Method for Testing TDEC and Related Device

Abstract

A method for testing a TDEC, where the method includes: obtaining a to-be-tested signal obtained through channel transmission of an original signal transmitted by a transmitter; performing first equalization processing on the to-be-tested signal based on at least one first tap coefficient to obtain a first signal; obtaining a product of a level corresponding to the original signal and a second tap coefficient, and performing second equalization processing on the first signal based on the product to obtain a second signal; obtaining an eye pattern of the second signal, and performing sampling at a target position of the eye pattern to obtain a third signal; obtaining a target symbol error rate (SER) obtained by superimposing target noise on the third signal; and adjusting the target noise based on a SER threshold to adjust the target SER.

Description

TECHNICAL FIELD

This application relates to the communication field, and in particular, to a method for testing a transmitter and dispersion eye closure (TDEC) and a related device.

BACKGROUND

In some optical communication network scenarios, for example, intra-data interconnection, to implement communication between one receiver and different transmitters, a consistency test may need to be performed on the different transmitters. Currently, a method for performing a consistency test on a transmitter includes measuring an eye mask, a transmitter and dispersion penalty (TDP), and a TDEC penalty. For a 4-order pulse-amplitude modulation (PAM-4) signal, transmitter and dispersion eye closure for PAM-4 (TDECQ) may need to be tested. A physical meaning of the TDEC is to calculate a ratio of amount of maximum superimposed noise on an ideal signal to power of maximum superimposed noise on a to-be-tested signal under given conditions of a bit error threshold and a reference receiver.

Typically, to make a TDEC result more valuable, parameters of a TDEC test receiver should be approximate to these of receivers in most actual systems. In a current solution of performing the consistency test on the transmitter, the TDEC test receiver performs equalization processing by using a feed forward equalizer (FFE) and calculates the TDEC. However, in a next-generation optical communication scenario, to handle severe signal impairment, the real receiver in the actual system uses a technology with a stronger equalization capability. Consequently, the TDEC test receiver has an equalization capability weaker than that of the receiver in the actual system, and the current TDEC test receiver and the current TDEC testing method have poor application effects in a severe signal impairment scenario.

SUMMARY

Embodiments of this application provide a method for testing a TDEC and a related device, to improve an equalization processing capability, and reduce complexity of estimating a symbol error rate (SER), so as to be better applied to a scenario with severe signal impairment.

According to a first aspect, an embodiment of this application provides a method for testing a TDEC. The method includes the following steps: obtaining a to-be-tested signal obtained through channel transmission of an original signal transmitted by a transmitter; performing first equalization processing on the to-be-tested signal based on at least one first tap coefficient to obtain a first signal; obtaining a product of a level corresponding to the original signal and a second tap coefficient, and performing second equalization processing on the first signal based on the product to obtain a second signal, where a difference obtained by subtracting the second tap coefficient from a sum of all the first tap coefficients is 1; obtaining an eye pattern of the second signal, and performing sampling at a target position of the eye pattern to obtain a third signal; obtaining a target SER obtained by superimposing target noise on the third signal; and adjusting the target noise to adjust the target SER, for calculating the TDEC based on adjusted target noise if the target SER is equal to a SER threshold.

In this implementation, in this application, the first equalization processing and the second equalization processing are combined, to improve an equalization capability, so that a high baud rate system is better supported. In addition, this application provides a method for estimating a target SER obtained through equalization processing. Decision feedback equalizer (DFE) error propagation is not first considered, and a DFE performs equalization processing by using the original signal, to obtain an initial SER without consideration of the error propagation of the DFE, and then calculates a modification factor and modifies the initial SER based on the modification factor to obtain the target SER. It should be understood that the modification factor may be considered as a SER increase multiple caused by the DFE error propagation, so that a real estimation of the SER obtained through the DFE equalization processing can be obtained. In this manner, under a condition in which a length of a test sequence of the original signal is limited, low operational complexity of estimating the SER is implemented, so that a scenario with severe signal impairment is better applied.

In some possible implementations, the obtaining a target SER obtained by superimposing target noise on the third signal includes: obtaining an initial SER obtained by superimposing the target noise on the third signal; and further modifying the initial SER based on a modification factor to obtain the target SER.

In some possible implementations, the modifying the initial SER based on a modification factor to obtain the target SER includes: multiplying the initial SER by the modification factor to obtain the target SER.

In some possible implementations, the modification factor satisfies the following formula 1:

$f_{\mathrm{modify}}=\frac{1}{1-\left(P_E-{\mathrm{SER}}_0\right)}$

where f_modify represents the modification factor, SER, represents the initial SER, P_E represents a probability that a second symbol is erroneously decided due to a decision error of a first symbol in the second equalization processing, and a receiving moment of the first symbol is earlier than a receiving moment of the second symbol.

In some possible implementations, the modification factor satisfies the following formula 2:

$f_{\mathrm{modify}}=\frac{1}{1-P_E}$

where f_modify represents the modification factor, P_E represents a probability that a second symbol is erroneously decided due to a decision error of a first symbol in the second equalization processing, and a receiving moment of the first symbol is earlier than a receiving moment of the second symbol.

In some possible implementations, P_E is determined by the second tap coefficient, a root mean square (RMS) value of the target noise, and a sampling level-related parameter, and the sampling level-related parameter includes at least one of frequency distribution of sampling levels of the third signal, average optical power of the eye pattern, and an outer optical modulation amplitude of the second signal.

In some possible implementations, P_E satisfies the following formula 3:

$P_E=\sum _{i=1}^K{P}_E^k$

where K is a quantity of eyes in the eye pattern, P_E^k represents a probability that the second symbol is erroneously decided at a k^th eye, and 1≤k≤K.

In some possible implementations, P_E satisfies the following formula 4:

$P_E^k=\left(\int {\mathrm{CF}}_k\left(y\right)G_{\mathrm{thk}}^{\mathrm{shift}+}\left(y\right)\mathrm{dy}+\int {\mathrm{CF}}_k\left(y\right)G_{\mathrm{thk}}^{\mathrm{shift}-}\left(y\right)\mathrm{dy}\right)/2$

where CF_k(y) represents a cumulative probability function of level distribution corresponding to the k^th eye, G_thk^shift+(y) represents a probability density function of superimposed noise obtained by offsetting a decision threshold of the k^th eye in a first direction, and G_thk^shift−(y) represents a probability density function of superimposed noise obtained by offsetting the decision threshold of the k^th eye in a second direction, the first direction is opposite to the second direction, and y represents a sampling level value of the third signal.

In some possible implementations, CF_k(y) satisfies the following formula 5:

${\mathrm{CF}}_k\left(y_i\right)=\{\begin{array}{cc}\sum _{y=P_{\mathrm{thk}}}^{y_i}F\left(y\right)& \mathrm{for}{y}_i\ge P_{\mathrm{thk}}\\ \sum _{y=y_i}^{P_{\mathrm{thk}}}F\left(y\right)& \mathrm{for}{y}_i<P_{\mathrm{thk}}\end{array}$

where P_thk is a decision threshold of a k^th eye, F(y) represents the frequency distribution function of the sampling levels of the third signal, and y_i is the representation of discretized y.

G_thk^shift+(y) and G_thk^shift−(y) satisfy the following formula 6:

$G_{\mathrm{thk}}^{\mathrm{shift}+}\left(y_i\right)={\int }_{y_i-\frac{\Delta y}{2}}^{y_i+\frac{\Delta y}{2}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{\mathrm{thk}}+w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{K}\right)}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$ $G_{\mathrm{thk}}^{\mathrm{shift}-}\left(y_i\right)={\int }_{y_i-\frac{\Delta y}{2}}^{y_i+\frac{\Delta y}{2}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{\mathrm{thk}}-w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{K}\right)}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$

where W_DFE represents the second tap coefficient, OMA_outer represents the outer optical modulation amplitude of the second signal, C_eq represents a coefficient of noise enhancement through the first equalization processing and the second equalization processing, σ_G represents the RMS value of the target noise, and Δy represents an interval between two adjacent y_i.

In some possible implementations, P_E satisfies the following formula 7:

$P_E=\{\begin{array}{c}0.5\times \mathrm{erfc}\left(\frac{1-2w_{\mathrm{DFE}}}{\sqrt{2}}\times \sqrt{\frac{3\times \mathrm{SNR}}{\left(M+1\right)\times \left(M-1\right)}}\right)\times \left(1-\frac{1}{M}\right),w_{\mathrm{DFE}}\le 0.5\\ \begin{array}{c}(1-0.5\times \mathrm{erfc}(\frac{2w_{\mathrm{DFE}}-1}{\sqrt{2}}\times \\ \sqrt{\frac{3\times \mathrm{SNR}}{\left(M+1\right)\times \left(M-1\right)}}))\times \left(1-\frac{1}{M}\right),0.5w_{\mathrm{DFE}}\le 1\end{array}\end{array}$

where erfc represents a complementary error function, an SNR represents a signal-to-noise ratio obtained by superimposing the target noise on the third signal, w_DFE represents the second tap coefficient, M represents a quantity of level values of the original signal, and M≥2.

In some possible implementations, the SNR satisfies the following formula 8:

$\mathrm{SNR}=\frac{P_{sig}}{{\sum }_{y_i}F\left(y_i\right)\times {\left(y_i-P_{ave}\right)}^2-P_{sig}+{\sigma }_G^2}$

where P_sig represents power of a signal component without noise and distortion in the third signal, y_i is a representation of a discretized sampling level value of the third signal, P_ave represents the average optical power of the eye pattern, and σ_G represents the RMS value of the target noise.

In some possible implementations, P_sig satisfies the following formula 9:

$P_{sig}=\frac{\left(M+1\right)\times \left(M-1\right)}{3}\times {\left(\frac{{\mathrm{OMA}}_{o\mathrm{uter}}}{2\times \left(M-1\right)}\right)}^2$

where OMA_outer represents the outer optical modulation amplitude of the second signal, M represents the quantity of level values of the original signal, and M≥2.

In some possible implementations, the original signal includes a test sequence periodically sent by the transmitter.

In some possible implementations, the first tap coefficient is a tap coefficient of an FFE, and the second tap coefficient is a tap coefficient of a DFE.

In some possible implementations, the target position of the eye pattern includes a

$\left(0.5-\frac{\Delta }{2}\right)$

unit interval (UI) and a

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI},$

and 0<Δ<1.

In some possible implementations, the third signal includes a first sub-signal and a second sub-signal, the first sub-signal is obtained through sampling at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}$

of the eye pattern, and the second sub-signal is obtained through sampling at the

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI}$

of the eye pattern. The obtaining a target SER obtained by superimposing target noise on the third signal includes: obtaining a first SER obtained by superimposing the target noise on the first sub-signal, and obtaining a second SER obtained by superimposing the target noise on the second sub-signal; and selecting a larger one of the first SER and the second SER as the target SER. For example, the first SER is greater than the second SER.

In some possible implementations, the original signal includes a signal using PAM-4 or a signal using 2-order pulse-amplitude modulation (PAM-2).

In some possible implementations, the level corresponding to the original signal is represented as

$P_{ave}+\frac{OM{A}_{outer}}{M-1}\left(i-\frac{M+1}{2}\right),$

P_ave represents the average optical power of the eye pattern, OMA_outer represents the outer optical modulation amplitude of the second signal, 1≤i≤M, M represents the quantity of level values of the original signal, and M≥2.

According to a second aspect, an embodiment of this application provides a processing chip. The processing chip includes a processor and a memory. The memory and the processor are connected to each other through a line, the memory stores instructions, and the processor is configured to perform the method described in any implementation of the first aspect.

According to a third aspect, an embodiment of this application provides a TDEC test receiver, including a receiver front end and a processing chip. The receiver front end and the processing chip are connected to each other through a line. The receiver front end is configured to: convert a received optical signal into an electrical signal, and transmit the electrical signal to the processing chip. The processing chip is configured to perform the following steps: obtaining a to-be-tested signal obtained through channel transmission of an original signal transmitted by a transmitter; performing first equalization processing on the to-be-tested signal based on at least one first tap coefficient to obtain a first signal; obtaining a product of a level corresponding to the original signal and a second tap coefficient, and performing second equalization processing on the first signal based on the product to obtain a second signal, where a difference obtained by subtracting the second tap coefficient from a sum of all the first tap coefficients is 1; obtaining an eye pattern of the second signal, and performing sampling at a target position of the eye pattern to obtain a third signal; obtaining a target SER obtained by superimposing target noise on the third signal; and adjusting the target noise to adjust the target SER, for calculating the TDEC based on adjusted target noise if the target SER is equal to a SER threshold.

In some possible implementations, the processing chip is specifically configured to: obtaining an initial SER obtained by superimposing the target noise on the third signal; and modifying the initial SER based on a modification factor to obtain the target SER.

In some possible implementations, the processing chip is specifically configured to: multiplying the initial SER by the modification factor to obtain the target SER.

In some possible implementations, the modification factor satisfies the following formula 1:

$f_{\mathrm{modify}}=\frac{1}{1-\left(P_E-SE{R}_0\right)}$

where f_modify represents the modification factor, SER₀ represents the initial SER, P_E represents a probability that a second symbol is erroneously decided due to a decision error of a first symbol in the second equalization processing, and a receiving moment of the first symbol is earlier than a receiving moment of the second symbol.

In some possible implementations, the modification factor satisfies the following formula 2:

$f_{\mathrm{modify}}=\frac{1}{1-P_E}$

where f_modify represents the modification factor, P_E represents a probability that a second symbol is erroneously decided due to a decision error of a first symbol in the second equalization processing, and a receiving moment of the first symbol is earlier than a receiving moment of the second symbol.

In some possible implementations, P_E is determined by the second tap coefficient, an RMS value of the target noise, and a sampling level-related parameter, and the sampling level-related parameter includes at least one of frequency distribution of sampling levels of the third signal, average optical power of the eye pattern, and an outer optical modulation amplitude of the second signal.

In some possible implementations, P_E satisfies the following formula 3:

$P_E=\sum _{i=1}^K{P}_E^k$

where K is a quantity of eyes in the eye pattern, P_E represents a probability that the second symbol is erroneously decided at a k^th eye, and 1≤k≤K.

In some possible implementations, P_E satisfies the following formula 4:

$P_E^k=\left(\int C{F}_k\left(y\right)G_{thk}^{\mathrm{shift}+}\left(y\right)dy+\int C{F}_k\left(y\right)G_{thk}^{sh\mathrm{ift}-}\left(y\right)\mathrm{dy}\right)/2$

where CF_k(y) represents a cumulative probability function of level distribution corresponding to the k^th eye, G_thk^shift+(y) represents a probability density function of superimposed noise obtained by offsetting a decision threshold of the k^th eye in a first direction, and G_thk^shift−(y) represents a probability density function of superimposed noise obtained by offsetting the decision threshold of the k^th eye in a second direction, the first direction is opposite to the second direction, and y represents a sampling level value of the third signal.

In some possible implementations, CF_k(y) satisfies the following formula 5:

$C{F}_k\left(y_i\right)=\{\begin{array}{c}\sum _{y=P_{\mathrm{thk}}}^{y_i}F\left(y\right)\mathrm{for}{y}_i\ge P_{\mathrm{thk}}\\ \sum _{y=y_i}^{P_{\mathrm{thk}}}F\left(y\right)\mathrm{for}{y}_i<P_{\mathrm{thk}}\end{array}$

where P_thk is a decision threshold of a k^th eye, F(y) represents the frequency distribution function of the sampling levels of the third signal, and y_i is the representation of discretized y.

G_thk^shift+(y) and G_thk^shift−(y) satisfy the following formula 6:

$G_{thk}^{sh\mathrm{ift}+}\left(y_i\right)={\int }_{y_i-\frac{\Delta y}{2}}^{y_i+\frac{\Delta y}{2}}\frac{1}{C_{eq}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{thk}+w_{DFE}\times \frac{OM{A}_{outer}}{K}\right)}{C_{eq}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$ $G_{thk}^{\mathrm{shift}-}\left(y_i\right)={\int }_{y_i-\frac{\Delta y}{2}}^{y_i+\frac{\Delta y}{2}}\frac{1}{C_{eq}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{thk}-w_{DFE}\times \frac{OM{A}_{outer}}{K}\right)}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}dy$

where w_DFE represents the second tap coefficient, OMA_outer represents the outer optical modulation amplitude of the second signal, C_eq represents a coefficient of noise enhancement through the first equalization processing and the second equalization processing, σ_G represents the RMS value of the target noise, and Δy represents an interval between two adjacent y_i.

In some possible implementations, P_E satisfies the following formula 7:

$P_E=\{\begin{array}{c}0.5\times \mathrm{erfc}\left(\frac{1-2w_{\mathrm{DFE}}}{\sqrt{2}}\times \sqrt{\frac{3\times \mathrm{SNR}}{\left(M+1\right)\times \left(M-1\right)}}\right)\times \left(1-\frac{1}{M}\right),w_{\mathrm{DFE}}\le 0.5\\ \begin{array}{c}(1-0.5\times \mathrm{erfc}(\frac{2w_{\mathrm{DFE}}-1}{\sqrt{2}}\times \\ \sqrt{\frac{3\times \mathrm{SNR}}{\left(M+1\right)\times \left(M-1\right)}}))\times \left(1-\frac{1}{M}\right),0.5w_{\mathrm{DFE}}\le 1\end{array}\end{array}$

where erfc represents a complementary error function, an SNR represents a signal-to-noise ratio obtained by superimposing the target noise on the third signal, w_DFE represents the second tap coefficient, M represents a quantity of level values of the original signal, and M≥2.

In some possible implementations, the SNR satisfies the following formula 8:

$\mathrm{SNR}=\frac{P_{\mathrm{sig}}}{{\sum }_{y_i}F\left(y_i\right)\times {\left(y_i-P_{\mathrm{ave}}\right)}^2-P_{\mathrm{sig}}+{\sigma }_G^2}$

where P_sig represents power of a signal component without noise and distortion in the third signal, y_i is a representation of a discretized sampling level value of the third signal, P_ave represents the average optical power of the eye pattern, and σ_G represents the RMS value of the target noise.

In some possible implementations, P_sig satisfies the following formula 9:

$P_{\mathrm{sig}}=\frac{\left(M+1\right)\times \left(M-1\right)}{3}\times {\left(\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{2\times \left(M-1\right)}\right)}^2$

where OMA_outer represents the outer optical modulation amplitude of the second signal, M represents the quantity of level values of the original signal, and M≥2.

In some possible implementations, the original signal includes a test sequence periodically sent by the transmitter.

In some possible implementations, the first tap coefficient is a tap coefficient of an FFE, and the second tap coefficient is a tap coefficient of a DFE.

In some possible implementations, the target position of the eye pattern includes a

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}\mathrm{and}a\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI},$

In some possible implementations, the third signal includes a first sub-signal and a

second sub-signal, the first sub-signal is obtained through sampling at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}$

of the eye pattern, and the second sub-signal is obtained through sampling at the

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI}$

of the eye pattern. The processing chip is specifically configured to: obtain a first SER obtained by superimposing the target noise on the first sub-signal, and obtain a second SER obtained by superimposing the target noise on the second sub-signal; and select a larger one of the first SER and the second SER as the target SER. For example, the first SER is greater than the second SER.

In some possible implementations, the original signal includes a signal using PAM-4 or a signal using PAM-2.

In some possible implementations, the level corresponding to the original signal is represented as

$P_{\mathrm{ave}}+\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{M-1}\left(i-\frac{M+1}{2}\right),$

P_ave represents the average optical power of the eye pattern, OMA_outer represents the outer optical modulation amplitude of the second signal, 1≤i≤M, M represents the quantity of level values of the original signal, and M≥2.

In this embodiment of this application, the DFE is added to the TDEC test receiver, to improve an equalization capability, so that a high baud rate system is better supported. In addition, this application provides a method for estimating a target SER obtained through equalization processing. DFE error propagation is not first considered, and a DFE performs equalization processing by using the original signal, to obtain an initial SER without consideration of the error propagation of the DFE, and then calculates a modification factor and modifies the initial SER based on the modification factor to obtain the target SER. It should be understood that the modification factor may be considered as a SER increase multiple caused by the DFE error propagation, so that a real estimation of the SER obtained through the DFE equalization processing can be obtained. In this manner, under a condition in which a length of a test sequence of the original signal is limited, low operational complexity of estimating the SER is implemented, so that a scenario with severe signal impairment is better applied.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 is a diagram of a TDEC test system;

FIG. 2 is a diagram of a structure of a TDEC test receiver;

FIG. 3 is a diagram of an embodiment of a method for testing a TDEC according to an embodiment of this application;

FIG. 4 is a diagram of equalization processing according to an embodiment of this application;

FIG. 5A is a diagram of a first estimation effect of a SER increase multiple caused by DFE error propagation;

FIG. 5B is a diagram of a second estimation effect of a SER increase multiple caused by DFE error propagation;

FIG. 6A is a diagram of a third estimation effect of a SER increase multiple caused by DFE error propagation;

FIG. 6B is a diagram of a fourth estimation effect of a SER increase multiple caused by DFE error propagation; and

FIG. 7 is a diagram of a structure of a TDEC test receiver according to an embodiment of this application.

DESCRIPTION OF EMBODIMENTS

Embodiments of this application provide a method for testing a TDEC and a related device, to improve an equalization processing capability, and reduce complexity of estimating a SER, so as to be better applied to a scenario with severe signal impairment. It should be noted that in the specification, claims, and the foregoing accompanying drawings of this application, the terms such as “first” and “second” are intended to distinguish between similar objects, but do not limit a specific order or sequence. It should be understood that the foregoing terms are interchangeable in appropriate cases, so that embodiments described in this application can be implemented in a sequence other than the content described in this application. In addition, the terms “including” and “having” and any other variants thereof are intended to cover a non-exclusive inclusion. For example, a process, a method, a system, a product, or a device that includes a list of steps or units is not necessarily limited to those expressly listed steps or units, but may include other steps or units not expressly listed or inherent to the process, the method, the product, or the device.

This application is mainly applied to a TDEC test scenario. A physical meaning of the TDEC is to calculate a ratio of power of maximum superimposed noise on an ideal signal to power of maximum superimposed noise on a to-be-tested signal under given conditions of a bit error threshold and a reference receiver. In some possible scenarios, for a PAM-4 signal, a TDEC test may also be denoted as a TDECQ test.

FIG. 1 is a diagram of a TDEC test system. As shown in FIG. 1, the TDEC test system includes a transmitter and a TDEC test receiver. The transmitter is configured to: generate a test sequence and modulate the test sequence into an optical signal, where the optical signal transmitted by the transmitter is transmitted to the TDEC test receiver through a channel. The TDEC test receiver processes the received optical signal to complete a TDEC test. The following further describes the TDEC test receiver.

FIG. 2 is a diagram of a structure of a TDEC test receiver. As shown in FIG. 2, the TDEC test receiver includes a polarization rotator 10, an optical splitter 20, a tunable reflector 30, a test fiber 40, a photoelectric conversion unit 50, a clock recovery unit 60, an oscilloscope 70, and a processor 80. Specifically, the optical signal from the transmitter is transmitted to the polarization rotator 10, and maximum relative intensity noise (relative intensity noise, RIN) can be obtained by adjusting the polarization rotator 10. The optical splitter 20 splits an optical signal obtained by the polarization rotator 10 through processing, one channel of split optical signals is transmitted to the tunable reflector 30, and the other channel of split optical signals is transmitted to the photoelectric conversion unit 50 through the test fiber 40. An optical return loss (optical return loss) may meet a standard specification by adjusting the tunable reflector 30. The optical return loss refers to a ratio of optical power of a signal that is returned to an input port along an original path due to reasons such as reflection and scattering in a transmission process of inputting the optical signal to incident optical power. The photoelectric conversion unit 50 converts the received optical signal into an electrical signal, and separately transmits the electrical signal to the clock recovery unit 60 and the oscilloscope 70. The clock recovery unit 60 performs clock recovery on the electrical signal to obtain a clock signal, and transmits the clock signal to the oscilloscope 70. The oscilloscope 70 acquires a complete sampled electrical signal. The processor 80 first filters the sampled electrical signal, and finally analyzes a filtered signal to complete the TDEC test. It should be understood that a manner of acquiring the complete sampled electrical signal by the oscilloscope 70 includes but is not limited to a mode triggering manner. A manner in which the processor 80 filters the sampled electrical signal includes but is not limited to performing filtering by using a fourth-order Bessel filter whose 3-dB bandwidth is half of a baud rate.

It should be noted that, the TDEC test receiver may be specifically used as a large instrument device such as an eye pattern analyzer, or may alternatively be used as a small-sized product dedicated to the TDEC test, to improve test efficiency. The method for testing a TDEC provided in this application is mainly implemented by the processor in the TDEC test receiver. A specific structure of the processor is not limited in this application. The following describes in detail the method for testing a TDEC provided in this application.

FIG. 3 is a diagram of an embodiment of a method for testing a TDEC according to an embodiment of this application. In this example, the method for testing a TDEC includes the following steps.

301: Obtain a to-be-tested signal. In this embodiment, a signal transmitted by a transmitter is referred to as an original signal, and the transmitter may periodically and repeatedly send the original signal. The original signal is an optical signal obtained by modulating a generated test sequence by the transmitter. A manner of modulating the test sequence includes but is not limited to PAM-4, PAM-2, and the like. The PAM-2 may also be referred to as OOK. For example, the test sequence may be a short stress pattern random quaternary (SSPRQ), and an SSPRQ pattern is one of PAM-4 patterns. It should be understood that the SSPRQ is formed by splicing and encoding four patterns that are selected from a PRBS31 pattern and that have great pressure on the transmitter, and a length of the SSPRQ is 2¹⁶−1. Advantages of the SSPRQ are that enough pressure can be applied to the tested transmitter, so that this case is closer to testing performance of the tested transmitter in a real service, and the SSPRQ has a feature of a short pattern, so that a sampling oscilloscope can capture an entire pattern for signal processing such as equalization. Refer to descriptions in FIG. 2. The original signal transmitted by the transmitter is received by a TDEC test receiver through channel transmission, and a front end of the TDEC test receiver processes the signal, and then transmits, to a processor, a signal obtained through processing, where the signal received by the processor is referred to as the to-be-tested signal.

302: Perform equalization processing on the to-be-tested signal. This application provides a target equalizer, configured to perform equalization processing on the to-be-tested signal. Specifically, the target equalizer includes an FFE and a DFE. For ease of description, in this application, equalization processing performed by the FFE is referred to as first equalization processing, and a tap coefficient of the FFE is referred to as a first tap coefficient; and equalization processing performed by the DFE is referred to as second equalization processing, and a tap coefficient of the DFE is referred to as a second tap coefficient.

FIG. 4 is a diagram of equalization processing according to an embodiment of this application. As shown in FIG. 4, a main tap of an FFE is denoted as r_L, and Z⁰ represents no delay. Front taps of the FFE are of a quantity N1, and are respectively denoted as r_L−N1, . . . , and r_L−1, and Z^−N, . . . , Z⁻², and Z¹ respectively represent delays of −1, −2, . . . , and −N1 symbol periodicities. Rear taps of the FFE are of a quantity N2, and are respectively denoted as r_L+1, . . . , and r_L+N2, and Z⁺¹, Z⁺², . . . , and Z^+N2 respectively represent delays of 1, 2, . . . , and N2 symbol periodicities. Both N1 and N2 are integers greater than or equal to 0. Specific values of N1 and N2 depend on an implementation requirement, and are not limited in this application. In an example, N1+N2=4 or 20. A first tap coefficient of the FFE is denoted as w_FFEⁱ, where i is an integer between −N1 and N2. It should be understood that the FFE is at an integer interval, to be specific, a signal interval corresponding to adjacent taps in the FFE is one symbol periodicity. A quantity of taps of a DFE is 1, and a second tap coefficient of the DFE is denoted as w_DFE. It should be noted that, a difference obtained by subtracting the second tap coefficient of the DFE from a sum of all first tap coefficients of the FFE is 1, that is, Σ_i=−N1^N2w_FFEⁱ−w_DFE=1.

Specifically, the FFE performs first equalization processing on a to-be-tested signal based on the first tap coefficient to obtain a first signal. For example, as shown in FIG. 4, the to-be-tested signal is denoted as r, and the FFE performs delay weighted superposition on the to-be-tested signal to obtain a signal Σ_i=−N1^N2w_FFEⁱ×r_L+i. Delay amounts are respectively −N₁, . . . , −1, 0, 1, . . . , and N₂ symbol periodicities, and a weight is the first tap coefficient of the FFE, namely, WPFE. The DFE calculates a product of a level d[L−1] corresponding to an original signal and the second tap coefficient w_DFE, that is, obtains an estimation of inter-symbol interference delayed in time, and performs second equalization processing on the first signal based on the product to obtain a second signal. The second signal is a signal output by performing equalization processing by a target equalizer. In an example, the DFE obtains, by calculating the product of d[L−1] and w_DFE, the inter-symbol interference (ISI) delayed in the time, and further performs ISI elimination on the first signal to obtain the second signal. It should be understood that the level corresponding to the original signal is represented as

$P_{\mathrm{ave}}+\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{M-1}\left(i-\frac{M+1}{2}\right),$

where P_ave represents average optical power of an eye pattern, OMA_outer represents an outer optical modulation amplitude of the second signal, 1≤i≤M, M represents a quantity of level values of the original signal, and M≥2. For example, if M=2, the original signal uses PAM-2; or if M=4, the original signal uses PAM-4.

It should be noted that the transmitter and the TDEC test receiver may agree on a test sequence used by the original signal, so that the DFE calculates the product of the level d[L−1] corresponding to the original signal and the second tap coefficient w_DFE. It should be understood that the equalization processing performed by the DFE provided in this application is different from other methods for performing equalization processing by a DFE. Impact of DFE error propagation is not considered, and the DFE provided in this application estimates the ISI by using the original signal instead of a decision feedback symbol. The DFE error propagation is explained as follows: The DFE calculates the ISI based on a symbol output through decision and performs equalization based on a calculation result. Once the symbol output through decision is erroneous, the DFE calculates the ISI erroneously based on the erroneously decided symbol, and cannot correctly eliminate the ISI. Consequently, a probability of an error decision of a symbol at a next moment is increased. This phenomenon that an erroneous symbol causes an error in a subsequent symbol is referred to as error propagation.

303: Obtain an eye pattern of a signal obtained through the equalization processing, and perform sampling at a target position of the eye pattern.

In an example, a method for calculating a xUI sampling point of an L^th symbol within a periodicity is as the following formula (1):

$\mathrm{Eq}\left(r_{L,x}\right)=\sum _{i=-N_1}^{N_2}{w}_{\mathrm{FFE}}^i\times r_{L+i,x}-w_{\mathrm{DFE}}\times d\left[L-1\right]$

r_L,x represents a value of the xUI sampling point of the received L^th symbol within the periodicity, 0<x<1, and d[L−1] is a level corresponding to an (L−1)^th symbol of the original signal. An example in which the original signal uses the PAM-4 is used. Four levels corresponding to the original signal are sequentially

$P_{\mathrm{ave}}-\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{2},P_{\mathrm{ave}}-\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{6},P_{\mathrm{ave}}+\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{6},\mathrm{and}{P}_{\mathrm{ave}}+\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{2}$

in ascending order.

In this embodiment, an eye pattern of the second signal obtained through the equalization processing is sampled to obtain a third signal. For example, a position of an inter-symbol crosspoint is first found based on average optical power, and the target position for the sampling is further determined. In an example, the target position of the eye pattern includes a

$\left(0.5-\frac{\Delta }{2}\right)$

unit interval (Unit interval, UI) and a

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI}.$

Δ is an interval between two groups of samples, is at a unit of UI, has a value ranging from 0 to 1, and is usually about 0.1. It should be noted that a sampling moment may also be adjusted slightly herein to minimize the TDEC, but an interval of a Δ UI may need to be ensured. For example, the sampling is performed at a

$\left(0.49-\frac{\Delta }{2}\right)\mathrm{UI}\mathrm{and}a\left(0.49+\frac{\Delta }{2}\right)\mathrm{UI}$

as adjusted. It should be understood that a horizontal width of each sampling window is a δ UI, 0<δ<0.5, and δ is usually about 0.04. For a vertical width, it may need to be ensured that the window includes a boundary of a waveform. A purpose of the sampling is to obtain two groups of level histograms, superpose target noise on the two groups of level histograms, and then separately analyze TDEC performance. One with poorer performance is a final TDEC indicator.

304: Estimate an initial SER obtained by superposing the target noise. In this embodiment, the initial SER obtained by superposing the target noise to the third signal may need to be first determined, and the initial SER is further modified to obtain a target SER. It should be understood that the initial SER is a SER without consideration of the impact of the DFE error propagation, and the target SER is a SER in consideration of the impact of the DFE error propagation. It should be noted that, in this application, a bit error rate (BER) may alternatively be used to replace the SER for representation, and the SER and the BER may be converted into each other. A person skilled in the art may know a specific manner, and details are not described herein.

The following describes a specific implementation of determining the initial SER by using an example in which the original signal uses the PAM-4.

A decision threshold of a lower eye in the eye pattern is denoted as P_th1, a decision threshold of a middle eye in the eye pattern is denoted as P_th2, and a decision threshold of an upper eye in the eye pattern is denoted as P_th3. P_th1, P_th2, and P_th3 satisfy the following formula (2):

$P_{\mathrm{th}1}=P_{\mathrm{ave}}-\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{3}$ $P_{\mathrm{th}2}=P_{\mathrm{ave}}$ $P_{\mathrm{th}3}=P_{\mathrm{ave}}+\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{3}$

Levels sampled at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}\mathrm{and}\mathrm{the}\left(0.5+\frac{\Delta }{2}\right)$

UI are normalized. Values of the levels are grouped into a plurality of ranges, an interval is Δy, a center of each range is y_i, and frequencies F(y_i) of the ranges are calculated. It is clear that a sum of the frequencies of different intervals is 1. For two groups of the levels sampled at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}$

and sampled at the

$\left(0.5+\frac{\Delta }{2}\right)$

cumulative probability functions for the eyes are calculated, and are respectively denoted as CF_L1(y_i), CF_L2y_i), CF_L3(y_i), CF_R1(y_i), CF_R2(y_i), and CF_R3(y_i), where L represents the sampling on a left side (namely, at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}),$

R represents the sampling on a right side (namely, at the

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI}),$

and subscript symbols 1, 2, and 3 respectively represent the lower eye, the middle eye, and the upper eye. A specific calculation manner is described by using CF_L1(y_i) as an example. A calculation manner of another eye is similar, and CF_L1(y_i) satisfies the following formula (3):

${\mathrm{CF}}_{L1}\left(y_i\right)=\{\begin{array}{c}\sum _{y=P_{\mathrm{th}1}}^{y_i}F\left(y\right)\mathrm{for}{y}_i\ge P_{\mathrm{th}1}\\ \sum _{y=y_i}^{P_{\mathrm{th}1}}F\left(y\right)\mathrm{for}{y}_i<P_{\mathrm{th}1}\end{array}$

y represents a sampling level value of the third signal, F(y) represents a frequency distribution function of the sampling levels of the third signal, and y_i is a representation of discretized y.

Then, SER_L1 of the lower eye is obtained through calculation by superimposing the target noise on the signal sampled at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}.$

A specific metnou is: multiplying each element of the cumulative probability function CF_L1(y_i) by G_th1(y_i) and then performing summation, to obtain SER_L1. Similarly, SER_L2 and SER_L3 are obtained, and a sum of SER_L1, SER_L2, and SER_L3 is SER_L. In the same way, SER_R obtained by superimposing the noise on the level sampled at the

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI}$

may be calculated. G_th1(y_i) is equivalent to a Gaussian probability density function using a root mean square (root mean square, RMS) value of the target noise, and has a center located at the lower eye threshold P_th1. G_th1(y_i) satisfies the following formula (4):

$G_{\mathrm{th}1}\left(y_i\right)={\int }_{y_i-\frac{\Delta y}{2}}^{y_i+\frac{\Delta y}{2}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-P_{\mathrm{th}1}}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$ $\mathrm{or}$ $G_{\mathrm{th}1}\left(y_i\right)=\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y_{i-P_{\mathrm{th}1}}}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\times \Delta y$

σ_G represents the RMS value of the target noise, C_eq represents a coefficient of noise enhancement through the first equalization processing and the second equalization processing, and Δy represents an interval between two adjacent y_i. Similarly, G_th2(y_i) and G_th3(y_i) may be obtained through calculation in a manner similar to Formula (4). A calculation method of C_eq is a product of normalized noise power spectral density N(f) at an input of a reference equalizer and a normalized frequency response H_eq(f) of the reference equalizer, and C_eq satisfies the following formula (5):

$C_{\mathrm{eq}}=\sqrt{\int N\left(f\right)\times {❘H_{\mathrm{eq}}\left(f\right)❘}^2\mathrm{df}}$

N(f) is a normalized noise power spectrum obtained by filtering white noise via a filter. N(f) satisfies the following formula (6):

$\int N\left(f\right)\mathrm{df}=H_{\mathrm{eq}}\left(f=0\right)=1$

305: Modify the initial SER to obtain the target SER. In this embodiment, the target SER may be obtained by multiplying the initial SER by a modification factor f_modify. The modification factor f_modify may be considered as a SER increase multiple caused by the DFE error propagation. It should be understood that, for the signal sampled at the

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}$

and the signal sampled at the

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI},$

two initial SERs may be modified in the same calculation manner to obtain two target SERs, and a larger one of the two target SERs is selected as a final target SER.

$\left(0.5-\frac{\Delta }{2}\right)\mathrm{UI}$

may be denoted as f_modify^L. For another example, a modification factor corresponding to the signal sampled at the

$\left(0.5+\frac{\Delta }{2}\right)\mathrm{UI}$

may be denoted as fint Imodify. For ease of description, f_modify is uniformly used to represent the modification factor below, and L and R are not distinguished.

The following describes several specific calculation manners of the modification factor.

In a first calculation manner of the modification factor f_modify, the modification factor f_modify satisfies the following formula (7):

$f_{\mathrm{modify}}=\frac{1}{1-\left(P_E-SE{R}_0\right)}$

SER₀ represents the initial SER, and P_E represents a probability that a second symbol is erroneously decided due to a decision error of a first symbol in the second equalization processing, where the first symbol is received by the TDEC test receiver before the second symbol. That is, because the symbol is erroneously decided currently, and the DFE equalizer performs erroneous ISI elimination, a probability of error decision of a symbol at a next moment is denoted as P_E.

FIG. 5A is a diagram of a first estimation effect of a SER increase multiple caused by DFE error propagation. FIG. 5B is a diagram of a second estimation effect of a SER increase multiple caused by DFE error propagation. As shown in FIG. 5A and FIG. 5B, a horizontal coordinate represents a signal-to-noise ratio (SNR), and a vertical coordinate represents the SER increase multiple caused by the DFE error propagation. FIG. 5A and FIG. 5B both show a comparison between modification factors f_modify obtained through calculation and simulation results of SER increase multiples caused by the DFE error propagation under different value conditions of w_DFE. It can be learned that the modification factors f_modify obtained in the first calculation manner are basically the same as the simulation results. The simulation results of the SER increase multiples caused by the DFE error propagation may be specifically calculated by using a Monte Carlo method.

In a second calculation manner of the modification factor f_modify, if a degree of linearity of the transmitter is good, noise levels are evenly distributed, and a BER threshold requirement is low, the modification factor f_modify satisfies the following formula (8):

$f_{\mathrm{modify}}=\frac{1}{1-P_E}$

It should be understood that, the second calculation manner of the modification factor f_modify has lower calculation complexity, but has a requirement on an application scope, and is usually applicable only to a case in which the BER threshold is below a −3 order of magnitude.

FIG. A is a diagram of a third estimation effect of a SER increase multiple caused by DFE error propagation. FIG. 6B is a diagram of a fourth estimation effect of a SER increase multiple caused by DFE error propagation. As shown in FIG. 6A and FIG. 6B, a horizontal coordinate represents an SNR, and a vertical coordinate represents the SER increase multiple caused by the DFE error propagation. FIG. 6A and FIG. 6B both show a comparison between modification factors f_modify obtained through calculation and simulation results of SER increase multiples caused by the DFE error propagation under different value conditions of w_DFE. It can be learned that the modification factors f_modify obtained in the second calculation manner are basically the same as the simulation results. The simulation results of the SER increase multiples caused by the DFE error propagation may be specifically calculated by using a Monte Carlo method.

It should be noted that, P_E is determined by the second tap coefficient w_DFE of the DFE, the RMS value σ_G of the target noise, and a sampling level-related parameter. It should be understood that the target noise is in Gaussian distribution, and the RMS value of the target noise is denoted as σ_G. The sampling level-related parameter includes at least one of frequency distribution of the sampling levels of the third signal, the average optical power of the eye pattern, and the outer optical modulation amplitude of the second signal. This application provides a plurality of specific calculation manners of P_E, which are separately described below.

In a first calculation manner of P_E, P_E satisfies the following formula (9):

$P_E=\sum _{i=1}^K{P}_E^k$

K is a quantity of eyes in the eye pattern, P_E^k represents a probability that the second symbol is erroneously decided at a k^th eye, and 1≤k≤K. An example in which the original signal uses the PAM-4 is used. K=3, and the eye pattern includes the lower eye, the middle eye, and the upper eye. k=1, 2, or 3, and values of k respectively correspond to the lower eye, the middle eye, and the upper eye.

Further, P_E^k satisfies the following formula (10):

$P_E^k=\left(\int C{F}_k\left(y\right)G_{thk}^{\mathrm{shift}+}\left(y\right)\mathrm{dy}+\int C{F}_k\left(y\right)G_{thk}^{\mathrm{shift}-}\left(y\right)\mathrm{dy}\right)/2$

CF_k(y) represents a cumulative probability function of level distribution corresponding to the k^th eye, G_thk^shift+(y) represents a probability density function of superimposed noise obtained by offsetting a decision threshold of the k^th eye in a first direction, and G_thk^shift−(y) represents a probability density function of superimposed noise obtained by offsetting the decision threshold of the k^th eye in a second direction, the first direction is opposite to the second direction, and y represents a sampling level value of the third signal.

Further, CF_k(y) satisfies the following formula (11):

$C{F}_k\left(y_i\right)=\{\begin{array}{c}\sum _{y=P_{\mathrm{thk}}}^{y_i}F\left(y\right)\mathrm{for}{y}_i\ge P_{\mathrm{thk}}\\ \sum _{y=y_i}^{P_{\mathrm{thk}}}F\left(y\right)\mathrm{for}{y}_i<P_{\mathrm{thk}}\end{array}$

P_thk is a decision threshold of the k^th eye, F(y) represents the frequency distribution function of the sampling levels of the third signal, and y_i is the representation of discretized y.

G_thk^shift+(y) and G_thk^shift−(y) satisfy the following formula (12):

$G_{\mathrm{thk}}^{\mathrm{shift}+}\left(y_i\right)={\int }_{y_{i-\frac{\Delta y}{2}}}^{y_{i+\frac{\Delta y}{2}}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{\mathrm{thk}}+w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{K}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$ $G_{\mathrm{thk}}^{\mathrm{shift}-}\left(y_i\right)={\int }_{y_{i-\frac{\Delta y}{2}}}^{y_{i+\frac{\Delta y}{2}}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{\mathrm{thk}}-w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{K}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$

w_DFE represents the second tap coefficient, OMA_outer represents the outer optical modulation amplitude of the second signal, C_eq represents a coefficient of noise enhancement through the first equalization processing and the second equalization processing, σ_G represents the RMS value of the target noise, and Δy represents an interval between two adjacent y_i. It should be understood that

$w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{K}$

represents an offset value of a level offset that is of the second symbol and that is caused by the DFE error propagation. This is equivalent to a case in which a decision threshold is offset while a level position remains unchanged.

It should be noted that the foregoing first calculation manner of P_E is applicable to different scenarios in which the original signal uses the PAM-4, the PAM-2, or the like.

In an example, in the scenario in which the original signal uses the PAM-4, P_E includes three SER components respectively corresponding to a lower eye error, a middle eye error, and an upper eye error, that is, K=3, and k=1, 2, or 3. P_E=P_E¹+P_E²+P_E³, where P_E¹, P_E², and P_E³ respectively represent the three SER components. The lower eye is used as an example. G_th1^shift+(y) represents a probability density function of superimposed noise obtained by offsetting a decision threshold of the lower eye in the first direction, and G_th1^shift−(y) represents a probability density function of superimposed noise obtained by offsetting the decision threshold of the lower eye in second direction. G_th1^shift+(y) and G_th1^shift−(y) satisfy the following formula (13):

$G_{th1}^{\mathrm{shift}+}\left(y_i\right)={\int }_{y_{i-\frac{\Delta y}{2}}}^{y_{i+\frac{\Delta y}{2}}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{th1}+w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{3}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$ $G_{th1}^{\mathrm{shift}-}\left(y_i\right)={\int }_{y_{i-\frac{\Delta y}{2}}}^{y_{i+\frac{\Delta y}{2}}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{th1}-w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{3}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$

Alternatively, G_th1^shift+(y) and G_th1^shift−(y) satisfy the following formula (14):

$G_{th1}^{\mathrm{shift}+}\left(y_i\right)=\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y_i-\left(P_{th1}+w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{3}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\times \Delta y$ $G_{th1}^{\mathrm{shift}-}\left(y_i\right)=\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y_i-\left(P_{th1}-w_{\mathrm{DFE}}\times \frac{{\mathrm{OMA}}_{\mathrm{outer}}}{3}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\times \Delta y$

In another example, in the scenario in which the original signal uses the PAM-2, an eye pattern of a PAM-2 signal includes only one eye, and therefore P_E does not have a plurality of SER components. The eye in the eye pattern is used as an example. G_thk^shift+(y) represents a probability density function of superimposed noise obtained by offsetting a decision threshold of the eye in the first direction, and G_th^shift−(y) represents a probability density function of superimposed noise obtained by offsetting the decision threshold of the eye in the second direction. G_th^shift+(y) and G_th^shift−(y) satisfy the following formula (15):

$G_{th}^{\mathrm{shift}+}\left(y_i\right)={\int }_{y_{i-\frac{\Delta y}{2}}}^{y_{i+\frac{\Delta y}{2}}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{th}+w_{\mathrm{DFE}}\times {\mathrm{OMA}}_{\mathrm{outer}}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$ $G_{\mathrm{th}}^{\mathrm{shift}-}\left(y_i\right)={\int }_{y_{i-\frac{\Delta y}{2}}}^{y_{i+\frac{\Delta y}{2}}}\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y-\left(P_{th}-w_{\mathrm{DFE}}\times {\mathrm{OMA}}_{\mathrm{outer}}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\mathrm{dy}$

Alternatively, G_th^shift+(y) and G_th^shift−(y) satisfy the following formula (16):

$G_{\mathrm{th}}^{\mathrm{shift}+}\left(y_i\right)=\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y_i-\left(P_{\mathrm{th}}+w_{\mathrm{DFE}}\times {\mathrm{OMA}}_{\mathrm{outer}}\right)}{c_{\mathrm{eq}}{\sigma }_G\sqrt{2}}\right)}^2}\times \Delta y$ $G_{\mathrm{th}}^{\mathrm{shift}-}\left(y_i\right)=\frac{1}{C_{\mathrm{eq}}{\sigma }_G\sqrt{2\pi }}\times e^{-{\left(\frac{y_i-\left(P_{\mathrm{th}}-w_{\mathrm{DFE}}\times {\mathrm{OMA}}_{\mathrm{outer}}\right)}{c_{eq}{\sigma }_G\sqrt{2}}\right)}^2}\times \Delta y$

In a second calculation manner of P_E, P_E satisfies the following formula (17):

$P_E=\{\begin{array}{c}\begin{array}{c}0.5\times \mathrm{erfc}\left(\frac{1-2w_{\mathrm{DFE}}}{\sqrt{2}}\times \sqrt{\frac{3\times \mathrm{SNR}}{\left(M+1\right)\times \left(M-1\right)}}\right)\times \\ \left(1-\frac{1}{M}\right),w_{\mathrm{DFE}}\le 0.5\end{array}\\ \begin{array}{c}\left(1-0.5\times \mathrm{erfc}\left(\frac{2w_{\mathrm{DFE}}-1}{\sqrt{2}}\times \sqrt{\frac{3\times \mathrm{SNR}}{\left(M+1\right)\times \left(M-1\right)}}\right)\right)\times \\ \left(1-\frac{1}{M}\right),0.5<w_{\mathrm{DFE}}\le 1\end{array}\end{array}$

erfc represents a complementary error function (complementary error function), the SNR represents a signal-to-noise ratio obtained by superimposing the target noise on the third signal, w_DFE represents the second tap coefficient, M represents the quantity of level values of the original signal, and M≥2. It should be understood that

$\mathrm{erfc}\left(x\right)=\frac{2}{\sqrt{\pi }}{\int }_x^{\infty }{e}^{-{\eta }^2}d\eta .$

Further, the SNR satisfies the following formula (18):

$\mathrm{SNR}=\frac{P_{\mathrm{sig}}}{{\sum }_{y_i}F\left(y_i\right)\times {\left(y_i-P_{ave}\right)}^2-P_{sig}+{\sigma }_{G^2}}$

P_sig represents power of a signal component without noise and distortion in the third signal, y_i is a representation of a discretized sampling level value of the third signal, P_ave represents the average optical power of the eye pattern, and σ_G represents the RMS value of the target noise.

Further, P_sig satisfies the following formula (19):

$P_{sig}=\frac{\left(M+1\right)\times \left(M-1\right)}{3}\times {\left(\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{2\times \left(M-1\right)}\right)}^2$

OMA_outer represents the outer optical modulation amplitude of the second signal, M represents the quantity of level values of the original signal, and M≥2.

It should be noted that the foregoing described two calculation manners of P_E is applicable to any one of the foregoing described calculation manners of the modification factor f_modify, and the modification factor f_modify obtained through calculation is accurate and reliable. Details may be specifically shown in FIG. 5A or FIG. 6B. In addition, compared with the first calculation manner of P_E, the second calculation manner of P_E reduces calculation complexity. However, when the degree of linearity of the transmitter is poor, and the noise levels are not evenly distributed, an error in estimating the SER increase multiple in the second calculation manner of P_E is greater than that in the first calculation manner of P_E.

306: Determine whether the target SER is equal to a SER threshold. If the target SER is not equal to the SER threshold, perform step 307; or if the target SER is equal to the SER threshold, perform step 308.

It should be noted that a specific value of the SER threshold depends on actual application, and is not limited in this application.

307: Adjust the target noise. If the target SER is not equal to the SER threshold, the target noise superimposed on the third signal may need to be adjusted, so that the target SER is re-determined according to the foregoing step 305, to cause a newly determined target SER to be equal to the SER threshold. For example, if the current target SER is greater than the SER threshold, the target noise superimposed on the third signal is reduced. For another example, if the current target SER is less than the SER threshold, the target noise superimposed on the third signal is increased.

It should be noted that, if the target noise is adjusted, the first tap coefficient and the second tap coefficient also change. In a possible implementation, the first tap coefficient and the second tap coefficient may be iterated based on adjusted target noise, to determine an optimal first tap coefficient and an optimal second tap coefficient. For example, if the SER of the second signal output through the equalization processing is the smallest, a first tap coefficient and a second tap coefficient in this state are optimal.

308: Calculate the TDEC based on the adjusted target noise. If the target SER is equal to the SER threshold, the TDEC is calculated based on the currently determined target noise. Specifically, total RMS noise is first calculated, and the RMS noise satisfies the following formula (20):

$R=\sqrt{{\sigma }_G^2+{\sigma }_S^2}$

R represents the total RMS noise, σ_G is the RMS value of the target noise, and σ_s is a noise standard deviation of background noise of the TDEC test receiver.

Further, the PAM-4 is used as an example. The TDEC satisfies the following formula (21):

$\mathrm{TDEC}=10{\log }_{10}\left(\frac{{\mathrm{OMA}}_{\mathrm{outer}}}{6}\times \frac{1}{Q_{t}R}\right)$

Further, based on a specific bit error threshold, Q_t satisfies the following formula (22):

$Q_t={\mathrm{erfc}}^{-1}\left(\frac{8}{3}\times \mathrm{BER}\right)\times \sqrt{2}$

erfc represents the complementary error function, and BER represents a BER obtained by superimposing the target noise on the third signal.

The following lists several Q_t values corresponding to the BER threshold, as shown in Table 1 below. It should be understood that, the BER threshold may be selected from Table 1 below based on an actual requirement and converted to obtain the SER threshold mentioned in step 306 above.

TABLE 1
BER threshold	2.4e−4	2e−3	4.5e−3	6e−3	8.5e−3	1.1e−2
Qt	3.414	2.786	2.512	2.409	2.279	2.179

It can be learned from the foregoing descriptions that, in this application, the DFE is added to the TDEC test receiver, to improve an equalization capability, so that a high baud rate system is better supported. In addition, this application provides a method for estimating a target SER obtained through equalization processing. DFE error propagation is not first considered, and a DFE performs equalization processing by using the original signal, to obtain an initial SER without consideration of the error propagation of the DFE, and then calculates a modification factor and modifies the initial SER based on the modification factor to obtain the target SER. It should be understood that the modification factor may be considered as a SER increase multiple caused by the DFE error propagation, so that a real estimation of the SER obtained through the DFE equalization processing can be obtained. In this manner, under a condition in which a length of a test sequence of the original signal is limited, low operational complexity of estimating the SER is implemented, so that a scenario with severe signal impairment is better applied.

The foregoing describes the method for testing a TDEC provided in this embodiment of this application. The following describes a TDEC test receiver provided in an embodiment of this application.

FIG. 7 is a diagram of a structure of a TDEC test receiver according to an embodiment of this application. As shown in FIG. 7, the TDEC test receiver includes a receiver front end 901 and a processing chip 902. The receiver front end 901 and the processing chip 902 are connected to each other through a line. The receiver front end 901 is configured to: convert a received optical signal into an electrical signal, and transmit the electrical signal to the processing chip 902. The processing chip 902 is configured to perform method steps in the embodiment shown in FIG. 3. In a possible implementation, the receiver front end 901 includes the polarization rotator 10, the optical splitter 20, the tunable reflector 30, the test fiber 40, the photoelectric conversion unit 50, the clock recovery unit 60, and the oscilloscope 70 shown in FIG. 2. For details, refer to related descriptions in FIG. 2. Details are not described herein again.

An embodiment of this application further provides a processing chip. A processor and one or more interfaces that are configured to implement the method steps in the embodiment shown in FIG. 3 are integrated into the processing chip. When a memory is integrated into the processing chip, the processing chip may complete the method steps in any one or more of the foregoing embodiments. When no memory is integrated into the processing chip, the digital processing chip may be connected to an external memory through an interface. The processing chip implements, based on program code stored in the external memory, actions performed by the receiver in the foregoing embodiments.

A person of ordinary skill in the art may understand that all or some of the steps in the foregoing embodiments may be implemented through hardware or a program instructing related hardware. The program may be stored in a computer-readable storage medium. The storage medium above may be a read-only memory (ROM), a random-access memory (RAM), or the like. Specifically, for example, the foregoing processing unit or processor may be a central processing unit, a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or another programmable logic device, a transistor logic device, a hardware component, or any combination thereof. Whether the functions are performed by hardware or software depends on particular applications and design constraint conditions of the technical solutions. A person skilled in the art may use different methods to implement the described functions for each particular application, but it should not be considered that the implementation goes beyond the scope of this application.

When software is used to implement the functions, all or some of the method steps described in the foregoing embodiments may be implemented in a form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the procedure or functions according to embodiments of this application are all or partially generated. The computer may be a general-purpose computer, a dedicated computer, a computer network, or another programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or may be transmitted from a computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website, computer, server, or data center to another website, computer, server, or data center in a wired (for example, a coaxial cable, an optical fiber, or a digital subscriber line (DSL)) or wireless (for example, infrared, radio, and microwave, or the like) manner. The computer-readable storage medium may be any usable medium accessible by the computer, or a data storage device, for example, a server or a data center, integrating one or more usable media. The usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, or a magnetic tape), an optical medium (for example, a digital versatile disc (DVD)), a semi-conductor medium (for example, a solid-state disk (SSD)), or the like.

Finally, it should be noted that the foregoing descriptions are merely specific implementations of this application, but are not intended to limit the protection scope of this application. Any variation or replacement readily figured out by a person skilled in the art within the technical scope disclosed in this application shall fall within the protection scope of this application. Therefore, the protection scope of this application shall be subject to the protection scope of the claims.

Claims

1. A method for testing a transmitter and dispersion eye closure (TDEC), the method comprising: obtaining a to-be-tested signal that is based on channel transmission of an original signal;performing first equalization processing on the to-be-tested signal based on at least one first tap coefficient to obtain a first signal;obtaining a product of a level corresponding to the original signal and a second tap coefficient;performing second equalization processing on the first signal based on the product to obtain a second signal, wherein a difference between the second tap coefficient and a sum of all the first tap coefficients is 1;obtaining an eye pattern of the second signal;performing sampling at a target position of the eye pattern to obtain a third signal;obtaining a target symbol error rate (SER) by superimposing target noise on the third signal;adjusting the target noise in order to adjust the target SER and obtain an adjusted target noise; andcalculating the TDEC based on the adjusted target noise when the target SER is equal to an SER threshold.

2. The method according to claim 1, wherein obtaining the target SER by superimposing the target noise on the third signal comprises: obtaining an initial SER by superimposing the target noise on the third signal; andmodifying the initial SER based on a modification factor in order to obtain the target SER.

3. The method according to claim 2, wherein modifying the initial SER based on the modification factor in order to obtain the target SER comprises multiplying the initial SER by the modification factor in order to obtain the target SER.

4. The method according to claim 2, wherein the modification factor satisfies the following formula:

5. The method according to claim 2, wherein the modification factor satisfies the following formula:

6. The method according to claim 4, wherein PE is based on the second tap coefficient, a root mean square (RMS) value of the target noise, and a sampling level-related parameter, and wherein the sampling level-related parameter comprises at least one of a frequency distribution of sampling levels of the third signal, an average optical power of the eye pattern, or an outer optical modulation amplitude of the second signal.

7. The method according to claim 4, wherein PE satisfies the following formula:

8. The method according to claim 7, wherein PEk satisfies the following formula:

9. The method according to claim 8, wherein CFk(y) satisfies the following formula:

10. The method according to claim 4, wherein PE satisfies the following formula:

11. An apparatus, comprising: a receiver front end configured to: convert a received optical signal into an electrical signal; andtransmit the electrical signal; anda processing chip coupled to the receiver front end and configured to: receive, from the receiver front end, the electrical signal, wherein the electrical signal is a to-be-tested signal;perform first equalization processing on the to-be-tested signal based on at least one first tap coefficient to obtain a first signal;obtain a product of a level corresponding to an original signal of the received optical signal and a second tap coefficient;perform second equalization processing on the first signal based on the product to obtain a second signal, wherein a difference between the second tap coefficient and a sum of all the first tap coefficients is 1;obtain an eye pattern of the second signal;perform sampling at a target position of the eye pattern to obtain a third signal;obtain a target symbol error rate (SER) by superimposing target noise on the third signal;adjust the target noise in order to adjust the target SER and obtain an adjusted target noise; andcalculate a transmitter and dispersion eye closure (TDEC) based on the adjusted target noise when the target SER is equal to an SER threshold.

12. The apparatus according to claim 11, wherein the processing chip is further configured to: obtain an initial SER by superimposing the target noise on the third signal; andmodify the initial SER based on a modification factor in order to obtain the target SER.

13. The apparatus according to claim 12, wherein the processing chip is further configured to multiply the initial SER by the modification factor in order to obtain the target SER.

14. The apparatus according to claim 12, wherein the modification factor satisfies the following formula:

15. The apparatus according to claim 12, wherein the modification factor satisfies the following formula:

16. The apparatus according to claim 14, wherein PE is based on the second tap coefficient, a root mean square (RMS) value of the target noise, and a sampling level-related parameter, and wherein the sampling level-related parameter comprises at least one of a frequency distribution of sampling levels of the third signal, an average optical power of the eye pattern, and an outer optical modulation amplitude of the second signal.

17. The apparatus according to claim 14, wherein PE satisfies the following formula:

18. The apparatus according to claim 17, wherein PEk satisfies the following formula:

19. The apparatus according to claim 18, wherein CFk(y) satisfies the following formula:

20. The apparatus according to claim 14, wherein PE satisfies the following formula: