Parallel data transmission method based on optical chip

Abstract

The invention discloses a parallel data transmission method based on an optical chip, and the method comprises the following steps: S1, receiving external digital data through an FPGA module, and carrying out the format verification and preprocessing; s2, segmenting the data into a plurality of segments; s3, sending the fragments into a digital-to-analog converter, and converting the fragments into analog signals under the control of the FPGA; s4, analog signals obtained through conversion are output to the middle of decomposition and wavelength division multiplexing, modulation of optical signals is carried out, and an optical calculation chip carries out calculation processing on the modulated signals; s5, performing interference modulation on the analog signal in the optical chip to generate a coherent parallel optical signal; s6, transmitting the optical signal to a receiving end through an independent channel by means of an optical transmission medium, and keeping clock synchronization; s7, the receiving end collects optical signals through an optical detector, and the optical signals are converted into digital signals through an analog-to-digital converter; and S8, the FPGA module carries out recombination, verification and feedback control on the digital signal, and channel parameters are adjusted to complete data transmission.

Description

Technical Field

[0001] The present invention relates to the technical field of data transmission, and in particular to a parallel data transmission method based on an optical chip. Background Art

[0002] In the prior art, with the continuous development of data transmission technologies such as optical communication and optical transmission, various optical signal acquisition, analog-to-digital conversion, digital signal processing and other technologies have been widely used in modern communication systems. In traditional optical transmission systems, the receiving end mainly relies on optical detectors to collect optical signals transmitted by each channel, and then converts the collected optical signals into digital signals through analog-to-digital converters, and then uses digital processing circuits to process the digital signals. However, due to the problems of signal distortion, insufficient sampling accuracy and inaccurate data reconstruction in the process of optical signal acquisition and digital conversion in the prior art, data transmission errors, reduced transmission rates and untimely signal feedback control occur in data transmission. In addition, in the process of multi-channel data processing, the existing system often lacks an effective mechanism for data verification and feedback control, which makes the entire data transmission process have great uncertainty and is difficult to meet the requirements of data transmission. When using optical detectors to collect optical signals, the traditional technology usually uses the optical signal formula such as Although the basic characteristics of optical signals can be described in the quantization formula D = round ((Y(t)-Y min ) / (Y max -Y min )·(2 n-1)) Although it can theoretically ensure a certain conversion accuracy, data distortion may still occur during the signal processing due to insufficient sampling frequency or quantization error, which directly affects the effective parsing of digital signals by the data processing unit. At the same time, the existing technology lacks an effective real-time adjustment mechanism in data verification and feedback control. For the error information that appears during the data transmission process, it is impossible to correct it quickly and accurately, resulting in a lag in the adjustment of the working parameters of multiple channels and reducing the stability and reliability of the entire data transmission system. Especially when using the FPGA module for data recombination, verification, and feedback control, due to the lack of sufficient cooperation between the various modules in the processing flow and the imperfect design of the feedback control algorithm, it is difficult for the existing technology to achieve real-time control and precise adjustment of multi-channel data transmission in complex optical communication systems. Generally speaking, although the existing technology has made certain progress in optical signal acquisition, digital signal conversion, and data processing, there are still defects such as low sampling accuracy, inaccurate data recombination, simple verification methods, and untimely feedback control. These problems are particularly obvious in data transmission and real-time data processing applications, and there is an urgent need for a technical solution that can effectively solve the above problems. Therefore, how to provide a parallel data transmission method based on an optical chip is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0003] An object of the present invention is to propose a parallel data transmission method based on an optical chip. According to an embodiment of the present invention, a parallel data transmission method based on an optical chip includes the following steps:

[0004] S1. Use the FPGA module to receive the original digital data transmitted externally, and perform format verification and preprocessing on the received data;

[0005] S2. Divide the preprocessed digital data within the FPGA module, divide the data into several data segments, and each data segment serves as the basic unit for parallel transmission;

[0006] S3. Send each data segment into the corresponding digital-to-analog converter, and under the control of the FPGA, convert the digital signal into an analog signal according to the preset conversion parameters;

[0007] S4. Use the FPGA module to synchronously control multiple digital-to-analog converters, and simultaneously output the converted analog signals to the middle of decomposition and wavelength division multiplexing for optical signal modulation, and the optical computing chip performs computing processing on the modulated signals;

[0008] S5. Perform precise interference modulation on the received analog signal inside the optical chip, and generate parallel optical signals with coherent characteristics through a predetermined modulation program;

[0009] S6. Using a dedicated optical transmission medium, transmit the modulated multiple parallel optical signals to the receiving end through each independent channel to ensure the clock synchronization of each channel signal;

[0010] S7, at the receiving end, using a light detector to collect the optical signals transmitted by each channel, and converting the collected optical signals into digital signals through an analog-to-digital converter;

[0011] S8. Use the FPGA module to perform data reorganization, verification and real-time feedback control on the received digital signal, and adjust the working parameters of each parallel channel according to the feedback information to complete the entire data transmission process.

[0012] Optionally, the S1 specifically includes:

[0013] S11. Use FPGA module to receive the original digital data D input from the outside in , where D in ∈{0,1} N , N is the number of data bits;

[0014] S12, for the D in Perform data integrity check and calculate the checksum:

[0015]

[0016] Among them, d i Indicates D in The i-th data in the , C represents the checksum;

[0017] S13. Calculate the error index based on the checksum C:

[0018]

[0019] Among them, d ideal,i is the reference value of the i-th bit in the preset ideal data mode, and determines whether it satisfies E <E th , E th is the preset error threshold;

[0020] S14, will satisfy E <E th The original data of the condition D in Divided into qualified data D pass With unqualified data D fail ,in:

[0021] D pass ={D in |E <E th};

[0022] For qualified data D pass Call the preprocessing function f pre , get the preprocessed data:

[0023] D pre＝ f pre (D pass );

[0024] Among them, the preprocessing function includes a non-linear filtering operation, and its form is:

[0025]

[0026] α is a scaling parameter;

[0027] S15. Statistically normalize the preprocessed data D pre and calculate the mean value:

[0028]

[0029] and the standard deviation:

[0030]

[0031] and obtain the normalized data:

[0032]

[0033] S16. Adopt a dynamic segmentation algorithm according to the preset segmentation parameter, and evenly segment the normalized data D norm into M data segments, where:

[0034]

[0035] and the optimal segment length L is determined according to the data volatility, and the calculation formula is:

[0036] L = k1σ + k2;

[0037] where k1 and k2 are design constants;

[0038] S17. Perform time synchronization processing on each segmented data segment D i , i = 1, 2,..., M, and add a timestamp using the global synchronization clock T sync to form a time-tagged data segment:

[0039] D' i ={D i , T sync +Δt i};

[0040] where the segment delay Δt i is calculated according to the formula:

[0041] Δt i =β·i;

[0042] where β is a constant;

[0043] S18. Use all the data segments with time stamps as the output and transport them for digital-to-analog conversion and optical signal modulation:

[0044] {D'1, D'2,..., D' M}}.

[0045] Optionally, S2 specifically includes:

[0046] S21. Preprocess and format-convert the data:

[0047] D out ={d'1, d'2,..., d' M}};

[0048] Segment it according to the preset data segment length L and calculate the number of data segments

[0049] S22. For any data segment P i , i ∈ [1, k], define the expression as:

[0050] P i ={d' (i-1)L+1 , d' (i-1)L+2 ,..., d' iL}};

[0051] where d' j represents the data unit after format conversion;

[0052] S23. Combine all data segments into a set P and store it in the data segment buffer B inside the FPGA module:

[0053] P={P1, P2,..., P k}};

[0054] S24. Check whether the bit length of each data segment P i is exactly equal to the preset L bits.

[0055] Optionally, S3 specifically includes:

[0056] S31. Uniformly sample the input signal I to obtain a discrete sampling sequence I[n], where n ∈ {0, 1, 2,..., N–1} represents the sampling point index and the sampling interval is T s :

[0057] I[n]=I(n·T s );

[0058] S32. Use the preprocessing module F to perform noise suppression and signal enhancement processing on the sampled signal I[n] to obtain the preprocessed signal I o [n], and the mathematical expression is:

[0059] I o [n] = F(I[n]);

[0060] Among them, F(I[n]) can be expressed as a time-domain or frequency-domain filtering function, and the specific form is:

[0061] F(I[n]) = I[n] - ρ·B(I[n]);

[0062] Among them, B(I[n]) is the noise floor estimation function, and ρ is the noise suppression coefficient;

[0063] S33. Send the preprocessed signal I o [n] to the filtering module G for band-pass filtering to extract the target frequency band signal, and the filtering process satisfies:

[0064] I1[n] = G(I0[n]);

[0065] Among them, G(I0[n]) is defined as:

[0066] G(I0[n]) = ∑ k h[k]·I0[n - k];

[0067] Among them, h[k] is the impulse response of the designed band-pass filter, and the filter frequency response H(ω) reaches the maximum value within the target frequency band;

[0068] S34. To achieve adaptive adjustment of signal weighting, calculate the mean value μ[n] and variance σ 2 [n] of I1[n] within the local time window [n – L, n]:

[0069] Among them, L is the window length, and this step provides local statistical information for adaptive weighting;

[0070] S35. Combine the linear weighting parameter P1 and the non-linear weighting parameter P2, and combine the local statistic, introduce the innovative adaptive correction term β·σ 2 [n], and perform weighting processing on the filtered signal (I1[n]) to obtain the weighted signal I2[n], and the formula is:

[0071] I2[n] = P1·I1[n] + P2·(I1[n]) 2 + β·σ 2 [n];

[0072] Among them, β is an adaptive correction coefficient. Through this term, the weight can be automatically adjusted according to local signal fluctuations to achieve more accurate signal feature extraction;

[0073] S36. Normalize the weighted signal I2[n] using the normalization function H so that the normalized signal I3[n] satisfies a predetermined amplitude range. At the same time, combine a non-linear mapping to improve the signal-to-noise ratio. The process is expressed as:

[0074] I3[n] = H(I2[n]) = (I2[n] - min(I2)) / (max(I2) - min(I2));

[0075] Among them, min(I2) and max(I2) are respectively the minimum and maximum values of I2[n] within the processing window. In addition, to further optimize the signal dynamic range, a Sigmoid mapping can be used:

[0076] I3[n] = 1 / (1 + exp(-γ·(I2[n] - δ)));

[0077] Among them, γ and δ are parameters that adjust the steepness and offset of the mapping, and exp is the exponential function;

[0078] S37. Output the normalized signal I3[n] as the final processing result of step S43.

[0079] Optionally, the specific steps of S4 include:

[0080] S41. Use the FPGA module to simultaneously drive multiple digital-to-analog converters through a synchronous control signal, and adjust the analog signals output by each digital-to-analog converter to ensure the consistency of the signals in each channel;

[0081] S42. Transmit the analog signals output by each digital-to-analog converter to the intermediate link of decomposition and wavelength division multiplexing in the optical signal modulation module through a dedicated circuit, and perform signal preprocessing in this intermediate link;

[0082] S43. During the signal preprocessing process, use the electro-optic modulator in the optical chip to convert each input analog signal into a corresponding modulation signal, and modulate the optical signal by controlling the voltage signal. The voltage amplitude of the modulation signal is:

[0083]

[0084] Among them, I(t) is the modulated optical intensity, I0 is the input optical intensity, V(t) is the control voltage signal, and V pi is the half-wave voltage of the modulator;

[0085] S44. Further compute and process the modulated signal using an optical computing chip to ensure the coherent characteristics of each signal and optimize the signal quality;

[0086] S45. Adjust the phase, amplitude, and frequency of each signal through the algorithm in the optical computing chip to ensure that the modulated signal has stable characteristics during transmission:

[0087]

[0088] where, θ(t) is the optimized phase, A i is the amplitude of the i-th frequency component, ω i is the angular frequency of this frequency component, is the phase of this frequency component;

[0089] S46. Further control the output of the modulated signal for the optimized signal to achieve an optical signal meeting the transmission requirements and ensure the synchronization and coherence of each parallel optical signal.

[0090] Optionally, the specific steps of S5 include:

[0091] S51. Receive the preliminary processed signal R output in step S4, and extract the calibration signal P and the normalized signal N1:

[0092] P = α·I + β;

[0093] S52. Perform power function enhancement processing on the preliminary signal R:

[0094] R1 = R η1 ;

[0095] where, η1 is used as the gain parameter to determine the overall amplification degree of the signal, and R1 is the result of the enhancement of R;

[0096] S53. Construct a transfer function and derive the proportionality coefficient F:

[0097] F = exp(λ1·N1), lnκ = η1·ln R - λ1·N1;

[0098] S54. Perform dynamic calibration on the proportionality coefficient κ:

[0099] κ' = κ·[1 + δ1·(P - N1)];

[0100] S55. Perform non-linear mapping on the signal using the calibrated coefficient and calculate the intermediate signal C:

[0101] C = ln(1 + κ′·R η2 );

[0102] where, η2 is the non-linear enhancement index;

[0103] S56. Normalize signal C and perform weighted fusion with transfer function F to construct composite signal U:

[0104] W = θ·U+(1 - θ)·F;

[0105] S57. Perform difference correction and non - linear boundary compression on composite signal W and calibration signal P:

[0106] D = W+λ2·(P-(α·I + β)), D' = tanh(D);

[0107] Output the processing result D' as the final output of step S6.

[0108] Optionally, S6 specifically includes:

[0109] S61. Denote the multiple modulated parallel optical signals as S1, S2,..., S n1 , where n1 is the number of independent channels, and each signal satisfies the definition of the formula M i = f(X, Y i );

[0110] S62. Use a dedicated optical transmission medium to transmit each signal S i through independent optical transmission channels to the receiving end, and the transmission process satisfies the formula:

[0111] T i = H i (S i )+B i ;

[0112] Where, T i is the transmission signal of channel i, H i is the channel transfer function, and B i is the channel offset;

[0113] S63. Embed the clock synchronization signal C i :

[0114] C i = F sync (T i , t offset );

[0115] Where, T i is the transmission signal of channel i, t offset is the clock offset, F sync is the clock synchronization function, and it is required that each t offset = 0;

[0116] S64. Introduce the reference clock signal T ref , and use the formula:

[0117] ΔT i =T i -T ref ;

[0118] Among them, clock deviation correction is performed on each channel signal to ensure that ΔT i =0;

[0119] S65. Input the corrected signal T i and the synchronization signal C i into the demodulation unit together:

[0120] D i =G(T i , C i );

[0121] Among them, D i is the data signal recovered after demodulation of channel i, and G is the demodulation function.

[0122] Optionally, the S7 specifically includes:

[0123] S71. Use the photodetector PD at the receiving end to collect the optical signals I(t) transmitted by each channel. Among them, I(t) satisfies the formula:

[0124]

[0125] Among them, I0 represents the DC component, ΔI represents the amplitude change of the optical signal, ω is the angular frequency, is the initial phase;

[0126] S72. Perform preprocessing on the collected optical signal I(t), and use a filter. Its output signal Y(t) satisfies:

[0127] Y(t)=∫[I(t)·h(t)]dt;

[0128] Among them, h(t) is the impulse response function of the filter;

[0129] S73. Send the filtered optical signal Y(t) to the analog-to-digital converter ADC. The sampling frequency f s of the ADC satisfies f s >2f max , where f max is the highest frequency component contained in Y(t);

[0130] S74. The analog-to-digital converter is based on the quantization formula:

[0131] D = round((Y(t)-Ymin ) / (Y max -Y min )·(2 n -1));

[0132] Convert Y(t) into a digital signal D, where Y min and Y max are the minimum and maximum values of Y(t) respectively, and n is the resolution of the ADC;

[0133] S75. Transmit the quantized digital signal D to the data processing unit, and perform data decoding and information extraction.

[0134] Optionally, the S8 specifically includes:

[0135] S81. Input the digital signal D output by the analog-to-digital converter ADC into the FPGA module, and perform data recombination on D to generate a recombined data sequence S, where S = {S1, S2,..., S N};

[0136] S82. Adopt the check formula:

[0137]

[0138] where S i represents the i-th data item in the recombined data sequence S, M is a preset modulus, calculate the check code C, and mod represents the modulo operation, which is used to calculate the remainder after dividing two numbers;

[0139] S83. Compare the check code C with the pre-stored reference check code C ref to obtain the error information E, which is defined as:

[0140] E = |C - C ref |;

[0141] S84. Generate a feedback control signal F based on the error information E and the data feedback function. The expression of the feedback function is:

[0142] F(E) = K·E;

[0143] where K is the feedback coefficient;

[0144] S85. Adjust the working parameters P of each parallel channel according to the signal F, so that the new working parameters satisfy

[0145] P new = P + F(E);

[0146] where P is the original working parameter of each channel, thus completing the entire data transmission process.

[0147] The beneficial effects of the present invention are as follows:

[0148] By combining the high-precision acquisition of optical signals, the efficient analog-to-digital conversion of digital signals, and the precise operations of the FPGA module in data recombination, verification, and feedback control, the present invention significantly improves the performance and reliability of the entire optical communication system. During the process of using a photodetector to collect optical signals at the receiving end, by preprocessing and filtering the signals, the influence of external noise and interference can be effectively reduced, ensuring that the transmitted signals meet the requirements of the preset formula, thereby providing high-quality input signals for digital processing. This improvement in the initial signal acquisition stage overcomes the sampling error problem caused by unstable DC components and amplitude variations in traditional systems, contributing to improving the sampling accuracy and signal restoration degree of the entire system during data transmission.

[0149] The analog-to-digital converter ADC finely quantifies the preprocessed signals according to the quantization formula, achieving high-precision conversion from analog signals to digital signals. During this conversion process, not only is it ensured that the sampling frequency meets the requirements of the Nyquist sampling theorem, but also the data distortion problem caused by quantization errors is effectively reduced through the quantization formula. Compared with the inaccurate digital signals caused by insufficient sampling frequency or low quantization accuracy in traditional technologies, the method of the present invention can more truly restore the information of the original optical signal, ensuring the accuracy and reliability of data processing and laying a solid foundation for data transmission.

[0150] In terms of digital signal processing, the FPGA module undertakes the key tasks of data recombination, verification, and real-time feedback control. The FPGA precisely recombines the input digital signals, combines the preset verification formula and feedback control function, and real-time detects and corrects errors during data transmission. This process calculates the error information through the verification sum of the recombined data sequence and generates a feedback control signal based on the error information, ultimately realizing the dynamic adjustment of the working parameters of each parallel channel. This real-time feedback control mechanism effectively compensates for the problems of untimely feedback control and inaccurate data verification in traditional optical communication systems, enabling the system to quickly respond to external interference and internal error changes during data transmission, thereby greatly improving the stability and reliability of the overall system.

[0151] In addition, while implementing data transmission, the technical solution of the present invention also has good adaptability and scalability. By combining various technical elements in the claims, the organic integration of the whole process from signal acquisition, digital conversion to data processing is achieved, effectively avoiding the problem of insufficient cooperation between modules in traditional technologies. This solution can not only meet the requirements of multi-channel data transmission, but also maintain high-precision and high-reliability performance in complex communication environments. Thus, the present invention has achieved remarkable beneficial effects in improving data transmission rate, reducing bit error rate and realizing dynamic parameter regulation, providing a new and efficient technical solution for optical communication and related fields, and having high application value and promotion prospects. Description of the Drawings

[0152] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, but do not constitute a limitation to the present invention. In the drawings:

[0153] Figure 1 is a flowchart of a parallel data transmission method based on an optical chip proposed by the present invention;

[0154] Figure 2 is a schematic diagram of a parallel data transmission method based on an optical chip proposed by the present invention;

[0155] Figure 3 is a data flow diagram of a parallel data transmission method based on an optical chip proposed by the present invention. Detailed Description of the Embodiments

[0156] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only illustrating the basic structure of the present invention in a schematic manner, so they only show the components related to the present invention.

[0157] Refer to Figures 1-3 , a parallel data transmission method based on an optical chip, comprising the following steps:

[0158] S1. Use the FPGA module to receive the original digital data transmitted externally, and perform format verification and preprocessing on the received data;

[0159] S2. Divide the preprocessed digital data within the FPGA module, divide the data into several data segments, and each data segment serves as the basic unit for parallel transmission;

[0160] S3. Send each data segment into the corresponding digital-to-analog converter respectively, and under the control of the FPGA, convert the digital signal into an analog signal according to the preset conversion parameters;

[0161] S4. Use the FPGA module to synchronously control multiple digital-to-analog converters, and at the same time output the converted analog signals to the middle of decomposition and wavelength division multiplexing for optical signal modulation. The optical computing chip performs computing processing on the modulated signals;

[0162] S5. Perform precise interference modulation on the received analog signals inside the optical chip to generate parallel optical signals with coherent characteristics through a predetermined modulation program;

[0163] S6. Use a dedicated optical transmission medium to transmit the modulated multiple parallel optical signals to the receiving end through independent channels to ensure the clock synchronization of the signals in each channel;

[0164] S7. Use a photodetector at the receiving end to collect the optical signals transmitted in each channel, and convert the collected optical signals into digital signals through an analog-to-digital converter;

[0165] S8. Use the FPGA module to perform data recombination, verification, and real-time feedback control on the received digital signals, and adjust the working parameters of each parallel channel according to the feedback information to complete the entire data transmission process.

[0166] In this embodiment, the specific content of S1 includes:

[0167] S11. Use the FPGA module to receive the externally input original digital data D in , where D in ∈{0,1} N , and N is the number of data bits;

[0168] S12. Perform data integrity verification on the D in and calculate the checksum:

[0169]

[0170] where d i represents the i-th bit data in D in , and C represents the checksum;

[0171] S13. Calculate the error index according to the checksum C:

[0172]

[0173] where d ideal,i is the reference value of the i-th bit in the preset ideal data pattern, and determine whether it satisfies E < E th , and E th is the preset error threshold;

[0174] S14. Divide the original data D th that satisfies the condition of E < E in into qualified data Dpass with unqualified data D fail , where:

[0175] D pass ={D in |E < E th};

[0176] For the qualified data D pass call the preprocessing function f pre , to obtain the preprocessed data:

[0177] D pre＝ f pre (D pass );

[0178] Among them, the preprocessing function includes a non - linear filtering operation, and its form is:

[0179]

[0180] α is a scaling parameter;

[0181] S15. Statistically normalize the preprocessed data D pre and calculate the mean value:

[0182]

[0183] and the standard deviation:

[0184]

[0185] and obtain the normalized data:

[0186]

[0187] S16. According to the preset segmentation parameter, use the dynamic segmentation algorithm to evenly divide the normalized data D norm into M data segments, where:

[0188]

[0189] and the optimal segment length L is determined according to the data volatility, and the calculation formula is:

[0190] L = k1σ + k2;

[0191] where k1 and k2 are design constants;

[0192] S17. For each segmented data segment D i , i = 1, 2,..., M, perform time synchronization processing, and use the global synchronization clock T sync to add time stamps to form time - marked data segments:

[0193] D' i ={D i , T sync +Δt i};

[0194] Among them, the segmented delay Δt i According to the formula:

[0195] Δt i =β·i;

[0196] Among them, β is a constant;

[0197] S18. Take the set of all data segments with time stamps as the output and transport them for use in digital-to-analog conversion and optical signal modulation:

[0198] {D'1, D'2,..., D' M}.

[0199] In this embodiment, the S2 specifically includes:

[0200] S21. The preprocessed and format-converted data:

[0201] D out ={d'1, d'2,..., d' M};

[0202] Segment according to the preset data segment length L and calculate the number of data segments

[0203] S22. For any data segment P i , i ∈ [1, k], the defined expression is:

[0204] P i ={d' (i-1)L+1 , d' (i-1)L+2 ,..., d' iL};

[0205] Among them, d' j represents the data unit after format conversion;

[0206] S23. Combine all data segments into a set P and store it in the data segment buffer B inside the FPGA module:

[0207] P={P1, P2,..., P k};

[0208] S24. Check whether the bit length of each data segment P i is strictly equal to the preset L bits.

[0209] In this embodiment, S3 specifically includes:

[0210] S31. Uniformly sample the input signal I to obtain a discrete sampling sequence I[n], where n ∈ {0, 1, 2, …, N–1} represents the sampling point index, and the sampling interval is T s :

[0211] I[n] = I(n·T s );

[0212] S32. Use the preprocessing module F to perform noise suppression and signal enhancement processing on the sampled signal I[n] to obtain a preprocessed signal I o [n], and the mathematical expression is:

[0213] I o [n] = F(I[n]);

[0214] Among them, F(I[n]) can be expressed as a time-domain or frequency-domain filtering function, and the specific form is:

[0215] F(I[n]) = I[n] - ρ·B(I[n]);

[0216] Among them, B(I[n]) is a noise floor estimation function, and ρ is a noise suppression coefficient;

[0217] S33. Send the preprocessed signal I o [n] to the filtering module G for band-pass filtering to extract the target frequency band signal, and the filtering process satisfies:

[0218] I1[n] = G(I0[n]);

[0219] Among them, G(I0[n]) is defined as:

[0220] G(I0[n]) = ∑ k h[k]·I0[n - k];

[0221] Among them, h[k] is the impulse response of the designed band-pass filter, and the filter frequency response H(ω) reaches the maximum value within the target frequency band;

[0222] S34. To achieve adaptive adjustment of signal weighting, calculate the mean value μ[n] and variance σ of I1[n] within the local time window [n–L, n] 2 [n]:

[0223] Among them, L is the window length, and this step provides local statistical information for adaptive weighting;

[0224] S35. Incorporate the linear weighting parameter P1 and the non-linear weighting parameter P2, and combine the local statistics, and introduce an innovative adaptive correction term β·σ2 [n], the filtered signal (I1[n]) is weighted to obtain a weighted signal I2[n], and the formula is:

[0225] I2[n] = P1·I1[n] + P2·(I1[n]) 2 +β·σ 2 [n];

[0226] Among them, β is an adaptive correction coefficient. Through this term, the weight can be automatically adjusted according to local signal fluctuations to achieve more accurate signal feature extraction;

[0227] S36. Normalize the weighted signal I2[n]. Using the normalization function H, make the normalized signal I3[n] satisfy the predetermined amplitude range, and at the same time combine non-linear mapping to improve the signal-to-noise ratio. The process is expressed as:

[0228] I3[n] = H(I2[n]) = (I2[n] - min(I2)) / (max(I2) - min(I2));

[0229] Among them, min(I2) and max(I2) are the minimum and maximum values of I2[n] within the processing window respectively; In addition, to further optimize the signal dynamic range, Sigmoid mapping can be used:

[0230] I3[n] = 1 / (1 + exp(-γ·(I2[n] - δ)));

[0231] Among them, γ and δ are parameters for adjusting the mapping steepness and offset, and exp is the exponential function;

[0232] S37. Output the normalized signal I3[n] as the final processing result of step S43.

[0233] In this embodiment, the said S4 specifically includes:

[0234] S41. Use the FPGA module to simultaneously drive multiple digital-to-analog converters through a synchronous control signal, and adjust the analog signals output by each digital-to-analog converter to ensure the consistency of the signals in each channel;

[0235] S42. Transmit the analog signals output by each digital-to-analog converter to the intermediate link of decomposition and wavelength division multiplexing in the optical signal modulation module through a dedicated circuit, and perform signal preprocessing in this intermediate link;

[0236] S43. During the signal preprocessing process, use the electro-optic modulator in the optical chip to convert each input analog signal into a corresponding modulated signal, and modulate the optical signal by controlling the voltage signal. The voltage amplitude of the modulated signal is:

[0237]

[0238] Among them, I(t) is the modulated optical intensity, I0 is the input optical intensity, V(t) is the control voltage signal, and V pi is the half-wave voltage of the modulator;

[0239] S44. Use the optical computing chip to further calculate and process the modulated signal to ensure the coherent characteristics of each signal and optimize the signal quality;

[0240] S45. Through the algorithm in the optical computing chip, adjust the phase, amplitude, and frequency of each signal to ensure that the modulated signal has stable characteristics during transmission:

[0241]

[0242] Among them, θ(t) is the optimized phase, A i is the amplitude of the i-th frequency component, ω i is the angular frequency of this frequency component, is the phase of this frequency component;

[0243] S46. Further control the output of the modulated signal with the optimized signal to achieve an optical signal that meets the transmission requirements, and ensure the synchronization and coherence of each parallel optical signal.

[0244] In this embodiment, the specific content of S5 includes:

[0245] S51. Receive the preliminary processing signal R output in step S4, and extract the correction signal P and the normalized signal N1:

[0246] P = α·I + β;

[0247] S52. Perform power function enhancement processing on the preliminary signal R:

[0248] R1 = R η1 ;

[0249] Among them, η1 is used as the gain parameter to determine the overall amplification degree of the signal, and R1 is the result of enhancing R;

[0250] S53. Construct a conversion function and deduce the proportionality coefficient F:

[0251] F = exp(λ1·N1), lnκ = η1·ln R - λ1·N1;

[0252] S54. Perform dynamic correction on the proportionality coefficient κ:

[0253] κ' = κ·[1 + δ1·(P - N1)];

[0254] S55. Use the corrected coefficient to perform non - linear mapping on the signal and calculate the intermediate signal C:

[0255] C = ln(1 + κ′·R η2 );

[0256] where η2 is the non - linear enhancement index;

[0257] S56. Normalize the signal C and perform weighted fusion with the transfer function F to construct the composite signal U:

[0258] W = θ·U+(1 - θ)·F;

[0259] S57. Perform difference correction and non - linear boundary compression on the composite signal W and the correction signal P:

[0260] D = W+λ2·(P-(α·I + β)), D' = tanh(D);

[0261] Output the processing result D′ as the final output of step S6.

[0262] In this embodiment, step S6 specifically includes:

[0263] S61. Denote the modulated multiple parallel optical signals as S1, S2,..., S n1 , where n1 is the number of independent channels, and each signal satisfies the definition of the formula M i = f(X, Y i );

[0264] S62. Use a dedicated optical transmission medium to transmit each signal S i through independent optical transmission channels to the receiving end, and the transmission process satisfies the formula:

[0265] T i = H i (S i )+B i ;

[0266] where T i is the transmission signal of channel i, H i is the channel transfer function, and B i is the channel bias;

[0267] S63. Embed the clock synchronization signal C i :

[0268] C i = F sync (T i , t offset );

[0269] Among them, T i is the transmission signal of channel i, t offset is the clock offset, F sync is the clock synchronization function, and it is required that each t offset = 0;

[0270] S64. Introduce a reference clock signal T ref , and use the formula:

[0271] ΔT i = T i - T ref ;

[0272] Among them, perform clock deviation correction on each channel signal to ensure that ΔT i = 0;

[0273] S65. Input the corrected signal T i and the synchronization signal C i into the demodulation unit together:

[0274] D i = G(T i , C i );

[0275] Among them, D i is the data signal recovered after demodulation of channel i, and G is the demodulation function.

[0276] In this embodiment, the S7 specifically includes:

[0277] S71. At the receiving end, use the photodetector PD to collect the optical signals I(t) transmitted by each channel. Among them, I(t) satisfies the formula:

[0278]

[0279] Among them, I0 represents the DC component, ΔI represents the amplitude change of the optical signal, ω is the angular frequency, is the initial phase;

[0280] S72. Preprocess the collected optical signal I(t), and use a filter. Its output signal Y(t) satisfies:

[0281] Y(t)=∫[I(t)·h(t)]dt;

[0282] Among them, h(t) is the impulse response function of the filter;

[0283] S73. Send the filtered optical signal Y(t) into the analog-to-digital converter ADC. The sampling frequency f s of the ADC satisfies f s > 2fmax , where f max is the highest frequency component contained in Y(t);

[0284] S74. The analog-to-digital converter converts Y(t) into a digital signal D according to the quantization formula:

[0285] D = round((Y(t) - Y min ) / (Y max - Y min )·(2 n - 1));

[0286] where Y min and Y max are the minimum and maximum values of Y(t) respectively, and n is the resolution of the ADC;

[0287] S75. Transmit the quantized digital signal D to the data processing unit, and perform data decoding and information extraction.

[0288] In this embodiment, the S8 specifically includes:

[0289] S81. Input the digital signal D output by the analog-to-digital converter ADC into the FPGA module, and perform data recombination on D to generate a recombined data sequence S, where S = {S1, S2,..., S N};

[0290] S82. Use the check formula:

[0291]

[0292] where S i represents the i-th data item in the recombined data sequence S, M is a preset modulus, calculate the check code C, and mod represents the modulo operation, which is used to calculate the remainder after two numbers are divided;

[0293] S83. Compare the check code C with the pre-stored reference check code C ref to obtain the error information E, which is defined as:

[0294] E = |C - C ref |;

[0295] S84. Generate a feedback control signal F based on the error information E and the data feedback function. The expression of the feedback function is:

[0296] F(E) = K·E;

[0297] where K is the feedback coefficient;

[0298] S85. Adjust the working parameters P of each parallel channel according to the signal F, so that the new working parameters satisfy

[0299] P new = P + F(E);

[0300] Wherein, P is the original working parameter of each channel, thus completing the entire data transmission process.

[0301] Embodiment 1:

[0302] In order to verify the feasibility of the present invention in implementation, the present invention is applied in the data center of a large optical communication operation company. To solve problems such as sampling error, inaccurate data recombination, and feedback control delay during data transmission, the company selects the data center located in Pudong New Area, Shanghai as the experimental scenario. There are numerous optical communication links of various types in the data center. During long-term operation, the traditional system frequently experiences data transmission distortion and error codes, affecting the overall stability of the system and the data transmission rate. Therefore, the experimental team introduces the optical signal acquisition, analog-to-digital conversion, and real-time verification and feedback control technologies of the FPGA module in the present invention into the actual scenario to transform and upgrade the original system. During the experiment, the system samples and processes the optical signals from each channel within 48 consecutive hours of operation, and maintains 24-hour monitoring records throughout the process. After the transformation, the system processes the acquired optical signals using a pre-designed filtering algorithm to ensure that the signals input to the analog-to-digital converter reach the ideal sampling quality; then the signals are converted into digital signals through a high-sampling-rate ADC module, and data recombination and verification are performed in the FPGA module to generate real-time feedback control signals to adjust the working parameters of each channel, thereby achieving dynamic compensation. Significant improvements have been achieved in each link of the system's sampling frequency, quantization accuracy, data verification, and feedback control.

[0303] In the actual application process, experimental data shows that the improved system has increased the sampling accuracy by about 15%, reduced the quantization error by about 20%, and the data verification error rate has been reduced from the original 5% to less than 1%. In addition, the delay of the entire data transmission process has been reduced from the traditional average delay of 20 milliseconds to less than 8 milliseconds, and the data transmission rate has increased by about 30%. To verify the feasibility of the present invention in implementation, the present invention is applied to the renovation project of an optical communication link in a large data center. According to the experimental data, during the period from March 15th to March 17th, 2024, the experimental site was the data center in Pudong New Area, Shanghai, and the key data recorded are as follows: the sampling frequency has been increased from the original 200 MS / s to 250 MS / s; the analog-to-digital conversion accuracy has been increased from the original 12 bits to 14 bits; the signal transmission delay has been reduced from the average of 20 milliseconds in the traditional system to the average of 7.8 milliseconds in the solution of the present invention; the verification error rate has been reduced from the original 5.3% to 0.9%; the data transmission rate has been increased from the original 1.2 Gbps to 1.56 Gbps. Through a large amount of data collection and statistical analysis, the experimental results show that the solution of the present invention can effectively alleviate the deficiencies of the traditional optical communication system in multi-channel data transmission, ensure the high precision and high reliability of the data transmission process, and at the same time achieve real-time feedback control.

[0304] To more intuitively display the improvement effect, a detailed data table was compiled in the experiment. This data table details the performance of each index in the system before and after the transformation during the experiment. The data table includes key parameters such as sampling frequency, analog-to-digital conversion accuracy, signal transmission delay, verification error rate, and data transmission rate. The data are all from actual monitoring records and are averaged after multiple experimental statistics. The data in the data table are true and reliable, and can comprehensively reflect the significant advantages of the present invention in solving problems such as sampling, conversion, data recombination, and feedback control in data transmission.

[0305]

[0306]

[0307] Table 1 Comparison Table of Improvement Effects of the Parallel Data Transmission Method Based on Optical Chips

[0308] From the above data, it can be seen that after adopting the solution of the present invention, each index has been significantly improved. Especially in terms of signal transmission delay and verification error rate, the decline is obvious, fundamentally solving the problem of unstable data transmission in traditional optical communication technology.

[0309] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, should be covered by the protection scope of the present invention.

Claims

1. A parallel data transmission method based on an optical chip, characterized in that, It includes the following steps: S1. Use the FPGA module to receive the original digital data transmitted externally, and perform format verification and preprocessing on the received data; S2. Divide the preprocessed digital data within the FPGA module, divide the data into several data segments, and each data segment serves as the basic unit for parallel transmission; S3. Send each data segment into the corresponding digital-to-analog converter respectively, and under the control of the FPGA, convert the digital signal into an analog signal according to the preset conversion parameters; S4. Use the FPGA module to synchronously control multiple digital-to-analog converters, and simultaneously output the converted analog signals to the middle of decomposition and wavelength division multiplexing for optical signal modulation, and the optical computing chip performs computing processing on the modulated signals; S5. Perform precise interference modulation on the received analog signals inside the optical chip, and generate parallel optical signals with coherent characteristics through a predetermined modulation program; S6. Use a dedicated optical transmission medium to transmit the modulated multiple parallel optical signals to the receiving end through each independent channel, and ensure the clock synchronization of the signals in each channel; S7. Use an optical detector at the receiving end to collect the optical signals transmitted through each channel, and convert the collected optical signals into digital signals through an analog-to-digital converter; S8. Use the FPGA module to perform data recombination, verification and real-time feedback control on the received digital signals, and adjust the working parameters of each parallel channel according to the feedback information to complete the entire data transmission process.

2. The parallel data transmission method based on an optical chip according to claim 1, wherein The S1 includes the following specific operations: S11. Use the FPGA module to receive the original digital data D input externally in , where D in ∈{0,1} N , and N is the number of data bits; S12. Perform data integrity verification on the said D in and calculate the checksum: where d i represents the i-th bit data in D in , and C represents the checksum; S13. Calculate the error index according to the checksum C: Among them, d ideal,i is the reference value of the i-th bit in the preset ideal data pattern, and it is judged whether E < E th , E th is the preset error threshold; S14. Divide the original data D that meets the condition E < E th into qualified data D in and unqualified data D pass , where: fail ​ D pass = {D in | E < E th}; For qualified data D pass Call the preprocessing function f pre , and obtain the preprocessed data: D pre＝ f pre (D pass ); Among them, the preprocessing function includes a non-linear filtering operation, and its form is: α is a scaling parameter; S15. Perform statistical normalization on the preprocessed data D pre and calculate the mean value: and the standard deviation: and obtain the normalized data: S16. Using a dynamic segmentation algorithm according to preset segmentation parameters, the normalized data D norm is evenly segmented into M data segments, where: And the optimal segment length L is determined according to the data volatility, and the calculation formula is: L = k1σ + k2; Among them, k1 and k2 are design constants; S17. Perform time synchronization processing on each segmented data segment D i , where i = 1, 2,..., M, and use the global synchronization clock T sync to add time stamps to form time-stamped data segments: D′ i ={D i , T sync +Δt i}; Among them, the segmented delay Δt i According to the formula: Δt i = β·i; Among them, β is a constant; S18. Take the set of all data segments with time stamps as the output and send them to the next step for digital-to-analog conversion and optical signal modulation: {D′1, D′2,..., D′ M}。 3. The parallel data transmission method based on an optical chip according to claim 1, wherein The S2 step includes the following steps: S21. The data after preprocessing and format conversion: D out = {d′1, d′2,..., d′ M}; Divide according to the preset data segment length L and calculate the number of data segments S22. For any data segment P i , where i ∈ [1, k], the expression is defined as: P i = {d′ (i-1)L+1 , d′ (i-1)L+2 ,..., d′ iL}; where d' j represents the data unit after format conversion; S23. Form a set P of all data segments and store them in the data segment buffer B inside the FPGA module: P = {P1, P2,..., P k}; S24. Verify whether the bit length of each data segment P i is strictly equal to the preset L bits.

4. The parallel data transmission method based on an optical chip according to claim 1, wherein The S3 includes the following specific operations: S31. Uniformly sample the input signal I to obtain a discrete sampling sequence I[n], where n ∈ {0, 1, 2, …, N–1} represents the sampling point index and the sampling interval is T s : I[n] = I(n·T s ); S32. Use the preprocessing module F to perform noise suppression and signal enhancement processing on the sampled signal I[n] to obtain the preprocessed signal I o [n], and the mathematical expression is: I o [n]=F(I[n]); Among them, F(I[n]) can be expressed as a time-domain or frequency-domain filtering function, and the specific form is: F(I[n]) = I[n] - ρ·B(I[n]); Among them, B(I[n]) is a noise floor estimation function, and ρ is a noise suppression coefficient; S33. Feed the preprocessed signal I o [n] into the filtering module G for band-pass filtering to extract the target frequency band signal, and the filtering process satisfies: I1[n] = G(I0[n]); Among them, G(I0[n]) is defined as: G(I0[n]) = ∑ k h[k]·I0[n - k]; Among them, h[k] is the impulse response of the designed band-pass filter, and the filter frequency response H(ω) reaches the maximum value within the target frequency band; S34. To achieve adaptive adjustment of signal weighting, calculate the mean μ[n] and variance σ 2 [n] of I1[n] within the local time window [n–L, n]: Among them, L is the window length, and this step provides local statistical information for adaptive weighting; S35. Introduce the innovative adaptive correction term β·σ 2 [n], combine the linear weighting parameter P1 and the non-linear weighting parameter P2 with the local statistic, and perform weighted processing on the filtered signal (I1[n]) to obtain the weighted signal I2[n]. The formula is as follows: I2[n] = P1·I1[n] + P2·(I1[n]) 2 + β·σ 2 [n]; Among them, β is an adaptive correction coefficient, and through this item, the weight can be automatically adjusted according to the local signal fluctuation to achieve more accurate signal feature extraction; S36. Normalize the weighted signal I2[n] using the normalization function H such that the normalized signal I3[n] satisfies a predetermined amplitude range, and at the same time, combine a non - linear mapping to improve the signal - to - noise ratio. The process is expressed as: I3[n]=H(I2[n])=(I2[n] - min(I2)) / (max(I2)-min(I2)); where min(I2) and max(I2) are the minimum and maximum values of I2[n] within the processing window respectively. In addition, to further optimize the signal dynamic range, a Sigmoid mapping can be used: I3[n]=1 / (1 + exp(-γ·(I2[n]-δ))); where γ and δ are parameters for adjusting the mapping steepness and offset, and exp is the exponential function; S37. Output the normalized signal I3[n] as the final processing result of step S43.

5. The parallel data transmission method based on an optical chip according to claim 1, wherein The said S4 includes the following specific operations: S41. Use the FPGA module to simultaneously drive multiple digital - to - analog converters through a synchronous control signal, and adjust the analog signals output by each digital - to - analog converter to ensure the consistency of the signals in each channel; S42. Transmit the analog signal output by each digital - to - analog converter through a dedicated circuit to the intermediate link of decomposition and wavelength division multiplexing in the optical signal modulation module, and perform signal pre - processing in this intermediate link; S43. In the process of signal pre - processing, use the electro - optical modulator in the optical chip to convert each input analog signal into a corresponding modulation signal, and modulate the optical signal by controlling the voltage signal. The voltage amplitude of the modulation signal is: Among them, I(t) is the modulated optical intensity, I0 is the input optical intensity, V(t) is the control voltage signal, and V pi is the half-wave voltage of the modulator; S44. Use the optical computing chip to further calculate and process the modulated signal to ensure the coherent characteristics of each signal and optimize the signal quality; S45. Through the algorithm in the optical computing chip, adjust the phase, amplitude and frequency of each signal to ensure that the modulation signal has stable characteristics during transmission: Among them, θ(t) is the optimized phase, and A i is the amplitude of the i-th frequency component, ω i is the angular frequency of this frequency component, and is the phase of this frequency component; S46. Further control the output of the modulation signal for the optimized signal to achieve an optical signal that meets the transmission requirements, and ensure the synchronization and coherence of each parallel optical signal.

6. The parallel data transmission method based on an optical chip according to claim 1, wherein The said S5 includes the following specific operations: S51. Receive the preliminary processing signal R output by step S4, and extract the calibration signal P and the normalized signal N1: P = α·I+β; S52. Perform power - function enhancement processing on the preliminary signal R: R1 = R η1 ; where η1 is the gain parameter, which determines the overall amplification degree of the signal, and R1 is the result of enhancing R; S53. Construct a transfer function and derive the proportionality coefficient F: S54. Perform dynamic calibration on the proportionality coefficient κ: κ′=κ·[1 + δ1·(P - N1)]; S55. Use the calibrated coefficient to perform non - linear mapping on the signal and calculate the intermediate signal C: C = ln(1 + κ′·R η2 ); where η2 is the non - linear enhancement exponent; S56. Normalize the signal C and perform weighted fusion with the transfer function F to construct the composite signal U: S57. Perform difference calibration and non - linear boundary compression on the composite signal W and the calibration signal P: D = W+λ2·(P-(α·I+β)), D′=tanh(D); Output the processing result D′ as the final output of step S6.

7. The parallel data transmission method based on an optical chip according to claim 1, wherein The said S6 includes the following specific operations: S61. Denote the multiple modulated parallel optical signals as S1, S2, …, Sn1 respectively, where n1 is the number of independent channels, and each signal satisfies the definition of the formula M n1 = f(X, Y i ); i ) S62. Use a dedicated optical transmission medium to transmit each signal S i to the receiving end through independent optical transmission channels respectively, and the transmission process satisfies the formula: T i = H i (S i ) + B i ; Among them, T i is the transmission signal of channel i, H i is the channel transfer function, B i is the channel bias; S63. Embed a clock synchronization signal C in each independent channel i : C i = F sync (T i , t offset ); Among them, T i is the transmission signal of channel i, t offset is the clock offset, F sync is the clock synchronization function, and it is required that each t offset = 0; S64. Introduce a reference clock signal T ref , and use the formula: ΔT i = T i - T ref ; Among them, clock deviation correction is performed on each channel signal to ensure that ΔT i = 0; S65. Input the corrected signal T i and the synchronization signal C i into the demodulation unit together: D i = G(T i , C i ); Among them, D i is the data signal recovered after demodulation of channel i, and G is the demodulation function.

8. The parallel data transmission method based on an optical chip according to claim 1, wherein The said S7 includes the following specific operations: S71. At the receiving end, use the optical detector PD to collect the optical signals I(t) transmitted by each channel, where I(t) satisfies the formula: where, I0 represents the DC component, ΔI represents the amplitude change of the optical signal, ω is the angular frequency, is the initial phase; S72. Preprocess the collected optical signals I(t) and use a filter, and its output signal Y(t) satisfies: Y(t) = ∫[I(t)·h(t)]dt; where h(t) is the impulse response function of the filter; S73. Feed the filtered optical signal Y(t) into an analog-to-digital converter ADC, and the sampling frequency f of the ADC s satisfies f s > 2f max , where f max is the highest frequency component contained in Y(t); S74. The analog-to-digital converter is based on the quantization formula: D = round((Y(t) - Y min ) / (Y max - Y min ) · (2 n - 1)); Convert Y(t) into a digital signal D, where Y min and Y max are the minimum and maximum values of Y(t) respectively, and n is the resolution of the ADC; S75. Transmit the quantized digital signal D to the data processing unit and perform data decoding and information extraction.

9. The parallel data transmission method based on an optical chip according to claim 1, wherein The said S8 includes the following specific operations: S81. Input the digital signal D output by the analog-to-digital converter ADC into the FPGA module, and perform data recombination on D to generate a recombined data sequence S, where S = {S1, S2, …, S N}; S82. Adopt the verification formula: where S i represents the i-th data item in the reconstructed data sequence S, M is a preset modulus, the check code C is calculated, and mod represents the modulo operation, which is used to calculate the remainder after dividing two numbers; S83. Compare the check code C with the pre-stored reference check code C ref to obtain error information E, which is defined as: E = |C - C ref |; S84. Generate a feedback control signal F based on the error information E and the data feedback function, and the feedback function expression is: F(E) = K·E; where K is the feedback coefficient; S85. Adjust the working parameters P of each parallel channel according to the signal F, and make the new working parameters satisfy P new = P + F(E); where P is the original working parameter of each channel, thus completing the entire data transmission process.

Images