Error propagation suppression method of differential decoder based on error endpoint positioning

Abstract

The invention provides an error propagation suppression method for a differential decoder based on error endpoint positioning, and the digital signal processing of a receiving end comprises the steps: recovering a part of response signals through employing the differential decoder, and detecting an input signal of a fourth-order pulse amplitude modulation PAM4 decision device in the differential decoder through employing a threshold detector, when the input signal of the PAM4 decision device is greater than a preset threshold value, judging that the current sampling point is the end point of the burst error, recording the polarity of the error and activating the error correction step; forward backtracking is carried out from the end point of the burst error, and judgment noise of the first L symbols is calculated; marking the symbol with the maximum noise as a starting point of error propagation; and finally, symbols in a range from an error propagation starting point to an error propagation ending point are corrected so as to suppress burst errors caused by a differential decoder. According to the invention, error propagation introduced by the differential decoder can be effectively suppressed, the sensitivity of the receiver is remarkably improved, and the transmission performance of the system is improved.

Description

Technical Field

[0001] The present invention relates to optical fiber communication technology, and particularly to an equalizer error propagation suppression technology for a direct detection optical fiber communication system. Background Art

[0002] With the explosive growth of the bandwidth demand for data center interconnect (DCI), the rate of the next-generation optical module is expected to reach 1.6 Tb / s, and the single-channel rate exceeds 200 Gb / s to reduce the integration complexity. Among various modulation formats, the PAM4 signal can effectively balance the spectral utilization rate and the noise tolerance, and is the most promising solution for realizing 200 Gb / s single-wavelength transmission. However, the increase in the transmission rate and low-cost optoelectronic devices have led to challenges in bandwidth limitation for the intensity modulation direct detection (IM / DD) system, resulting in severe inter-symbol interference (ISI). The traditional feed-forward equalizer (FFE) is commonly used to eliminate ISI, but its performance is limited due to the noise enhancement effect; although the decision feedback equalizer (DFE) can suppress ISI without amplifying noise, there are problems of error propagation and it is difficult to meet the high-speed timing requirements; although the maximum likelihood sequence estimation (MLSE) can effectively eliminate ISI, its computational complexity increases exponentially with the modulation order and the memory length, making it difficult to be practical in high-speed systems. Therefore, in order to meet the requirements of low-cost high-speed transmission, a reasonable method is needed to effectively solve the problem of ISI.

[0003] The partial response equalization (PRE) can achieve spectral shaping through the FFE at the receiving end, which can avoid the problem of noise amplification, but it depends on the decoder to recover the original signal. The candidate symbol set of the traditional MLSE decoder is huge, resulting in huge consumption of computing resources; on the other hand, a simple differential decoder will cause error propagation. Therefore, how to suppress the error propagation of the differential decoder to adapt to the PRE system is of great research significance for realizing high-speed optical transmission. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an effective method for suppressing the error propagation of a differential decoder.

[0005] The technical solution adopted by the present invention to solve the above technical problem is an error propagation suppression method for a differential decoder based on error endpoint positioning. The digital signal processing at the receiving end includes the following steps:

[0006] Burst error detection step: A threshold detector is used in the differential decoder to detect the input signal of the PAM4 decision maker. When the input signal of the PAM4 decision maker is greater than a preset threshold, it is determined that the current sampling point is the end point of a burst error, the error polarity is recorded, and the error correction step is activated;

[0007] Error correction steps: Starting from the detected end point of error propagation, trace back forward, perform hard decision on the first L symbols using 7th-order pulse amplitude modulation PAM7, then subtract the corresponding hard decision result from the symbol sampling value to obtain the hard decision error of the first L symbols; then find the symbol corresponding to the hard decision error with the largest absolute value, and mark this symbol as the starting point of the burst error; then subtract the continuous error sequence with alternating +2 and -2 from the input sequence of the PAM4 decision maker within the range from the starting point to the end point of the burst error; then determine whether the output result is valid. If so, output the result after subtraction to correct the error propagation caused by the differential decoder. If not, skip this error correction.

[0008] Since this method can directly detect the starting point and end point of the burst error, during error correction, it only needs to subtract the continuous error sequence from the input sequence of the PAM4 decision maker within the range from the starting point to the end point of the burst error to suppress the error propagation of the differential decoder. This method only uses comparators and adders and has a low computational complexity.

[0009] The beneficial effects of the present invention are as follows: The method proposed by the present invention can effectively suppress the error propagation introduced by the differential decoder, significantly improve the receiver sensitivity, and improve the transmission performance of the system. Description of the Drawings

[0010] Figure 1 is the principle block diagram of the error propagation suppression of the differential decoder based on the positioning of the error start and end points of the present invention;

[0011] Figure 2 is the experimental setup diagram of the PAM4 intensity modulation direct detection optical fiber communication system. Detailed Embodiments

[0012] The digital signal processor DSP at the receiving end includes a partial response feed-forward equalizer differential decoder, a threshold detector, and an error correction module, and realizes the error propagation suppression of the differential decoder based on the positioning of the error start and end points by performing the following steps:

[0013] While the received signal eliminates inter-symbol interference ISI through the partial response feed-forward equalizer FFE and performs spectral shaping, then a differential decoder with a response of G(z) = 1 / (1 + z –1 ) is used for decoding. z –1 represents the delay of a single symbol interval. Then, an error correction module based on the positioning of the error start and end points is used to suppress the error propagation of the differential decoder, and the specific process is as Figure 1 shown.

[0014] The error correction module based on the positioning of the error start and end points is divided into two steps:

[0015] First step, use a threshold detector in the differential decoder to detect the end point of the burst error. Assuming a PAM4 signal, the upper threshold of the threshold detector is 4 and the lower threshold is -4. Symbols outside this range are determined as the end points of the burst error.

[0016] Second step, when the input signal of the PAM4 decision maker exceeds the threshold of the threshold detector, activate the error correction module to correct the burst error caused by the differential decoder.

[0017] The partial response feed-forward equalizer FFE in the receiving-end DSP is connected to the differential decoder. The input terminals of the PAM4 decision maker in the differential decoder are respectively connected to the threshold detector and the error correction module.

[0018] The error correction module includes a maximum value calculation unit, a polarity judgment unit, an error sequence generation unit, and an error correction unit. Taking the second-order partial response as an example, the differential decoder includes a PAM7 decision maker, a PAM4 decision maker, two adders, and a single-symbol interval delay element D. The output terminal of the FFE is connected to the input terminal of the PAM7 decision maker, and the signal y(k) processed by the FFE is input to the PAM7 decision maker. The output terminal of the PAM7 decision maker and the output terminal of the PAM4 decision maker are both connected to the first adder through the single-symbol interval delay element D. The output terminal of the first adder is respectively connected to the input terminal of the PAM4 decision maker, the input terminal of the threshold detector, and the input terminal of the error correction unit in the error correction module.

[0019] The input terminal and the output terminal of the PAM7 decision maker in the differential decoder are respectively connected to the input terminal of the second adder. The output terminal of the second adder is connected to the input terminal of the maximum value calculation unit in the error correction module. The output terminal of the maximum value calculation unit in the error correction module is connected to one input terminal of the error sequence generation unit. The output of the threshold detector is connected to the other input terminal of the error sequence generation unit after passing through the polarity judgment unit. The output terminal of this error sequence generation unit is connected to one input terminal of the error correction unit. The input sequence of the PAM4 decision maker in the differential decoder is input to the error correction module through the other input terminal of the error correction unit.

[0020] The signal y(k) is input into the PAM7 decision maker in the differential decoder. The output signal y’(k) after PAM7 decision is subtracted from the input signal y(k) of the PAM7 decision maker by the second adder to obtain the decision noise n(k), where k represents the k-th sampling point of the signal. y’(k) is subtracted from the signal x’(k-1) which is the output of the PAM4 decision maker delayed by a single symbol interval D by the first adder to obtain the input signal x(k) of the PAM4 decision maker. When the threshold detector detects that x(k) exceeds the threshold, this position is marked as the end point of the burst error. Looking back L symbols from the end point of the burst error, the maximum value calculation unit finds the noise with the largest absolute value within the range from n(k-L) to n(k) to determine the start point of the burst error and calculate the burst error length. The error sequence generation unit generates a predicted error propagation sequence using the burst error length and the output result of the polarity judgment unit. The error correction unit subtracts the predicted error propagation sequence from the input of the PAM4 decision maker to correct the burst error caused by the differential decoder.

[0021] When the obtained corrected signal is an unreasonable situation, that is, the symbol corrected according to the predicted error propagation sequence exceeds the maximum or minimum level value of the PAM4 signal, we consider this predicted error propagation sequence to be invalid. At this time, no modification needs to be made to the output of the differential decoder. On the contrary, the predicted error propagation sequence is used to correct the burst error caused by the differential decoder. That is, the method for judging whether the predicted error propagation sequence is valid is: the obtained correction is valid within the effective level range.

[0022] The embodiment takes a four-level pulse amplitude modulation PAM4 direct detection optical fiber communication system as an example to illustrate the specific working principle of the present invention, specifically as Figure 2 shown.

[0023] In the digital signal processing at the transmitting end, first, the original pseudo-random sequence is modulated into a PAM4 signal, and then the root raised cosine filter is used to perform pulse shaping on the signal and upsampling. The upsampled signal is input into the digital-to-analog converter DAC for digital-to-analog conversion after resampling to obtain an analog radio frequency signal. Then, the electrical signal is amplified by the radio frequency amplifier EA and input into the Mach-Zehnder MZM modulator together with the laser output by the laser for electro-optic modulation to obtain the transmitted optical signal.

[0024] When the transmitted signal is transmitted through the optical fiber, at the receiving end, a photodiode PD is used to achieve photoelectric conversion. The electrical signal received by the PD is collected by the oscilloscope and then undergoes subsequent offline digital signal processing DSP including resampling, matched filtering, downsampling, partial response feedforward equalization, differential decoding, burst error correction, PAM4 demapping, and bit error rate calculation.

[0025] In the offline receiving end's digital signal processing DSP, the received data is equalized by a partial response feed-forward equalizer (FFE) to eliminate inter-symbol interference (ISI) and simultaneously perform spectral shaping. Then, a differential decoder decodes the output of the FFE, and a threshold detector is introduced before the PAM4 decision maker to detect the end point of burst errors. When the input symbol does not exceed the threshold, the differential decoder directly outputs the decoded result. Once a symbol exceeding the threshold is detected, the end point of the burst error is recorded and the error correction module is activated to correct the burst error caused by the differential decoder. The starting point of the burst error can be located by finding the PAM7 decision error with the largest absolute value among the L symbols before the end point of the burst error. After locating the starting point and end point of the burst error, the input signal of the PAM4 decision maker within this range is subtracted by the predicted error sequence to correct the burst error caused by the differential decoder. Finally, the corrected signal is demapped by PAM4 into a binary bit stream, and the bit error rate is calculated by comparing it with the original binary bit stream.

Claims

1. Error propagation suppression method for a differential decoder based on error endpoint positioning, characterized in that The digital signal processing at the receiving end of the system includes the following steps: Burst error detection step: In the differential decoder, a threshold detector is used to detect the input signal of the fourth-order pulse amplitude modulation PAM4 decision maker. When the input signal of the PAM4 decision maker is greater than the preset threshold, it is determined that the current sampling point is the end point of the burst error, the error polarity is recorded, and the error correction step is activated; Error correction step: Starting from the detected end point of error propagation, trace back a set length forward, perform hard decision on the seventh-order pulse amplitude modulation PAM7 of the traced symbols, and then subtract the corresponding PAM7 hard decision result from the sampling value of the symbol to obtain the hard decision error of the traced symbol. Mark the symbol corresponding to the hard decision error with the largest absolute value as the starting point of the burst error; Then, within the range between the starting point and the end point of the burst error, subtract the predicted error sequence from the input signal of the PAM4 decision maker to correct the error propagation caused by the differential decoder.

2. The method according to claim 1, wherein Within the range from the starting point to the end point of the burst error, subtract the continuous error sequence with alternating +2 and –2 from the input sequence of the PAM4 decision maker respectively; then determine whether the corrected sequence is valid. If it is, output the corrected sequence; if not, skip this error correction.

3. The method according to claim 1, wherein The upper threshold of the threshold detector is M, and the lower threshold is –M; where M is the modulation order.

4. The method according to claim 1, wherein The traced length L should be greater than the maximum error propagation length.

5. The method according to claim 1, wherein The continuous error sequence with alternating +2 and –2 is {+2, –2, +2, …} or {–2, +2, –2, …}, which is determined by the error polarity at the end point of the burst error.

6. The method according to claim 5, wherein The method for determining whether the corrected sequence is valid is: the symbols of the corrected sequence are valid within the PAM4 valid level range.

Images