Decision feedback coefficient training method and device, electronic equipment and electronic equipment

Abstract

The invention provides a decision feedback coefficient training method and device, electronic equipment and electronic equipment, and relates to the technical field of communication, and the method comprises the steps: obtaining a current error signal corresponding to a current receiving signal entering a DFE; averaging the current error signal to obtain a current average signal; and updating the current tap coefficient of the DFE according to the current mean value signal. As the mean value of the noise is 0, the noise can be reduced and the signal-to-noise ratio can be improved by processing the error signal through the mean value method, so that the fluctuation of the DFE tap coefficient is reduced, and the stability and the reliability of the DFE tap coefficient are improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a decision feedback coefficient training method, device, electronic equipment and electronic equipment. Background Art

[0002] When data passes through a high-speed channel, inter-symbol interference (ISI) will occur. In order to recover the data, the receiver generally needs a DFE (Decision Feedback Equalization). The traditional decision feedback coefficient training method is the LMS (Least Mean Squares) method, which continuously iterates the DFE tap coefficient (i.e., decision feedback coefficient) by descending along the tangent direction according to the calculated error.

[0003] The traditional decision feedback coefficient training method is sensitive to noise. When the noise is large, the tap coefficient obtained by training will deviate from the ideal value, and the greater the noise, the greater the fluctuation of the tap coefficient. Summary of the invention

[0004] The object of the present invention is to provide a decision feedback coefficient training method, device, electronic device and electronic device to reduce noise, improve signal-to-noise ratio, and increase the stability and reliability of DFE tap coefficients.

[0005] In a first aspect, the present invention provides a decision feedback coefficient training method, which is applied to DFE. The method comprises:

[0006] Obtaining a current error signal corresponding to a current received signal entering the DFE;

[0007] Calculate the average of the current error signal to obtain the current average signal;

[0008] According to the current mean value signal, the current tap coefficients of the DFE are updated.

[0009] Furthermore, obtaining a current error signal corresponding to a current received signal entering the DFE includes:

[0010] Determine a current equalization signal and a current decision signal according to a current received signal;

[0011] A current error signal is obtained by calculation according to the current equalization signal and the current decision signal.

[0012] Further, according to the current received signal, determining the current equalization signal and the current decision signal includes:

[0013] According to the current tap coefficient of the DFE and the historical decision signal output by the DFE, the current received signal is processed by the equalizer of the DFE to obtain a current equalized signal;

[0014] The current equalization signal is judged by the judgement device of the DFE to obtain a current judgment signal.

[0015] Furthermore, according to the current equalization signal and the current decision signal, a current error signal is calculated, including:

[0016] According to the preset main calibration coefficient corresponding to the current received signal, the current error signal is calculated by the following formula:

[0017] e(k)=u(k)-s(k)*a0;

[0018] Among them, e(k) represents the error signal at time k, u(k) represents the equalization signal at time k, s(k) represents the decision signal at time k, and a0 represents the main coefficient.

[0019] Furthermore, the current error signal is averaged to obtain a current mean signal, including:

[0020] Determine the current filter coefficient corresponding to the current training stage;

[0021] According to the current filter coefficient, the current mean signal is calculated by the following formula:

[0022] f(k)=(1-P(k))f(k-1)+P(k)e(k);

[0023] Among them, f(k) represents the mean signal at time k, f(k-1) represents the mean signal at time k-1, P(k) represents the filter coefficient at time k, and e(k) represents the error signal at time k.

[0024] Furthermore, the current filter coefficient is determined by the following formula:

[0025]

[0026] Wherein, T represents a preset period.

[0027] Furthermore, according to the current mean value signal, the current tap coefficient of the DFE is updated, including:

[0028] The current tap coefficient of DFE is updated by the following formula:

[0029] w i (k+1)=w i (k)+μ*f(k)*s(ki);

[0030] Among them, wi (k+1) represents the i-th tap coefficient at time k+1, w i (k) represents the i-th tap coefficient at time k, μ / 2 represents the preset training step size, f(k) represents the mean signal at time k, and s(ki) represents the decision signal at time ki.

[0031] In a second aspect, the present invention further provides a decision feedback coefficient training device, applied to DFE, the device comprising:

[0032] An acquisition module, used to acquire a current error signal corresponding to a current received signal entering the DFE;

[0033] The mean value module is used to average the current error signal to obtain the current mean value signal;

[0034] The updating module is used to update the current tap coefficients of the DFE according to the current mean value signal.

[0035] In a third aspect, the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the decision feedback coefficient training method of the first aspect is implemented.

[0036] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the decision feedback coefficient training method of the first aspect is executed.

[0037] The decision feedback coefficient training method, device, electronic device and electronic device provided by the present invention can obtain the current error signal corresponding to the current received signal entering the DFE; average the current error signal to obtain the current mean signal; and update the current tap coefficient of the DFE according to the current mean signal. Since the mean value of the noise is 0, the error signal can be processed by the mean method to reduce the noise and improve the signal-to-noise ratio, thereby reducing the fluctuation of the DFE tap coefficient and increasing the stability and reliability of the DFE tap coefficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 A flowchart of a decision feedback coefficient training method provided by an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of the working principle of a DFE provided in an embodiment of the present invention;

[0041] Figure 3 A schematic diagram of a curve showing changes in an original error function with training provided by an embodiment of the present invention;

[0042] Figure 4 A schematic diagram of a curve showing the variation of an error function after noise reduction processing with training provided by an embodiment of the present invention;

[0043] Figure 5 A schematic diagram of the structure of a decision feedback coefficient training device provided by an embodiment of the present invention;

[0044] Figure 6 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than 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.

[0046] The existing decision feedback coefficient training method is the LMS method, which is easily affected by noise, resulting in large fluctuations in the tap coefficients. Based on this, the decision feedback coefficient training method, device, electronic device and electronic device provided by the embodiments of the present invention can reduce the noise in the error value, and have the advantages of simple implementation, significant noise reduction, improved signal-to-noise ratio, and increased stability and reliability of the DFE tap coefficients.

[0047] To facilitate understanding of this embodiment, a decision feedback coefficient training method disclosed in an embodiment of the present invention is first introduced in detail.

[0048] The embodiment of the present invention provides a decision feedback coefficient training method, which is applied to DFE and can be executed by an electronic device that implements the DFE function. Figure 1 The flowchart of a decision feedback coefficient training method shown in FIG. 1 mainly includes the following steps S110 to S130:

[0049] Step S110, obtaining a current error signal corresponding to a current received signal entering the DFE.

[0050] In some possible embodiments, the above step S110 may include: determining a current equalization signal and a current decision signal according to a current received signal; and calculating a current error signal according to the current equalization signal and the current decision signal.

[0051] DFE includes an equalizer and a slicer, wherein the equalizer is also called a filter, which is used to process the input signal to reduce the linear distortion caused by the channel; the slicer is used to receive the signal processed by the filter and make symbol decisions. Based on these decisions, a feedback signal is generated to further eliminate the interference between subsequent symbols. Based on this, the current equalized signal and the current decision signal can be obtained in the following way: according to the current tap coefficient of the DFE and the historical decision signal output by the DFE, the current received signal is processed by the DFE equalizer to obtain the current equalized signal; the current equalized signal is judged by the DFE decider to obtain the current decision signal.

[0052] The physical meaning of DFE equalization is to use tap coefficients to simulate post-mark coefficients (the interference value generated by the symbol sent before the current moment is called the post-mark coefficient), and remove the inter-symbol interference corresponding to the post-mark coefficient from the received signal. The tap coefficients represent the weight or gain value of the filter, which is used to adjust different parts of the input signal to achieve the desired output effect. Based on this, the current equalized signal can be calculated by the following formula:

[0053]

[0054] Among them, u(k) represents the equalized signal at time k, z(k) represents the input signal at time k, and w i (k) represents the i-th tap coefficient at time k, m represents the number of tap coefficients, and s(ki) represents the decision signal at time ki.

[0055] The size of the current transmitted signal corresponding to the current received signal is called the main calibration coefficient. In a possible implementation, the current error signal can be calculated by the following formula according to the preset main calibration coefficient corresponding to the current received signal:

[0056] e(k)=u(k)-s(k)*a0;

[0057] Among them, e(k) represents the error signal at time k, u(k) represents the equalization signal at time k, s(k) represents the decision signal at time k, and a0 represents the main coefficient.

[0058] Step S120, calculating the average of the current error signal to obtain a current average signal.

[0059] In order to reduce noise, this embodiment proposes a new DFE coefficient training algorithm, which uses the mean method to process the noise in the error value. The difference between the slicer input and the ideal output is recorded as the error function, the error function contains the post-standard residue, and the noise is completely retained, wherein the post-standard residue is generally small and can be ignored. When the noise mean is 0, when the error function is averaged, one of the noise items will tend to 0. Specifically, after the error function e(k) is averaged, the noise will be significantly reduced, thereby reducing the fluctuation of the tap coefficient.

[0060] The process of finding the mean value can be divided into two categories: FIR (Finite Impulse Response) design or IIR (Infinite Impulse Response) design. If a P-order FIR design is used, P is a configurable value. Assuming that the output of e(k) after mean processing is f(k), f(k) can be expressed as:

[0061]

[0062] Preferably, a flexible P value can be used to cope with different scenarios. When the DFE tap coefficient is not stable, a smaller P value is configured to speed up convergence; when the DFE tap coefficient is stable, a larger P value is configured to make the result more accurate.

[0063] In a possible implementation, a second-order IIR design is used. Based on this, the current filter coefficient (i.e., P value) corresponding to the current training stage can be determined first; then, according to the current filter coefficient, the current mean signal is calculated by the following formula:

[0064] f(k)=(1-P(k))f(k-1)+P(k)e(k);

[0065] Where, f(k) represents the mean signal at time k, f(k-1) represents the mean signal at time k-1, P(k) represents the filter coefficient at time k, and e(k) represents the error signal at time k. The initial value f(0) of f(k) is 0, and f(1) = e(1).

[0066] Optionally, the current filter coefficient may be determined by, but is not limited to, the following formula:

[0067]

[0068] Wherein, T represents a preset period.

[0069] Step S130: updating the current tap coefficients of the DFE according to the current mean value signal.

[0070] Along In the direction of , DFE training converges fastest. Based on this, the current tap coefficient of DFE can be updated by the following formula:

[0071] w i (k+1)=w i (k)+μ*f(k)*s(ki);

[0072] Among them, w i (k+1) represents the i-th tap coefficient at time k+1, w i (k) represents the i-th tap coefficient at time k, μ / 2 represents the preset training step size, f(k) represents the mean signal at time k, and s(ki) represents the decision signal at time ki.

[0073] The decision feedback coefficient training method provided in the embodiment of the present invention can obtain the current error signal corresponding to the current received signal entering the DFE; average the current error signal to obtain the current mean signal; and update the current tap coefficient of the DFE according to the current mean signal. Since the mean value of the noise is 0, the error signal can be processed by the mean method to reduce the noise and improve the signal-to-noise ratio, thereby reducing the fluctuation of the DFE tap coefficient and increasing the stability and reliability of the DFE tap coefficient.

[0074] For ease of understanding, the above-mentioned decision feedback coefficient training method is introduced in detail below.

[0075] Due to the influence of the channel, the received signal contains the inter-symbol interference of adjacent symbols. The interference value generated by the previously sent symbols is called the post-signal coefficient, and the size of the current sent signal is called the main sign coefficient. Figure 2 A schematic diagram of the working principle of a DFE is shown in FIG. Figure 2 In, Z -1 represents a delay of one symbol period). When the signal z(k) enters the DFE equalizer, there is interference from some past symbols on the current symbol. Let a0 be the main coefficient, [a1,…,a j ] are multiple postscript coefficients, [x k ,x k-1 ,…,x k-j ] represents the sending signal corresponding to the main label and multiple post-labels, n(k) is the noise, k is the time sequence number, and the signal z(k) has the following expression:

[0076]

[0077] DFE equalization is an adaptive feedback process. Assume that the tap coefficients at time k are [w1(k)w2(k)…w m (k)](e.g. Figure 2Among w1, w2, w3), the decision output of the slicer is s(k), and the decision values of the past m symbols (m < j) are [s(k - 1)s(k - 2)…s(k - m)] (such as Figure 2 s(k - 1), s(k - 2), s(k - 3) in). The physical meaning of DFE equalization is to use [w1(k)w2(k)…w m (k)] to simulate the postcursor coefficient, and remove the intersymbol interference corresponding to the postcursor coefficient from z(k). The expression of the output u(k) of DFE is:

[0078]

[0079] u(k) enters the slicer. Assuming that there is no error in the decision result of the slicer, then s(k) = x(k). Substituting it into z(k) gives:

[0080]

[0081] If all postcursors can be eliminated, the ideal output of u(k) is x(k)*a0. Denote the difference between the input and the ideal output of the slicer as the error function e(k), and the calculation expression is:

[0082]

[0083] Therefore, e(k) contains postcursor residues, and the noise is completely retained. Generally, the postcursor residues are small and can be ignored.

[0084] Since the training of the DFE tap coefficients uses e(k) as the input, when the noise in e(k) is large, the trained tap coefficients will deviate from the ideal values, and the greater the noise, the greater the fluctuation of the tap coefficients. To reduce the noise, this embodiment proposes a new DFE coefficient training algorithm, which uses the mean method to process the noise in the error value.

[0085] According to the above expression of e(k), when the mean value of the noise is 0, when taking the mean value of e(k), one term of the noise will tend to 0:

[0086]

[0087] Specifically, after taking the mean value of the error function e(k), the noise will be significantly reduced, thereby reducing the fluctuation of the tap coefficient w(k).

[0088] The process of taking the mean value can be divided into two categories: FIR design or IIR design. If a P-order FIR design is adopted, P is a configurable value. Assuming that the output after the mean value processing of e(k) is f(k), f(k) can be expressed as:

[0089]

[0090] Flexible P values ​​can be used to cope with different scenarios. When the DFE tap coefficient is not stable, configure a smaller P value to speed up convergence. When the DFE tap coefficient is stable, configure a larger P value to make the result more accurate.

[0091] If a 2nd-order IIR design is used, the expression for f(k) is as follows:

[0092] f(k)=(1-P(k))f(k-1)+P(k)e(k);

[0093] Among them, P(k) changes as k increases to adapt to the needs of DFE tap coefficient convergence. The following function is used as an example, and can also be expanded into more forms according to needs. T is the configuration period:

[0094]

[0095] Taking the 2nd order IIR design as an example, Figure 3 and Figure 4 (The horizontal axis is time, and the vertical axis is the amplitude of the error function) shows the changes of the original error function e(k) and the error function e(k) after noise reduction during the training process. Figure 3 and Figure 4 It can be seen that after a period of training, both function curves of e(k) decrease rapidly. The difference is that after stabilization, the original e(k) maintains a large fluctuation, while the fluctuation of e(k) after denoising is small.

[0096] Next, follow DFE training converges fastest in the direction of i=1,…,m:

[0097]

[0098] Assuming μ / 2 is the DFE training step size, according to Update the direction w i (k) then:

[0099]

[0100] Where, for example, μ=10 -3 After continuous iteration, finally [w1(k)w2(k)…w m (k)] is close to [a1,…,a m ], the DFE coefficients will converge.

[0101] This method significantly improves SNR. According to the variance formula, since noise generally conforms to a Gaussian random process and has a mean of 0, the noise variance is σ 2 =E((n-0) 2 )=E(n2 ), where n is n(k). After averaging the P points, assuming that each variance is the same and uncorrelated, then:

[0102]

[0103] That is, by processing the mean, the variance is reduced to 1 / P of the original value. Note that σ 2 =E(n 2 ) represents the square mean of the noise.

[0104] according to It can be seen that the method provided in this embodiment reduces the noise to 1 / P, and the SNR is improved by 10*lg(P)dB.

[0105] In summary, the embodiments of the present invention propose a new decision feedback coefficient training method and structure, propose a feasibility analysis for reducing noise, and propose a quantization formula for improving SNR.

[0106] Based on the above content, a possible implementation step of the decision feedback coefficient training method provided by an embodiment of the present invention is as follows:

[0107] 1. Set [w1(k)w2(k)…w m (k)] is the initial value.

[0108] 2. Calculate the balanced output u(k):

[0109]

[0110] 3. u(k) is judged by the judge and the error e(k) is calculated:

[0111] e(k)=u(k)-s(k)*a0.

[0112] 4. Determine the filter coefficient P(k) and calculate f(k):

[0113] f(k)=(1-P(k))f(k-1)+P(k)e(k).

[0114] 5. Update DFE tap coefficients:

[0115] w i (k+1)=w i (k)+μ*f(k)*s(ki).

[0116] 6. Repeat steps 2 to 5.

[0117] Corresponding to the above-mentioned decision feedback coefficient training method, an embodiment of the present invention further provides a decision feedback coefficient training device, which is applied to DFE. Figure 5A structural schematic diagram of a decision feedback coefficient training device is shown, the device comprising:

[0118] An acquisition module 501 is used to acquire a current error signal corresponding to a current received signal entering the DFE;

[0119] A mean value module 502 is used to average the current error signal to obtain a current mean value signal;

[0120] The updating module 503 is used to update the current tap coefficients of the DFE according to the current mean value signal.

[0121] The decision feedback coefficient training device provided in the embodiment of the present invention can obtain the current error signal corresponding to the current received signal entering the DFE; average the current error signal to obtain the current mean signal; and update the current tap coefficient of the DFE according to the current mean signal. Since the mean value of the noise is 0, the error signal can be processed by the mean method to reduce the noise and improve the signal-to-noise ratio, thereby reducing the fluctuation of the DFE tap coefficient and increasing the stability and reliability of the DFE tap coefficient.

[0122] Furthermore, the acquisition module 501 is specifically used to: determine a current equalization signal and a current decision signal according to a current received signal; and calculate a current error signal according to the current equalization signal and the current decision signal.

[0123] Furthermore, the acquisition module 501 is also used to: process the current received signal through the equalizer of the DFE according to the current tap coefficient of the DFE and the historical decision signal output by the DFE to obtain the current equalized signal; and make a decision on the current equalized signal through the decision device of the DFE to obtain the current decision signal.

[0124] Furthermore, the acquisition module 501 is further used to calculate the current error signal according to the preset main calibration coefficient corresponding to the current received signal by the following formula:

[0125] e(k)=u(k)-s(k)*a0;

[0126] Among them, e(k) represents the error signal at time k, u(k) represents the equalization signal at time k, s(k) represents the decision signal at time k, and a0 represents the main coefficient.

[0127] Furthermore, the above-mentioned mean value module 502 is specifically used to: determine the current filter coefficient corresponding to the current training stage;

[0128] According to the current filter coefficient, the current mean signal is calculated by the following formula:

[0129] f(k)=(1-P(k))f(k-1)+P(k)e(k);

[0130] Among them, f(k) represents the mean signal at time k, f(k-1) represents the mean signal at time k-1, P(k) represents the filter coefficient at time k, and e(k) represents the error signal at time k.

[0131] Furthermore, the above current filter coefficient is determined by the following formula:

[0132]

[0133] Wherein, T represents a preset period.

[0134] Furthermore, the updating module 503 is specifically used to update the current tap coefficient of the DFE by the following formula:

[0135] w i (k+1)=w i (k)+μ*f(k)*s(ki);

[0136] Among them, w i (k+1) represents the i-th tap coefficient at time k+1, w i (k) represents the i-th tap coefficient at time k, μ / 2 represents the preset training step size, f(k) represents the mean signal at time k, and s(ki) represents the decision signal at time ki.

[0137] The implementation principle and technical effects of the decision feedback coefficient training device provided in this embodiment are the same as those of the aforementioned decision feedback coefficient training method embodiment. For the sake of brief description, for matters not mentioned in the decision feedback coefficient training device embodiment, reference may be made to the corresponding contents in the aforementioned decision feedback coefficient training method embodiment.

[0138] like Figure 6 As shown, an electronic device 600 provided by an embodiment of the present invention includes: a processor 601, a memory 602 and a bus, the memory 602 stores a computer program that can be run on the processor 601, when the electronic device 600 is running, the processor 601 and the memory 602 communicate through the bus, and the processor 601 executes the computer program to implement the above-mentioned decision feedback coefficient training method.

[0139] Specifically, the memory 602 and the processor 601 can be general-purpose memories and processors, which are not specifically limited here.

[0140] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the decision feedback coefficient training method in the above method embodiment is executed. The computer-readable storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a RAM, a magnetic disk or an optical disk, and other media that can store program codes.

[0141] The term "and / or" herein is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can represent including any one or more elements selected from the set consisting of A, B, and C.

[0142] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limiting, and thus other examples of the exemplary embodiments may have different values.

[0143] The flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and a part of a module, a program segment or a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0144] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0145] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0146] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A decision feedback coefficient training method, characterized in that: Applied to a decision feedback equalizer DFE, the method comprises: Acquire a current error signal corresponding to a current received signal entering the DFE; averaging the current error signal to obtain a current mean signal; According to the current mean value signal, current tap coefficients of the DFE are updated.

2. The method according to claim 1, characterized in that The obtaining of a current error signal corresponding to a current received signal entering the DFE includes: Determine a current equalization signal and a current decision signal according to the current received signal; The current error signal is obtained by calculation according to the current equalization signal and the current decision signal.

3. The method according to claim 2, characterized in that The determining, according to the current received signal, a current equalization signal and a current decision signal comprises: Processing the current received signal through an equalizer of the DFE according to the current tap coefficient of the DFE and a historical decision signal output by the DFE to obtain the current equalized signal; The current equalized signal is judged by the judger of the DFE to obtain the current judged signal.

4. The method according to claim 2, characterized in that: The step of calculating the current error signal according to the current equalization signal and the current decision signal comprises: According to the preset main calibration coefficient corresponding to the current received signal, the current error signal is calculated by the following formula: e(k)=u(k)-s(k)*a0; Among them, e(k) represents the error signal at time k, u(k) represents the equalization signal at time k, s(k) represents the decision signal at time k, and a0 represents the main coefficient.

5. The method according to claim 1, characterized in that The step of averaging the current error signal to obtain a current mean value signal includes: Determine the current filter coefficient corresponding to the current training stage; According to the current filter coefficient, the current mean signal is calculated by the following formula: f(k)=(1-P(k))f(k-1)+P(k)e(k); Among them, f(k) represents the mean signal at time k, f(k-1) represents the mean signal at time k-1, P(k) represents the filter coefficient at time k, and e(k) represents the error signal at time k.

6. The method according to claim 5, characterized in that The current filter coefficient is determined by the following formula: Wherein, T represents a preset period.

7. The method according to claim 1, characterized in that The updating of the current tap coefficient of the DFE according to the current mean value signal includes: The current tap coefficients of the DFE are updated by the following formula: w i (k+1)=w i (k)+μ*f(k)*s(k-i); Among them, w i (k+1) represents the i-th tap coefficient at time k+1, w i (k) represents the i-th tap coefficient at time k, μ / 2 represents the preset training step size, f(k) represents the mean signal at time k, and s(ki) represents the decision signal at time ki.

8. A decision feedback coefficient training device, characterized in that: Applied to DFE, the device comprises: An acquisition module, configured to acquire a current error signal corresponding to a current received signal entering the DFE; A mean value module, used for averaging the current error signal to obtain a current mean value signal; An updating module is used to update the current tap coefficients of the DFE according to the current mean signal.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the decision feedback coefficient training method described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the decision feedback coefficient training method according to any one of claims 1 to 7 is executed.

Images