Optical module bit error rate statistical method and system based on Gaussian fitting under PAM4

Abstract

The invention provides an optical module bit error rate statistical method and system based on Gaussian fitting under PAM4, and the method comprises the steps: collecting balanced sampling data on levels, updating a corresponding statistical histogram statistical result, calculating a mean value corresponding to each group of levels according to the statistical histogram statistical result, and setting a fitting distance; calculating a fitting mean value, calculating a fitting standard deviation of a single side, generating a Gaussian curve, and calculating a mean square error of the Gaussian curve and the statistical histogram curve in a fitting interval; updating the fitting distances, and calculating mean square errors under different fitting distances; selecting the fitting distance and the fitting standard deviation corresponding to the minimum mean square error as an optimal fitting result; fitting calculation is carried out on the level of each PAM4, a Q value is calculated, the bit error rate corresponding to each PAM4 is calculated, the average value is taken, the current bit error rate statistical value of the optical module is obtained, the Q average value is calculated, and then the SNR estimated value is calculated. According to the invention, the purpose of reducing the hardware implementation cost by fully utilizing idle resources of software is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication, in particular to a method and system for statistical analysis of bit error rate of optical module based on Gaussian fitting under PAM4. BACKGROUND

[0002] The conventional Q value estimation method widely used at present is based on the normal distribution assumption of each PAM data, that is, the received data is only under the influence of white noise, and the hist statistical result of the data shows the characteristics of Gaussian curve. However, the signal components in the actual link are more complex, in addition to signal and white noise, there are MPI interference, nonlinear interference of high and low temperature devices, etc., which will affect the distribution characteristics of the data, so that the data does not completely obey the normal distribution, resulting in a large estimation error between the estimated ber and the actual ber, and affecting the monitoring of the link performance.

[0003] Therefore, although the Q value estimation method widely used at present can reflect the performance status in the link to some extent, the error is large, the fluctuation is large due to the influence of interference, and the accurate and stable estimation performance cannot be guaranteed in complex application scenarios.

[0004] Patent document CN102577490B discloses a method for improving signal quality, a pre-error correction bit error rate feedback device and an optical transmission network equipment, relating to the field of communication. The solution is: when FEC decoding, the error correction bit rate of the data bits of the signal corresponding to each virtual channel is counted; the pre-error correction bit error rate of the physical channel corresponding to each virtual channel is calculated according to the error correction bit rate of the data bits corresponding to each virtual channel; the pre-error correction bit error rate of each physical channel is fed back to the optical module; the optical module adjusts the receiving parameters of each physical channel to reduce the pre-error correction bit error rate of each physical channel. However, the present application cannot guarantee accurate and stable estimation performance in complex application scenarios. SUMMARY

[0005] In view of the defects in the prior art, the purpose of the present application is to provide a method and system for statistical analysis of bit error rate of optical module based on Gaussian fitting under PAM4.

[0006] According to the method for statistical analysis of bit error rate of optical module based on Gaussian fitting under PAM4 provided by the present application, the following steps are included:

[0007] Step S1: collect the sampling data after level equalization and update the corresponding statistical histogram statistical result, calculate the mean value corresponding to each group of levels according to the statistical histogram statistical result, and set the fitting distance;

[0008] Step S2: calculate the fitting mean value according to the fitting distance and the initial mean value, calculate the one-sided fitting standard deviation, generate the Gaussian curve, and calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval.

[0009] Step S3: update the fitting distance, calculate the mean square error under different fitting distances; compare the mean square error under different fitting distances, select the fitting distance and fitting standard deviation corresponding to the minimum mean square error as the best fitting result; calculate the Q value for each PAM4 level fitting calculation:

[0010] Step S4: calculate the respective bit error rate according to the Q value and take the average to obtain the current bit error rate statistical value of the optical module, calculate the Q average according to the bit error rate statistical average, and then calculate the SNR estimation value.

[0011] Preferably, in the step S1:

[0012] The PAM4 signal technology uses four different signal levels for signal transmission. The software collects the equalized sample data on each PAM level of PAM4 as needed when there is idle scheduling resource. Each PAM level data is stored in the form of statistical histogram statistical results;

[0013] The equalized sample data is collected multiple times and the corresponding statistical histogram statistical results are updated until the statistical amount reaches the software threshold;

[0014] From the first PAM level, calculate the initial mean value μ i , set a fitting distance d, the initial value of d is 0; where i represents the i-th PAM4 level, i=0, 1, 2 or 3;

[0015]

[0016] hist is the statistical value of the number of sample values on each level value in the range of 0-255 levels, and j is the index of the hist statistical result.

[0017] Preferably, in the step S2:

[0018] Calculate the fitting mean value according to the fitting distance and the initial mean value Calculate the fitting mean value according to the fitting mean value And the statistical histogram statistical value to calculate the one-sided fitting standard deviation

[0019]

[0020]

[0021] Generate a Gaussian curve according to the fitting mean value and the fitting standard deviation, calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval, and the selection of the fitting interval is the tailing part of the two curves.

[0022] The Gaussian curve calculation formula is:

[0023]

[0024] The mean square error calculation method is:

[0025]

[0026] Wherein, thd is an integer in the interval of 0-255, configured by software.

[0027] Preferably, in the step S3:

[0028] The fitting distance is updated, the mean square error under different fitting distances is calculated, and the calculation is repeated until the number of times reaches the calculation number configured by the software;

[0029] The mean square errors under different fitting distances are compared, and the fitting distance corresponding to the minimum mean square error is selected And the fitting standard deviation As the best fitting result;

[0030] The electrical mean of each PAM4 is calculated, and after the fitting calculation of all 4 groups of PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance, and the fitting standard deviation:

[0031]

[0032] Preferably, in the step S4:

[0033] The 3 Q values are calculated according to the complementary error function erfc function to obtain the bit error rate of the optical module, and the mean value is taken to obtain the bit error rate statistical value of the optical module:

[0034]

[0035]

[0036]

[0037] Wherein, the coefficients a, b, and c are configured by software;

[0038] The Q mean value is calculated according to the bit error rate statistical mean value, and the SNR estimation value is calculated:

[0039]

[0040] SNR = 20log10(Q_m)

[0041] Wherein, erfcinv is the inverse function of erfc, and the positive function and inverse function are realized by table lookup, and Q_m is the mean value of Q value calculated by the three eyes of PAM4 signal.

[0042] According to the application, a PAM4-based optical module bit error rate statistical system based on Gaussian fitting is provided, comprising:

[0043] Module M1: Collects the sampling data after equalization at each level and updates the corresponding statistical histogram statistical result, calculates the mean value corresponding to each group of levels according to the statistical histogram statistical result, and sets the fitting distance.

[0044] Module M2: calculates the fitting mean value according to the fitting distance and the initial mean value, calculates the one-sided fitting standard deviation, generates the Gaussian curve, and calculates the mean square error of the Gaussian curve and the statistical histogram curve within the fitting interval.

[0045] Module M3: updates the fitting distance, calculates the mean square error under different fitting distances, compares the mean square errors under different fitting distances, selects the fitting distance and fitting standard deviation corresponding to the minimum mean square error as the best fitting result, performs fitting calculation for each PAM4 level, and calculates the Q value.

[0046] Module M4: calculates the bit error rate corresponding to each Q value and takes the mean value to obtain the current bit error rate statistical value of the optical module, calculates the Q mean value according to the bit error rate statistical mean value, and further calculates the SNR estimation value.

[0047] Preferably, in the module M1:

[0048] The PAM4 signal technology uses four different signal levels for signal transmission. The software collects the sampling data after equalization at each PAM4 level as needed when there is idle scheduling resource. Each PAM4 level data is stored in the form of statistical histogram statistical result.

[0049] The sampling data after equalization is collected multiple times and the corresponding statistical histogram statistical result is updated until the statistical amount reaches the software threshold.

[0050] From the first PAM level, the initial mean value μ i is calculated according to the hist statistical result, a fitting distance d is set, and the initial value of d is 0; wherein i represents the i-th PAM4 level, i = 0, 1, 2 or 3.

[0051]

[0052] hist is the statistical value of the number of sampling values at each level value within the range of 0-255 levels, and j is the index of the hist statistical result.

[0053] Preferably, in the module M2:

[0054] The fitting mean value is calculated according to the fitting distance and the initial mean value According to the fitting mean value And the statistical histogram statistical value, the one-sided fitting standard deviation is calculated

[0055]

[0056]

[0057] According to the fitting mean value and the fitting standard deviation, a Gaussian curve is generated, and the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval is calculated, and the selection of the fitting interval is the tailing part of the two curves;

[0058] The Gaussian curve calculation formula is:

[0059]

[0060] The mean square error calculation method is:

[0061]

[0062] Wherein, thd is an integer in the interval of 0-255, which is configured by software.

[0063] Preferably, in the module M3:

[0064] The fitting distance is updated, and the mean square error under different fitting distances is calculated until the number of repeated calculations reaches the number of calculations configured by the software;

[0065] The mean square errors under different fitting distances are compared, and the fitting distance corresponding to the minimum mean square error is selected And the fitting standard deviation As the best fitting result;

[0066] The electrical mean value of each PAM4 is calculated, and after the fitting calculation of all 4 groups of PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance and the fitting standard deviation:

[0067]

[0068] Preferably, in the module M4:

[0069] The 3 Q values calculate the respective corresponding bit error rate according to the complementary error function erfc function and take the mean value to obtain the current bit error rate statistical value of the optical module:

[0070]

[0071]

[0072]

[0073] wherein the coefficients a, b, c are configured by software;

[0074] According to the mean value of the bit error rate statistics, the Q mean value is calculated, and the SNR estimation value is calculated:

[0075]

[0076] SNR = 20log10(Q_m)

[0077] Where erfcinv is the inverse function of erfc, and the forward function and inverse function evaluation is realized by table lookup, and Q_m is the mean value of Q value calculated by the three eyes of PAM4 signal.

[0078] Compared with the prior art, the present application has the following beneficial effects:

[0079] 1. The present application realizes the purpose of reducing the hardware implementation cost by fully utilizing the idle software resources by adopting the software method to statistically calculate the ber under PAM4, and has higher flexibility;

[0080] 2. The present application improves the resistance to device nonlinear interference by adopting the sampling fitting Gaussian method to statistically calculate the ber, greatly reducing the estimation error caused by the software estimation method. BRIEF DESCRIPTION OF DRAWINGS

[0081] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments with reference to the attached drawings:

[0082] Figure 1 The present application is a flowchart. DETAILED DESCRIPTION

[0083] The present application will be described in detail below with reference to specific embodiments. The following embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any form. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, a number of changes and improvements can be made. These all belong to the protection scope of the present application.

[0084] Example 1:

[0085] According to the PAM4-based optical module bit error rate statistics method provided by the present application, as shown in Figure 1 , comprising:

[0086] Step S1: Collecting the sampling data after the equalization of the level and updating the corresponding statistical histogram statistics result, calculating the mean value corresponding to each group of levels according to the statistical histogram statistics result, and setting the fitting distance;

[0087] Specifically, in the step S1:

[0088] The PAM4 signal technology adopts 4 different signal levels for signal transmission, and the software collects the equalized sample data on each PAM level of PAM4 as needed when there is a free scheduling resource, and each PAM level data is stored in the form of statistical histogram statistical results;

[0089] The equalized sample data is collected multiple times and the corresponding statistical histogram statistical results are updated until the statistical quantity reaches the software threshold;

[0090] From the first PAM level, the initial mean μi is calculated according to the hist statistical results, respectively: i Set a fitting distance d, the initial value of d is 0; Where i represents the i-th PAM4 level, i = 0, 1, 2 or 3;

[0091]

[0092] The hist is the statistical value of the number of sample values on each level value in the range of 0-255 levels, and j is the index of the hist statistical result.

[0093] Step S2: Calculate the fitting mean value according to the fitting distance and the initial mean value, calculate the one-sided fitting standard deviation, generate the Gaussian curve, and calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval;

[0094] Specifically, in the step S2:

[0095] Calculate the fitting mean value according to the fitting distance and the initial mean value Calculate the one-sided fitting standard deviation according to the fitting mean value And the statistical histogram statistical value

[0096]

[0097]

[0098] Generate the Gaussian curve according to the fitting mean value and the fitting standard deviation, calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval, and the selection of the fitting interval is the tailing part of the two curves;

[0099] The Gaussian curve calculation formula is:

[0100]

[0101] The mean square error calculation method is:

[0102]

[0103] wherein thd is an integer in the interval of 0-255, configured by software to the parameter.

[0104] Step S3: update the fitting distance, calculate the mean square error under different fitting distances; compare the mean square error under different fitting distances, select the fitting distance and fitting standard deviation corresponding to the minimum mean square error as the best fitting result; calculate the Q value for each PAM4 level:

[0105] Specifically, in the step S3:

[0106] The fitting distance is updated, the mean square error under different fitting distances is calculated, and the calculation is repeated until the number of times reaches the number of calculations configured by software;

[0107] The mean square error under different fitting distances is compared, and the fitting distance corresponding to the minimum mean square error and the fitting standard deviation are selected as the best fitting result;

[0108] Each PAM4 level is calculated, and after the fitting calculation of all 4 groups of PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance, and the fitting standard deviation:

[0109]

[0110] Step S4: calculate the respective corresponding bit error rate according to the Q value and take the mean value to obtain the current bit error rate statistical value of the optical module, calculate the Q mean value according to the bit error rate statistical mean value, and then calculate the SNR estimation value.

[0111] Specifically, in the step S4:

[0112] 3 Q values are calculated according to the complementary error function erfc function to obtain the current bit error rate statistical value of the optical module:

[0113]

[0114]

[0115]

[0116] wherein the coefficients a, b, and c are configured by software;

[0117] The Q mean value is calculated according to the bit error rate statistical mean value, and the SNR estimation value is calculated:

[0118]

[0119] SNR = 20log10(Q_m)

[0120] where erfcinv is the inverse function of erfc, the forward function and inverse function evaluation is realized by table lookup, and Q_m is the average Q value calculated by the three eyes of the PAM4 signal.

[0121] Embodiment 2:

[0122] Embodiment 2 is a preferred example of Embodiment 1, to more specifically illustrate the present application.

[0123] The present application also provides a PAM4-based optical module error rate statistical system based on Gaussian fitting, which can be realized by executing the process steps of the PAM4-based optical module error rate statistical method based on Gaussian fitting, that is, the PAM4-based optical module error rate statistical method based on Gaussian fitting can be understood by those skilled in the art as the preferred embodiment of the PAM4-based optical module error rate statistical system based on Gaussian fitting.

[0124] According to the present application, a PAM4-based optical module error rate statistical system based on Gaussian fitting is provided, which comprises:

[0125] Module M1: Collecting the sampling data after equalization at each level and updating the corresponding statistical histogram statistical result, calculating the mean value corresponding to each group of levels according to the statistical histogram statistical result, and setting the fitting distance;

[0126] Specifically, in the module M1:

[0127] The PAM4 signal technology uses four different signal levels for signal transmission, and the software collects the sampling data after equalization at each PAM level of PAM4 as needed when there is idle scheduling resource, and each group of PAM level data is stored in the form of statistical histogram statistical result;

[0128] Collecting the sampling data after equalization multiple times and updating the corresponding statistical histogram statistical result until the statistical amount reaches the software threshold;

[0129] Starting from the first PAM level, the initial mean value μ i is calculated according to the hist statistical result, a fitting distance d is set, and the initial value of d is 0; wherein i represents the i-th PAM4 level, i=0, 1, 2 or 3;

[0130]

[0131] hist is the statistical value of the number of sampling values at each level value in the range of 0-255 levels, and j is the index of the hist statistical result.

[0132] Module M2: calculate the fitting mean value according to the fitting distance and the initial mean value, calculate the one-side fitting standard deviation, generate the Gaussian curve, and calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval;

[0133] Specifically, in the module M2:

[0134] calculate the fitting mean value according to the fitting distance and the initial mean value according to the fitting mean value and the statistical histogram statistical value, calculate the one-side fitting standard deviation

[0135]

[0136]

[0137] generate the Gaussian curve according to the fitting mean value and the fitting standard deviation, calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval, and the selection of the fitting interval is the tailing part of the two curves;

[0138] The Gaussian curve calculation formula is:

[0139]

[0140] The mean square error calculation method is:

[0141]

[0142] thd is an integer in the interval of 0-255, which is configured by software.

[0143] Module M3: update the fitting distance, calculate the mean square error under different fitting distances, compare the mean square errors under different fitting distances, select the fitting distance and the fitting standard deviation corresponding to the minimum mean square error as the best fitting result, and calculate the Q value for each PAM4 level:

[0144] Specifically, in the module M3:

[0145] update the fitting distance, calculate the mean square error under different fitting distances, and repeat the calculation until the number of repetitions reaches the calculation number configured by software;

[0146] compare the mean square errors under different fitting distances, and select the fitting distance corresponding to the minimum mean square error and the fitting standard deviation as the best fitting result;

[0147] Each PAM4 level is calculated, and after the fitting calculation of all four groups of PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance, and the fitting standard deviation.

[0148]

[0149] Module M4: Calculate the respective bit error rate according to the Q value and take the average to get the current bit error rate statistics of the optical module, calculate the Q average according to the bit error rate statistics average, and then calculate the SNR estimation value.

[0150] Specifically, in the module M4:

[0151] 3 Q values are calculated according to the complementary error function erfc function to calculate the respective bit error rate and take the average to get the current bit error rate statistics of the optical module:

[0152]

[0153]

[0154]

[0155] Wherein the coefficients a, b, c are configured by software;

[0156] Calculate the Q average according to the bit error rate statistics average, and calculate the SNR estimation value:

[0157]

[0158] SNR = 20log10(Q_m)

[0159] Where erfcinv is the inverse function of erfc, and the forward function and inverse function are evaluated by table lookup, and Q_m is the average of the Q values calculated by the PAM4 signal three eyes.

[0160] Example 3:

[0161] Example 3 is a preferred example of Example 1 to more specifically illustrate the present application.

[0162] Step 1: PAM4 signal technology is a modulation technology that uses 4 different signal levels for signal transmission. The software collects the equalized sample data on each PAM level of PAM4 as needed when there is idle scheduling resource. Each PAM level data is stored in the form of histogram (histogram, hereinafter referred to as hist) statistics results;

[0163] Step 2: Collect the equalized sample data multiple times and update the corresponding hist statistics results until the statistics reaches the software threshold;

[0164] Step 3: Calculate from the first PAM level: Calculate the initial mean μ i, set a fitting distance d, the initial value of d is 0. Where i represents the i-th PAM4 level, i = 0 / 1 / 2 / 3;

[0165]

[0166] Step 4: Calculate the fitting mean value according to the fitting distance and the initial mean value According to the fitting mean value And the hist statistical value, calculate the one-sided fitting standard deviation

[0167]

[0168]

[0169] Step 5: Generate a Gaussian curve according to the fitting mean value and the fitting standard deviation, and calculate the mean square error (MSE) of the Gaussian curve and the hist curve in the fitting interval. The selection of the fitting interval is the tailing part of the two curves;

[0170] The Gaussian curve calculation formula is:

[0171]

[0172] The MSE calculation method is:

[0173]

[0174] Step 6: Update the fitting distance, repeat steps 4 and 5, and calculate the MSE under different fitting distances until the number of repeated calculations reaches the software configured calculation number;

[0175] Step 7: Compare the MSE under different fitting distances, and select the fitting distance corresponding to the minimum MSE And the fitting standard deviation As the best fitting result.

[0176] Step 8: The mean value of each PAM4 is calculated by steps 3 to 7. After the fitting calculation of all 4 groups of PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance, and the fitting standard deviation:

[0177]

[0178] Step 6: Calculate the bit error ratio (bit error ratio, hereinafter referred to as ber) corresponding to each Q value according to the complementary error function (erfc function), and take the mean value to obtain the current ber statistical value of the optical module.

[0179]

[0180]

[0181]

[0182] Wherein the coefficients a / b / c can be configured by software.

[0183] Step 7: Calculate the Q mean value according to the ber statistical mean, and then calculate the SNR estimation value:

[0184]

[0185] SNR = 20log10(Q_m)

[0186] Wherein erfcinv is the inverse function of erfc, and the forward function and inverse function evaluation can be realized by table lookup.

[0187] The patent adopts the way of Gaussian fitting and tail fitting to recalculate the mean and standard deviation, and finds the optimal solution by iteration.

[0188] Those skilled in the art know that, in addition to implementing the system and each device, module and unit thereof provided by the present application in the form of pure computer readable program code, the system and each device, module and unit thereof provided by the present application can also be realized in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers and embedded microcontrollers to achieve the same function by logically programming the method steps. Therefore, the system and each device, module and unit thereof provided by the present application can be considered as a hardware component, and the devices, modules and units included therein for realizing various functions can also be considered as structures within the hardware component; the devices, modules and units for realizing various functions can also be considered as both software modules realizing the method and structures within the hardware component.

[0189] The specific embodiments of the present application are described above. It should be understood that the present application is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essential content of the present application. In the case of no conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for PAM4-based optical module bit error rate statistics based on Gaussian fitting, characterized in that, Comprising: Step S1: Collecting the sampling data after equalization on each level and updating the corresponding statistical histogram statistics, calculating the mean value corresponding to each group of levels according to the statistical histogram statistics, and setting the fitting distance; Step S2: Calculating the fitting mean value according to the fitting distance and the initial mean value, calculating the one-sided fitting standard deviation, generating the Gaussian curve, and calculating the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval; Step S3: Updating the fitting distance and calculating the mean square error under different fitting distances; comparing the mean square errors under different fitting distances, selecting the fitting distance and fitting standard deviation corresponding to the minimum mean square error as the best fitting result; calculating the Q value for each PAM4 level: Step S4: Calculating the respective bit error rate according to the Q value and taking the mean value to obtain the current bit error rate statistics of the optical module, calculating the Q mean value according to the bit error rate statistics mean value, and then calculating the SNR estimate value.

2. The PAM4 based optical module bit error rate statistics method based on Gaussian fitting of claim 1, wherein, In the step S1: The PAM4 signal technology uses four different signal levels for signal transmission. The software collects the sampling data after equalization on each PAM4 level as needed when there is idle scheduling resource. Each PAM level data is stored in the form of statistical histogram statistics; Collecting the sampling data after equalization multiple times and updating the corresponding statistical histogram statistics until the statistical quantity reaches the software threshold; Calculate from the first PAM level: Calculate the initial mean value μ according to the hist statistics result respectively i Set a fitting distance d, the initial value of d is 0; wherein i represents the i-th PAM4 level, i = 0, 1, 2 or 3; hist is the statistical value of the number of sampling values in the range of 0-255 levels for each level value, and j is the index of the hist statistics.

3. The PAM4 based optical module bit error rate statistics method based on Gaussian fitting of claim 1, wherein, In the step S2: Fitted mean value is calculated from the fitted distance and the initial mean value Fitted mean value is calculated from the fitted mean value One-sided fitted standard deviation is calculated from the statistical histogram statistics and the fitted mean value Generate the Gaussian curve according to the fitting mean value and the fitting standard deviation, calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval, and the selection of the fitting interval is the tailing part of the two curves; The Gaussian curve calculation formula is: The mean square error calculation method is: Where thd is an integer in the range of 0-255, which is configured by the software.

4. The PAM4 based optical module bit error rate statistics method based on Gaussian fitting of claim 1, wherein, In the step S3: Update the fitting distance and calculate the mean square error under different fitting distances until the number of repeated calculations reaches the number of calculations configured by the software; The fitting distance corresponding to the minimum mean square error is selected by comparing the mean square errors under different fitting distances and the fitting standard deviation as the best fitting result; Each PAM4 level is calculated, and after the fitting calculation of all four PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance, and the fitting standard deviation:

5. The PAM4 based optical module bit error rate statistics method based on Gaussian fitting of claim 1, wherein, In the step S4: The three Q values calculate their respective bit error rates according to the complementary error function erfc function and take the mean value to obtain the current bit error rate statistics of the optical module: Where the coefficients a, b, and c are configured by the software; Calculate the Q mean value according to the bit error rate statistics mean value to calculate the SNR estimate value: SNR = 20log10(Q_m) Where erfcinv is the inverse function of erfc, the positive function and inverse function evaluation are realized by table lookup, and Q_m is the mean value of the Q value calculated by the PAM4 signal three eyes.

6. A PAM4 based Gaussian fitting based optical module bit error rate statistics system, characterized in that, Comprising: Module M1: Collecting the sampling data after equalization on each level and updating the corresponding statistical histogram statistics, calculating the mean value corresponding to each group of levels according to the statistical histogram statistics, and setting the fitting distance; Module M2: Calculate the fitting mean value according to the fitting distance and the initial mean value, calculate the fitting standard deviation of one side, generate the Gaussian curve, and calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval; Module M3: Update the fitting distance, calculate the mean square error under different fitting distances, compare the mean square errors under different fitting distances, select the fitting distance and the fitting standard deviation corresponding to the minimum mean square error as the best fitting result, and calculate the Q value for each PAM4 level: Module M4: Calculate the bit error rate corresponding to each Q value and take the mean value to obtain the current bit error rate statistical value of the optical module, calculate the Q mean value according to the bit error rate statistical mean value, and then calculate the SNR estimation value.

7. The PAM4 below Gaussian fitting based optical module bit error rate statistics system of claim 6, wherein, In the module M1: The PAM4 signal technology uses four different signal levels for signal transmission. The software collects the equalized sample data on each PAM level of PAM4 as needed when there is idle scheduling resource. Each PAM level data is stored in the form of statistical histogram statistical results. Multiple equalized sample data are collected and the corresponding statistical histogram statistical results are updated until the statistical amount reaches the software threshold. Calculate from the first PAM level: Calculate the initial mean value μ according to the hist statistics result respectively i Set a fitting distance d, the initial value of d is 0; wherein i represents the i-th PAM4 level, i = 0, 1, 2 or 3; hist is the statistical value of the number of sample values on each level value in the range of 0-255, and j is the index of the hist statistical result.

8. The PAM4 below Gaussian fitting based optical module bit error rate statistics system of claim 6, wherein, In the module M2: Calculate fitted mean from fitted distance and initial mean Calculate fitted mean from fitted mean Calculate one-sided fitted standard deviation from fitted mean and statistical histogram statistics Generate a Gaussian curve according to the fitting mean value and the fitting standard deviation, calculate the mean square error of the Gaussian curve and the statistical histogram curve in the fitting interval, and the selection of the fitting interval is the tailing part of the two curves. The Gaussian curve calculation formula is: The mean square error calculation method is: Where thd is an integer in the range of 0-255, which is configured by software.

9. The PAM4 below Gaussian fitting based optical module bit error rate statistics system of claim 6, wherein, In the module M3: Update the fitting distance, calculate the mean square error under different fitting distances, and repeat the calculation until the number of calculations reaches the calculation number configured by the software. The fitting distance corresponding to the minimum mean square error is selected by comparing the mean square errors under different fitting distances and the fitting standard deviation as the best fitting result; Each PAM4 level is calculated, and after the fitting calculation of all four PAM levels is completed, the Q value is calculated according to the initial mean value, the fitting distance, and the fitting standard deviation:

10. The PAM4 below Gaussian fitting based optical module bit error rate statistics system of claim 6, wherein, In the module M4: Three Q values calculate the bit error rate corresponding to each Q value and take the mean value to obtain the current bit error rate statistical value of the optical module: Where the coefficients a, b, and c are configured by software; Calculate the Q mean value according to the bit error rate statistical mean value, and calculate the SNR estimation value: SNR = 20log10(Q_m) Where erfcinv is the inverse function of erfc, the positive function and inverse function are evaluated by table lookup, and Q_m is the mean value of the Q value calculated by the PAM4 signal three eyes.

Images