Sampling frequency deviation compensation method for optical fiber communication system

Abstract

The invention belongs to the technical field of optical fiber communication, and particularly relates to an optical fiber communication system sampling frequency deviation compensation method. Sampling frequency deviation caused by ADC and DAC in a high-speed optical fiber communication system is compensated, and the method specifically comprises sampling frequency deviation compensation and signal optimal sampling point calibration, specifically, sampling frequency deviation is compensated in a filter equalization mode, and sampling time points of signals subjected to sampling frequency deviation compensation are rapidly corrected through a frequency domain time shifting method; the method is easy to implement, low in complexity and high in flexibility. The method is realized in digital signal processing, so that the flexibility of the system can be greatly improved.

Description

Technical Field

[0001] The invention belongs to the technical field of optical fiber communication, and in particular relates to a method for compensating sampling frequency offset of an optical fiber communication system. Background Art

[0002] With the development of technology, fiber-optic communication systems have become the most widely used, longest-distance, and largest-capacity wired communication solutions. In order to meet the growing demand, the rate and capacity of fiber-optic communication systems are also continuously improving, and increasing the symbol rate of the signal has become the most direct means to increase capacity. However, the higher symbol rate requires that the sampling rate of the system's digital-to-analog conversion and analog-to-digital conversion must also reach a corresponding level, usually requiring a sampling rate of tens of GHz or even hundreds of GHz.

[0003] However, due to the limitations of physical devices, the actual sampling rate of digital-to-analog and analog-to-digital conversion often deviates from the nominal sampling rate. When the sampling rate reaches tens of GHz or even hundreds of GHz, the impact of this frequency offset on signal recovery is extremely significant. Therefore, if the problem of sampling frequency offset can be effectively solved, it will provide important guarantees for high-baud-rate high-speed optical fiber communication systems and play a key role in future 6G and 7G communications. Summary of the invention

[0004] The purpose of the present invention is to provide a method for compensating sampling frequency offset of an optical fiber communication system, which is used for intensity modulation direct detection of a single-carrier pulse amplitude modulation signal in a high-speed optical fiber communication system, and aims to solve the frequency sampling offset problem caused by a digital-to-analog converter (DAC) and an analog-to-digital converter (ADC).

[0005] The optical fiber communication system sampling frequency offset compensation method provided by the present invention compensates for the sampling frequency offset (SFO) caused by the ADC and DAC in the high-speed optical fiber communication system, and specifically includes sampling frequency offset compensation and signal optimal sampling point calibration:

[0006] (I) Sampling frequency offset compensation:

[0007] When the transmitter sends a single-carrier pulse amplitude modulation signal for normal communication, the signal is converted into a digital sequence by an analog-to-digital converter (ADC) at the receiving end. At this time, the digital sequence at the receiving end is passed through a filter with a given coefficient. The structure and calculation method of the filter are as follows:

[0008]

[0009] Among them, y[n] and z[n] are the input and output of the filter respectively; c q (i) is the fixed coefficient of the filter; given in Table 1 below. m n and μ nis the correction number for the sampling frequency offset sequence and is given by:

[0010]

[0011] Among them, int[] is the rounding function;

[0012] Table 1, fixed coefficients c of the filter q (i)

[0013] 0 1 2 3 4 -3 -0.0596 0.1865 -0.0744 -0.0291 -0.0079 -2 0.1732 -0.517 0.1845 0.1171 0.0155 -1 -0.2643 1.074 -0.419 -0.3206 -0.0266 0 0.1408 -1.3808 1.2717 0.923 0.0426 1 0.1408 0.835 -1.9481 -0.0116 0.999 2 -0.2643 -0.0497 1.2137 -0.9299 0.0131 3 0.1732 -0.154 -0.3431 0.3514 -0.0155 4 -0.0596 0.0445 0.132 -0.1351 0.0112

[0014] Among them, the row represents c q The i in (i) represents the column c q q in (i).

[0015] (II) Signal optimal sampling point calibration:

[0016] After the sampling frequency offset compensation is completed, the time delay of each sampling point relative to the optimal sampling point is fixed rather than uneven. Therefore, the optimal sampling point calibration of the signal can be completed by using the frequency domain time shift method with extremely low complexity.

[0017] Let the signal sequences before and after correction be z[n] and w[n], then:

[0018] w[n]=IFFT[FFT[z[n]]exp(j2πfτ)], (3)

[0019] Where τ is the time difference between the sampling point and the optimal sampling time, w[n] is the signal that corrects the sampling frequency offset and shifts the sampling point of the signal to the optimal point. FFT[·] represents fast Fourier transform, IFFT[·] represents inverse fast Fourier transform, f represents the sampling rate of the signal sequence, and j represents an imaginary number;

[0020] Compared with the prior art, the present invention uses filter equalization to compensate for sampling frequency offset and uses frequency domain time shift method to complete the calibration of the optimal sampling point of the signal, which is simpler, less complex and more flexible. Since it is implemented in digital signal processing, the flexibility of the system is greatly improved. It lays a solid foundation for further improvement of the capacity and rate of future optical fiber communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The optical fiber communication system architecture of the present invention is utilized.

[0022] Figure 2 It is a graph comparing the received optical power and received bit error rate of an actual experimental system constructed when the present invention is used or not.

[0023] The numbers in the figure are: 1 is the digital signal processing module (DSP) at the transmitting end of the optical fiber communication system; 2 is the transmitting end of the system (including DAC); 3 is the optical fiber link of the system; 4 is the receiving end of the system (including ADC); 5 is the sampling frequency offset compensation module, 6 is the sampling time frequency domain correction module; 7 is other DSPs at the receiving end. DETAILED DESCRIPTION

[0024] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] Using the optical fiber communication system architecture of the present invention, such as Figure 1 As shown, it includes a digital signal processing module (DSP) 1 at the system transmitting end; a system transmitting end (including a DAC) 2, a system optical fiber link 3; a system receiving end (including an ADC) 4; a sampling frequency offset compensation module 5, a sampling moment frequency domain correction module 6; and other DSPs 7 at the receiving end.

[0026] Embodiment 1, sampling frequency offset compensation:

[0027] When the digital signal processing module (DSP) 1 of the system transmitting end and the pulse amplitude modulation signal of the system transmitting end (including DAC) 2 for normal communication are transmitted, the signal reaches the receiving end 4 through the transmission of the optical fiber link 3, and the signal is converted into a digital sequence through the analog-to-digital converter (ADC); at this time, the digital sequence of the receiving end is passed through a filter with a given coefficient (i.e., the sampling frequency offset compensation module 5) to perform sampling frequency offset compensation; the structure and calculation method of the filter are as follows:

[0028]

[0029] Among them, y[n] and z[n] are the input and output of the filter respectively; c q (i) is the fixed coefficient of the filter; given in Table 1 below. m n and μ n is given by the following formula:

[0030] m n =int[(1+Δ)n]

[0031] μ b =(1+Δ)nm

[0032] Among them, int[] is the rounding function;

[0033] Table 1, fixed coefficients c of the filterq (i) is:

[0034] 0 1 2 3 4 -3 -0.0596 0.1865 -0.0744 -0.0291 -0.0079 -2 0.1732 -0.517 0.1845 0.1171 0.0155 -1 -0.2643 1.074 -0.419 -0.3206 -0.0266 0 0.1408 -1.3808 1.2717 0.923 0.0426 1 0.1408 0.835 -1.9481 -0.0116 0.999 2 -0.2643 -0.0497 1.2137 -0.9299 0.0131 3 0.1732 -0.154 -0.3431 0.3514 -0.0155 4 -0.0596 0.0445 0.132 -0.1351 0.0112

[0035] Among them, the row represents c q The i in (i) represents the column c q q in (i).

[0036] Embodiment 2, sampling time correction:

[0037] After the signal completes the recovery of the sampling frequency offset filter of the sampling frequency offset compensation module 5, it is sent to the sampling time correction module 6, which is specifically implemented as follows:

[0038] The signal after oversampling frequency offset compensation is quickly corrected at the sampling time point after compensation. The signal sequences before and after correction are z[n] and w[n], then:

[0039] w[n]=IFFT[FFT[z[n]]exp(j2πfτ)]

[0040] Among them, τ is the time difference between the sampling point and the optimal sampling time, and w[n] is the signal that corrects the sampling frequency offset and shifts the sampling point of the signal to the optimal position.

[0041] This completes the work of the sampling time correction module 6. The signal is then handed over to other DSPs for further processing to restore the original signal.

[0042] The present invention uses filter equalization to compensate for SFO, and based on the characteristic that there is a fixed time shift at the sampling point of the signal after compensating for SFO, uses simple frequency domain time shift to achieve accurate recovery of the sampling point, which is simpler, less complex and more flexible. Since it is implemented in digital signal processing, the flexibility of the system is greatly improved. It lays a solid foundation for the further development of future optical fiber communication systems.

[0043] Example 3, actual system verification:

[0044] according to Figure 1A fiber optic transmission system was built for verification. At the transmitting DSP 1, a 50G baud pulse amplitude modulation (PAM) signal is generated and sent to the DAC of the transmitting end 2 to generate an electrical signal waveform. The electro-optical modulator in the transmitting end 2 modulates the electrical signal waveform onto the light of the 1310 nm laser source. After the modulated light enters the optical amplifier in the transmitting end 2, it is transmitted through the optical fiber link 3 (including 20 km of standard single-mode optical fiber), and then reaches the optical amplifier in the receiving end 4 for amplification and is detected by the photodiode in the receiving end 4, which can convert the modulated optical signal into an electrical signal, which is collected by the ADC in the receiving end 4. The collected signal is corrected at the sampling moment by the SFO compensation module 5 and the correction module 6, and the bit error rate is calculated after being processed by other DSPs 7.

[0045] Figure 2 The bit error rate performance of three cases, not using the scheme proposed by the present invention, only using SFO compensation, and using the scheme flow of the present invention for compensation, is compared under different received optical powers of the input photodiode. It can be observed that the bit error rate of the scheme of the present invention is the lowest under different received optical powers, proving that the scheme has great superiority.

Claims

1. A method for compensating sampling frequency offset of an optical fiber communication system, characterized in that: Compensation for sampling frequency offset (SFO) caused by ADC and DAC in high-speed fiber-optic communication systems, including sampling frequency offset compensation and signal optimal sampling point calibration: (I) Sampling frequency offset compensation: When the transmitter sends a single-carrier pulse amplitude modulation signal for normal communication, the signal is converted into a digital sequence by an analog-to-digital converter (ADC) at the receiving end. At this time, the digital sequence at the receiving end is passed through a filter with a given coefficient. The structure and calculation method of the filter are as follows: Among them, y[n] and z[n] are the input and output of the filter respectively; c q (i) is the fixed coefficient of the filter; given in Table 1 below; m n and μ n is the correction number for the sampling frequency offset sequence and is given by: Among them, int[] is the rounding function; Table 1, fixed coefficients c of the filter q (i) Among them, the row represents c q The i in (i) represents the column c q q in (i); (II) Calibration of optimal signal sampling point: After completing the sampling frequency offset compensation, the time delay of each sampling point relative to the optimal sampling point is fixed, so the optimal sampling point of the signal is calibrated by the frequency domain time shift method; Let the signal sequences before and after correction be z[n] and w[n], then: w[n]=IFFT[FFT[z[n]]exp(j2πfτ)], (3) Where τ is the time difference between the sampling point and the optimal sampling time, w[n] is the signal that corrects the sampling frequency offset and shifts the sampling point of the signal to the optimal point; FFT[·] represents fast Fourier transform, IFFT[·] represents inverse fast Fourier transform, f represents the sampling rate of the signal sequence, and j represents an imaginary number.

Images