) is introduced to map the information to a non-uniform signal that meets the Gaussian distribution. At the same time, the modulation format is as high as possible, and the ultra-high-order modulation technology can increase the number of bits carried by each symbol under the condition of limited wavelength resources, improve the spectral efficiency and reduce the cost per bit.

[0004] With the increase of modulation order, the requirement of signal for digital signal processing algorithm is also higher and higher, and how to improve the performance of ultra-high-order modulation system under the condition of existing devices is a hot spot in the field of communication. SUMMARY

[0005] In view of this, embodiments of the present application provide a pre-equalization-based ultra-high-order QAM modulation transmission system to eliminate or improve one or more defects in the prior art.

[0006] One aspect of the present application provides a pre-equalization-based ultra-high-order QAM modulation transmission system, which comprises:

[0007] A transmitting end algorithm module, the transmitting end algorithm module comprises a pre-equalization processing layer, the pre-equalization processing layer processes a to-be-sent data matrix based on a weight matrix, and the weight matrix is updated based on the number of 0 taps in the weight matrix after processing the to-be-sent data matrix each time;

[0008] A data transmitting module, the data transmitting module comprises an arbitrary wave generator and an IQ modulator, the data transmitting module receives a data matrix output by the transmitting end algorithm module, and outputs data through the arbitrary wave generator and the IQ modulator;

[0009] A data receiving module, which comprises a coherent receiver and an oscilloscope, the coherent receiver is connected with the IQ modulator through a transmission optical fiber, and the incoming data is sequentially processed by the coherent receiver and the oscilloscope.

[0010] A receiving end algorithm module, which comprises a decoding layer, receives the data transmitted by the oscilloscope, and obtains a receiving data matrix through the decoding layer.

[0011] By adopting the above scheme, firstly, the pre-equalization processing layer is arranged in the transmitting end algorithm module, and the smaller the normalized tap coefficient is, the less important the tap is, and the smaller the influence of deleting the tap on the system equalization performance is; the number of 0 taps in the weight matrix is used to update the weight of the pre-equalization processing layer, so that the pre-equalization calculation is more accurate and simplified, and the accuracy of the digital signal processing algorithm for data calculation is improved.

[0012] In some embodiments of the present application, the transmitting end algorithm module comprises a pre-equalization processing layer, an up-sampling layer and a root raised cosine shaping filter, and the to-be-sent data matrix is sequentially processed by the pre-equalization processing layer, the up-sampling layer and the root raised cosine shaping filter.

[0013] In some embodiments of the present application, the to-be-sent data matrix is processed by a probability shaping algorithm before being input into the pre-equalization processing layer.

[0014] In some embodiments of the present application, the receiving end algorithm module comprises a resampling layer, an orthogonalization normalization layer, a root raised cosine matching filter, a down-sampling layer, a clock recovery layer, a post-equalization layer, a phase estimation layer and a decoding layer, and the data transmitted by the oscilloscope is sequentially processed by the resampling layer, the orthogonalization normalization layer, the root raised cosine matching filter, the down-sampling layer, the clock recovery layer, the post-equalization layer, the phase estimation layer and the decoding layer.

[0015] In some embodiments of the present application, the phase estimation layer is sequentially processed by a principal component analysis method and a BPS phase recovery algorithm.

[0016] BPS (Blind Phase Search, blind phase search algorithm).

[0017] In some embodiments of the present application, after the receiving end algorithm module obtains the receiving data matrix, the bit error rate is calculated based on the receiving data matrix and the to-be-sent data matrix before being input into the transmitting end algorithm module.

[0018] In some embodiments of the present application, in the step of processing the to-be-sent data matrix based on the weight matrix in the pre-equalization processing layer, the processing is performed based on the following formula:

[0019] y(k) = w T (k)x(k);

[0020] y(k) represents the data matrix output by the pre-equalization processing layer, w T (k) represents the transpose matrix of the current weight matrix, and x(k) represents the to-be-sent data matrix input to the pre-equalization processing layer.

[0021] In some embodiments of the present application, in the step of updating the weight matrix based on the number of 0 taps in the weight matrix, the weight matrix is updated based on the following formula:

[0022]

[0023] wherein w(k+1) represents the updated weight matrix, w(k) represents the weight matrix before updating, μ is a learning rate parameter, e(k) represents an error signal, x(k) represents the to-be-sent data matrix input to the pre-equalization processing layer, γ represents a regularization parameter, represents the sub-differential of f(w(k)), and f(w(k)) represents the l0 norm of the weight matrix before updating.

[0024] In some embodiments of the present application, the error signal is calculated according to the following formula:

[0025] e(k) = d(k) - y(k);

[0026] wherein e(k) represents an error signal, d(k) represents a preset standard value, and y(k) represents the data matrix output by the pre-equalization processing layer.

[0027] Another aspect of the present application provides a pre-equalization-based ultra-high-order QAM modulation transmission method of the above-mentioned system, and the steps of the method include:

[0028] inputting the to-be-sent data matrix to a transmitting-end algorithm module, the transmitting-end algorithm module including a pre-equalization processing layer, the pre-equalization processing layer processing the to-be-sent data matrix based on a weight matrix, the weight matrix being updated based on the number of 0 taps in the weight matrix after processing the to-be-sent data matrix each time;

[0029] a data transmitting module receiving the data matrix output by the transmitting-end algorithm module and outputting data through an arbitrary wave generator and an IQ modulator;

[0030] The data receiving module receives the data transmitted by the data transmitting module, and the data receiving module comprises a coherent receiver and an oscilloscope, the coherent receiver is connected with the IQ modulator through a transmission optical fiber, and the incoming data is sequentially processed by the coherent receiver and the oscilloscope.

[0031] The receiving end algorithm module receives the data transmitted by the oscilloscope, and a receiving data matrix is obtained by analyzing a decoding layer of the receiving end algorithm module.

[0032] Additional advantages, objects, and features of the application will be set forth in part in the description which follows, and will become apparent to those skilled in the art upon examination of the following or can be learned by practice of the application. The objects and other advantages of the application can be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

[0033] It will be understood by those skilled in the art that the objects and advantages of the present application can be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0034] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the principles of the application.

[0035] Figure 1 A schematic diagram of a first embodiment of the pre-equalization-based ultra-high-order QAM modulation transmission system of the present application;

[0036] Figure 2 An embodiment flowchart of the pre-equalization-based ultra-high-order QAM modulation transmission system of the present application;

[0037] Figure 3 Another embodiment flowchart of the pre-equalization-based ultra-high-order QAM modulation transmission system of the present application;

[0038] Figure 4 A spectrum change graph (left) caused by the influence of nonlinearity, etc., and a correct spectrum graph (right) to be achieved after compensation;

[0039] Figure 5 A pre-equalization algorithm schematic diagram;

[0040] Figure 6 A cascade BPS algorithm schematic diagram;

[0041] Figure 7 A 1GBaud PDM-PS-1024QAM transmission optical signal-to-noise ratio and bit error rate relationship curve. DETAILED DESCRIPTION

[0042] In order to make the objects, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application in combination with the embodiments and drawings. Herein, the illustrative embodiments of the present application and the description thereof are used to explain the present application, but not as a limitation to the present application.

[0043] Herein, it is also needed to be explained that, in order to avoid the present application being obscured by unnecessary details, only the structures and / or processing steps closely related to the solutions according to the present application are shown in the drawings, and other details not closely related to the present application are omitted.

[0044] It should be emphasized that the term "comprises / comprising" as used herein is used to indicate the presence of a feature, element, step or component, but not to exclude the presence or addition of one or more other features, elements, steps or components.

[0045] Herein, it is also needed to be explained that, if not specially stated, the term "connected" as used herein can not only mean direct connection, but also mean indirect connection with an intermediate.

[0046] Hereinafter, the embodiments of the present application will be described with reference to the drawings. In the drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0047] To solve the above problems, as shown in Figures 1-3 The present application proposes a pre-equalization-based ultra-high-order QAM modulation transmission system, which comprises:

[0048] As shown in Figure 5 The transmitting-end algorithm module 100 comprises a pre-equalization processing layer, which processes the to-be-sent data matrix based on a weight matrix, and updates the weight matrix based on the number of 0 taps in the weight matrix after processing the to-be-sent data matrix each time.

[0049] The to-be-sent data matrix is input to the transmitting-end algorithm module in the form of an ultra-high-order layer QAM signal.

[0050] In some embodiments of the present application, the pre-equalization processing layer is realized by a Viterbi equalizer.

[0051] In some embodiments of the present application, the number of 0 taps is a preset parameter, the largest tap coefficient in the weight matrix is extracted, the ratio of each tap coefficient to the largest tap coefficient is calculated, and the tap coefficients corresponding to the number of smaller ratios of the number of 0 taps are deleted.

[0052] With the above solutions, as shown in Figure 5As shown, pre-equalization is realized through two stages: first, in the channel estimation stage, the required pre-equalization filter coefficients are calculated after the training sequence transmission is completed, and when the equalizer reaches a steady state, the required pre-equalization filter is obtained. Then in the actual transmission stage, the signal is convolved with the pre-equalization filter on the time domain at the transmitting end, and the signal pre-equalization technology is completed.

[0053] The pre-equalization principle is as shown in Figure 4 As shown, with the increase of the modulation order, the difficulty of signal generation and recovery is greatly increased, and the requirements for various devices are also increasing, and the advantage of introducing pre-processing is that it can solve the problems of device nonlinearity and other defect distortion at a relatively small cost. In the present example, device nonlinearity will cause the signal spectrum to change, as shown in Figure 4 As shown, the introduction of pre-equalization can add the opposite pre-compensation to the signal spectrum part to be compensated in advance, so as to eliminate the non-linear distortion and restore the spectrum.

[0054] The data transmitting module 200 comprises an arbitrary wave generator and an IQ modulator, the data transmitting module receives the data matrix output by the transmitting end algorithm module, and outputs the data through the arbitrary wave generator and the IQ modulator.

[0055] In some embodiments of the present application, the arbitrary wave generator (AWG) is connected with the IQ modulator, the IQ modulator is provided with a Mach-Zehnder modulator, and the Mach-Zehnder modulator is connected with a signal light source (LD).

[0056] The data receiving module 300 comprises a coherent receiver and an oscilloscope, the coherent receiver is connected with the IQ modulator through a transmission optical fiber, and the incoming data is sequentially processed through the coherent receiver and the oscilloscope.

[0057] In some embodiments of the present application, the coherent receiver is connected with a local oscillator light source LO.

[0058] The receiving end algorithm module 400 comprises a decoding layer, the receiving end algorithm module receives the data transmitted by the oscilloscope, and obtains a receiving data matrix through analysis of the decoding layer.

[0059] In some embodiments of the present application, the transmitting end algorithm module is further provided with an encoding layer, and the encoding layer and the decoding layer both adopt a Gray coding mode.

[0060] With the above scheme, firstly, the scheme is provided with a pre-equalization processing layer in the transmitting end algorithm module. Since the smaller the normalized tap coefficient is, the less important the tap is, and the smaller the influence of deleting the tap on the system equalization performance is, the scheme updates the weight of the pre-equalization processing layer based on the number of 0 taps in the weight matrix, so that the pre-equalization calculation is more accurate and simplified, and the accuracy of the digital signal processing algorithm for data calculation is improved.

[0061] As shown in Figure 2 In some embodiments of the application, the transmitting end algorithm module includes a pre-equalization processing layer, an up-sampling layer and a root raised cosine shaping filter, and the to-be-sent data matrix sequentially passes through the pre-equalization processing layer, the up-sampling layer and the root raised cosine shaping filter for processing.

[0062] In some embodiments of the application, the to-be-sent data matrix is processed by a probability shaping algorithm before being input into the pre-equalization processing layer.

[0063] In some embodiments of the application, the receiving end algorithm module includes a resampling layer, an orthogonalization normalization layer, a root raised cosine matching filter, a down-sampling layer, a clock recovery layer, a post-equalization layer, a phase estimation layer and a decoding layer, and the data transmitted by the oscilloscope sequentially passes through the resampling layer, the orthogonalization normalization layer, the root raised cosine matching filter, the down-sampling layer, the clock recovery layer, the post-equalization layer, the phase estimation layer and the decoding layer for processing.

[0064] In the specific implementation process, the orthogonalization normalization layer uses the GSOP algorithm for orthogonalization, the clock recovery layer uses the Gardner algorithm for clock recovery, and the post-equalization layer uses the least mean square error algorithm for processing.

[0065] As shown in Figure 6 In some embodiments of the application, the phase estimation layer sequentially uses the principal component analysis method and the BPS phase recovery algorithm for processing.

[0066] BPS (Blind Phase Search, blind phase search algorithm).

[0067] With the above scheme, the existence of the laser linewidth will cause the signal to be affected by the phase noise, so that the constellation point of the signal is rotated, and the carrier phase algorithm is needed for compensation. The mainstream algorithm in the prior art is BPS algorithm, the line width tolerance of the BPS algorithm is high, the phase noise estimation accuracy is high, but the complexity increases with the increase of the modulation format. In addition, principal component analysis is a data processing algorithm used for extracting the most critical data features, at the receiving end, the BPS algorithm is combined with the principal component analysis method to realize a cascade BPS algorithm, like the conventional BPS algorithm, phase demapping is also needed to remove the phase ambiguity, and then the obtained data is subjected to the BPS algorithm. The cascade BPS algorithm has the advantages of reducing the overall complexity of the algorithm in 1024QAM and the like super-high-order modulation formats, and can obtain good estimation performance under different SNRs.

[0068] Figure 7 The curve of the bit error rate of the PDM-PS-1024QAM signal with the change of the optical signal-to-noise ratio in the transmission environment of the BTB is shown. The equalization effects of the first-order and third-order Walter filters are compared. It can be seen that the super-high-order transmission system adopted in the application can ensure that the bit error rate is below the FEC threshold under the processing of the third-order Walter filter, and the spectral efficiency reaches 16.56bps / Hz, so that high-spectral-efficiency transmission under the super-high-order modulation format is realized.

[0069] In some embodiments of the application, the receiving end algorithm module calculates the bit error rate based on the received data matrix and the to-be-sent data matrix before input into the transmitting end algorithm module after the received data matrix is parsed.

[0070] In some embodiments of the application, in the step of processing the to-be-sent data matrix based on the weight matrix in the pre-equalization processing layer, the processing is performed based on the following formula:

[0071] y(k)=w T (k)x(k);

[0072] y(k) represents the data matrix output by the pre-equalization processing layer, w T (k) represents the transpose matrix of the current weight matrix, and x(k) represents the to-be-sent data matrix input into the pre-equalization processing layer.

[0073] In some embodiments of the application, in the step of updating the weight matrix based on the number of 0 taps in the weight matrix, the weight matrix is updated based on the following formula:

[0074]

[0075] wherein w(k+1) represents an updated weight matrix, w(k) represents a weight matrix before updating, μ is a learning rate parameter, e(k) represents an error signal, x(k) represents a matrix of data to be transmitted input to the pre-equalization processing layer, γ represents a regularization parameter, represents a sub-differential of f(w(k)), and f(w(k)) represents an l0 norm of the weight matrix before updating.

[0076] In some embodiments of the present application, the error signal is calculated according to the following formula:

[0077] e(k) = d(k) - y(k);

[0078] wherein e(k) represents an error signal, d(k) represents a preset standard value, and y(k) represents a matrix of data output by the pre-equalization processing layer.

[0079] In some embodiments of the present application, the number of 0 taps is a preset parameter, the largest tap coefficient in the weight matrix is extracted, the ratio of each tap coefficient to the largest tap coefficient is calculated, and the tap coefficients corresponding to the number of smaller ratios are deleted.

[0080] In the specific implementation process, the degree of simplification of the pre-equalization processing layer of the present scheme relative to the traditional equalization processing is calculated according to the following formula:

[0081] M1 = l1

[0082]

[0083]

[0084] C1 = M1 + 2M2 + 3M3

[0085] C2 = M1 + 2(M2 - N2) + 3(M3 - N3)

[0086] wherein l1, l2 and l3 represent the order memory lengths of 1, 2 and 3, M1, M2 and M3 are the tap numbers of the 1st, 2nd and 3rd order kernels respectively, N2 and N3 are the numbers of deleted tap coefficients of the 2nd and 3rd order kernels respectively,

[0087] C1 and C2 are the complexity of the calculation of the traditional algorithm and the complexity of the calculation of the present scheme respectively.

[0088] It can be seen that C2 depends on the values of l1, l2, l3, N2 and N3, so the calculation complexity of the algorithm is also related to these parameters only, and the reduced complexity P can be represented as:

[0089]

[0090] Therefore, compared with the conventional Voltera filter, this scheme introduces the concept of regularization and proposes a stable complexity reduction based on the l0 / l1 norm. The advantage is that it can significantly reduce the computational complexity while achieving the same bit error rate performance.

[0091] To address the high complexity of the Voltera equalizer, this solution introduces the sparsity principle. When multiple zero taps appear in the system and the locations of non-zero taps are uncertain, the norm-based regularization algorithm has the best mean square error performance. This solution defines the normalized tap coefficient as the ratio of the tap coefficient of the k-th order kernel to the maximum tap coefficient of the k-th order kernel. Obviously, the smaller the normalized tap coefficient, the less important the tap, and the smaller the impact of deleting the tap on the system's equalization performance. At this point, after the number of taps to be deleted is determined, the normalized tap coefficients are sorted from small to large, and the first tap is selected for deletion in turn to obtain a simplified system. The system is then retrained to obtain the highly robust sparse minimum mean square error-Voltera.

[0092] Another aspect of the present invention provides a pre-equalization-based ultra-high-order QAM modulation transmission method as described above, the method comprising the steps of:

[0093] Inputting the data matrix to be transmitted into a transmitting-end algorithm module, the transmitting-end algorithm module including a pre-equalization processing layer, the pre-equalization processing layer processing the data matrix to be transmitted based on a weight matrix, and updating the weight matrix based on the number of zero taps in the weight matrix after each processing of the data matrix to be transmitted;

[0094] The data transmission module receives the data matrix output by the transmitting end algorithm module and outputs the data through an arbitrary wave generator and an IQ modulator;

[0095] The data receiving module receives the data transmitted by the data transmitting module. The data receiving module includes a coherent receiver and an oscilloscope. The coherent receiver is connected to the IQ modulator via a transmission optical fiber. The incoming data is sequentially processed by the coherent receiver and the oscilloscope.

[0096] The receiving-end algorithm module receives the data input by the oscilloscope, and obtains a receiving data matrix through the decoding layer analysis of the receiving-end algorithm module.

[0097] Those of ordinary skill in the art will appreciate that the various illustrative components, systems and methods described in connection with the embodiments disclosed herein can be implemented as hardware, software, or both. The particular implementation is dependent on the specific application and design constraints imposed on the overall system. Skilled persons can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application. When implemented in hardware, the hardware can comprise, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, a plug-in, a functional card, etc. When implemented in software, the elements of the present application are the program or code segments to perform a specific task. The program or code segments can be stored in a machine-readable medium, or transmitted by a carrier wave as data signals over a transmission medium or communication link.

[0098] It is to be understood that the application is not limited to particular configurations and processes described herein. For the sake of brevity, conventional methods will not necessarily be described in detail. In the above embodiments, several specific steps are described and illustrated in order to provide a thorough disclosure of the application. However, the skilled person will appreciate that the method processes of the application are not limited to the specific steps described and illustrated, and that various modifications, permutations, and additions thereto can be made without departing from the spirit and scope of the present application.

[0099] In the present application, features described and / or illustrated in relation to one embodiment can be used in the same or a similar way in one or more other embodiments, and / or combined with or instead of features of other embodiments.

[0100] The above description is only preferred embodiments of the present application, and is not intended to limit the present application. The embodiments of the present application can be variously changed and modified by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the scope of the present application.

Claims

1. A pre-equalization based ultra-high order QAM modulation transmission system, characterized in that: The system comprises: A transmitter algorithm module includes a pre-equalization processing layer, wherein the pre-equalization processing layer processes the data matrix to be transmitted based on a weight matrix, and after each processing of the data matrix to be transmitted, the weight matrix is ​​updated based on the number of 0 taps in the weight matrix, and the weight matrix is ​​updated based on the following formula: in, represents the updated weight matrix, represents the weight matrix before update, is the learning rate parameter, represents the error signal, represents the data matrix to be sent input to the pre-equalization processing layer, represents the regularization parameter, express The subdifferential of Represents the weight matrix before update norm; A data transmission module, comprising an arbitrary wave generator and an IQ modulator, receives the data matrix output by the transmitting-end algorithm module, and outputs data through the arbitrary wave generator and the IQ modulator; A data receiving module, comprising a coherent receiver and an oscilloscope, wherein the coherent receiver is connected to the IQ modulator via a transmission optical fiber, and the incoming data is sequentially processed by the coherent receiver and the oscilloscope; The receiving end algorithm module includes a decoding layer. The receiving end algorithm module receives data input by the oscilloscope and obtains a receiving data matrix through parsing by the decoding layer.

2. The ultra-high-order QAM modulation transmission system based on pre-equalization according to claim 1, characterized in that: The transmitting end algorithm module includes a pre-equalization processing layer, an upsampling layer and a root raised cosine shaping filter, and the data matrix to be sent is processed sequentially by the pre-equalization processing layer, the upsampling layer and the root raised cosine shaping filter.

3. The ultra-high-order QAM modulation transmission system based on pre-equalization according to claim 1, characterized in that: The data matrix to be sent is processed by a probability shaping algorithm before being input into the pre-equalization processing layer.

4. The ultra-high-order QAM modulation transmission system based on pre-equalization according to claim 1, characterized in that: The receiving end algorithm module includes a resampling layer, an orthogonalization normalization layer, a root raised cosine matched filter, a downsampling layer, a clock recovery layer, a post-equalization layer, a phase estimation layer and a decoding layer. The data input by the oscilloscope is processed sequentially through the resampling layer, the orthogonalization normalization layer, the root raised cosine matched filter, the downsampling layer, the clock recovery layer, the post-equalization layer, the phase estimation layer and the decoding layer.

5. The pre-equalization based ultra-high order QAM modulation transmission system according to claim 4, characterized in that: The phase estimation layer uses principal component analysis and BPS phase recovery algorithm to perform processing in sequence.

6. The pre-equalization based ultra-high order QAM modulation transmission system according to claim 1, characterized in that: After parsing and obtaining the received data matrix, the receiving-end algorithm module calculates the bit error rate based on the received data matrix and the to-be-sent data matrix before inputting into the transmitting-end algorithm module.

7. The ultra-high-order QAM modulation transmission system based on pre-equalization according to claim 1, characterized in that: In the step of processing the data matrix to be transmitted based on the weight matrix in the pre-equalization processing layer, the processing is performed based on the following formula: represents the data matrix output by the pre-equalization layer, The transposed matrix of the current weight matrix, represents the data matrix to be sent input to the pre-equalization processing layer.

8. The pre-equalization based ultra-high order QAM modulation transmission system according to claim 1, characterized in that: The error signal is calculated according to the following formula: in, represents the error signal, Indicates the preset standard value. The data matrix representing the output of the pre-equalization layer.

9. A pre-equalization based ultra-high order QAM modulation transmission method using the system according to any one of claims 1 to 8, characterized in that: The steps of the method include: The data matrix to be transmitted is input into a transmitting end algorithm module, and the transmitting end algorithm module includes a pre-equalization processing layer. The pre-equalization processing layer processes the data matrix to be transmitted based on a weight matrix. After each processing of the data matrix to be transmitted, the weight matrix is ​​updated based on the number of zero taps in the weight matrix. The weight matrix is ​​updated based on the following formula: in, represents the updated weight matrix, represents the weight matrix before update, is the learning rate parameter, represents the error signal, represents the data matrix to be sent input to the pre-equalization processing layer, represents the regularization parameter, express The subdifferential of Represents the weight matrix before update norm; The data transmission module receives the data matrix output by the transmitting end algorithm module and outputs the data through an arbitrary wave generator and an IQ modulator; The data receiving module receives the data transmitted by the data transmitting module. The data receiving module includes a coherent receiver and an oscilloscope. The coherent receiver is connected to the IQ modulator via a transmission optical fiber. The incoming data is sequentially processed by the coherent receiver and the oscilloscope. The receiving-end algorithm module receives the data input by the oscilloscope, and obtains a receiving data matrix through the decoding layer analysis of the receiving-end algorithm module.

Images